Don't forget Microsoft | Hacker News Reader

Don't forget Microsoft

luttig.substack.com

429 pointsSamvitJ453 comments

I feel like the elephant in the room is culture.

* Facebook seems to be a bunch of smart people working on pet projects. Monopoly profits drive a political empire where people at the top think up something random, and it gets built.

* Google has customer contempt. They started with brilliant people who were used to being smarter than everyone else. They also started in algorithm-driven markets like search and ad-words, where everything was statistical and individuals didn't matter. They've lost the smarts and the ethics, and they're in a bit of a hole. I think they've reached the end of the growth line.

* I know nothing about Apple. Too secretive.

* Microsoft has a bunch of cut-throat teams, competing with each other. Their technology is middling. However, they're the only one of the bunch you'd want to partner with for B2B.

* ... except for Amazon, which is hyper-customer-focused, and has a track record of successful forward-looking projects. AWS has been rock solid. On the other hand, I'd never want to work there; they treat employees like crap. But it somehow works out for them.

This cartoon still holds much truth:

http://joyreactor.com/post/287724

Please try to find the original (in this case, it's from Manu Cornet's webseries Bonkers World: https://bonkersworld.net/organizational-charts)

Thanks!

I wasn't able to hunt that down.

Microsoft is an enterprise company, has a bureaucratic manner and bureaucratic ways of doing things, for example going all in on the XML craze of the late 1990's and early 2000's. There's a lot of ceremony in their projects, TypeScript being a notable exception (at least one that comes to my mind.) It reminds me of a complaint that is interesting to me about India being in love with bureaucracy and paperwork [1]. Somehow it all works and I think they are making strides to be more nimble, but I wonder if it's part of Microsoft's DNA.

[1] It’s 2021 and the Indian bureaucracy remains the greatest impediment to progress

https://theprint.in/opinion/its-2021-and-the-indian-bureaucr...

Not knowing much about India, but excessive gov. bureaucracy is often a just side effect of rampant corruption.

It goes both ways:

- people wanting to fix the corruption problem through technical means will make it more difficult for individual clerks to make decisions alone, and push for more paper trails.

- the harder the system is to navigate, the easier it is for clerks to get bribes and quid for pro. At some point it can become impossible to get anything done without bribing to accelerate or bypass the checks.

I'm talking more about people who like rules, regulations, organization and order as part of their nature. People who keep their ducks in a row.

I find it pretty funny that you would attribute something to the nature of a people after getting a well reasoned explanation of why people act that way and what the effect is, without even going into the argument.

You are right, I didn't even get involved in the corruption argument. If you think I said something about it, I didn't. It's not interesting to me. You say it's well reasoned but I have no opinion it. The commenter is probably right about corruption and bureaucracies. I was talking about Microsoft. There isn't any corruption of that sort at Microsoft. Maybe in India, I don't know. I was comparing them as organizations that are bureaucratic in nature. I find it interesting that some people like that kind of system and create it, because I don't. To me it seems stifling of creativity and discovery. But maybe I'm missing something and could use more rules and order in my life. I'll keep studying it. I'm glad you pointed out that I was misunderstood.

Not always. Bureaucracy can also result from a risk-averse culture.

That's almost always the case. Most corporates are risk-averse and thus bureaucratic.

The impulse to increase bureaucracy comes from risk-aversion, but you always have other forces to balance. For instance cash transporters are risk averse, but they haven't build secure underground tunnels under every shop and ATM.

Ever increasing bureaucracy probably doesn't come from ever increasing risk-aversion, and more from a dysfunctionning in the balancing mechanisms. Basically it needs to benefit a large portion of the system to keep creeping up.

i wouldn't really form opinion on print's article. I won't criticize it either. Just, refer to couple more places too. Print is often agenda driven than news

Would you mind expanding on the "ceremony" around Typescript a little, please? Or if there's an article which explains this that'd be great too. That statement interested me, is all!

Parent said "TypeScript being a notable exception"

I am not sure if you are claiming that MS is bureaucratic inside or the tools and applications they create are hard to use. If we are talking about the inside then it does not seem to have stopped them from becoming gigantic. If you mean that their tools and products are hard to use then I think it differs a lot between products. VSCode is for example very easy to start and use. GitHub seems to get better all the time and is also easy compared to other similar tools and so on. You could say that things like Office is hard but I think they are easy to start using. The hard part is learning all/many of the advanced features and that is ok IMHO.

I'm looking at it from a programmer's point of view, not a users.

Any organization that designs a system (defined broadly) will produce a design whose structure is a copy of the organization's communication structure — Melvin E. Conway

https://en.wikipedia.org/wiki/Conway%27s_law

I'm saying Microsoft is "bureaucratic" in it's nature. I'm not saying it can't change and I do recognize it has been improving in many areas. I celebrate their success. I used to work there. I didn't think what I am saying is controversial because Microsoft itself recognizes this and is taking many steps to become more open, nimble and to iterate development faster.

My thinking is not original and probably out dated - posted for those who haven't heard this before

Inside the Bureaucracy That Crippled Microsoft

https://www.theatlantic.com/technology/archive/2012/07/insid...

There is much truth in this post.

Perhaps controversially, I think Microsoft's real "B2B" success comes from treating end users like crap. Arguably their only successful "end user" product line is Gaming, and that (on Windows) has grown with comparatively modest investment on Microsoft's part or with phenomenal (bought, third-party IP-based) software and a bucketload of cash (Xbox division).

Microsoft's true success in the business world comes from ultimately targeting managers – be they in an IT department or otherwise. They are afforded real power over users. GPC can deny a user the ability to do something with their computer that other users can do. Permissions can be very finely grained. Office has, for many years, come with some form of document-based DRM. Microsoft as a company seems to love DRM; they invented activation, for god's sake. The net result of this is that they know what managers like, which is, in fact, to manage. They centralise. They absorb. They have created a huge, confusing, proprietary computer-verse completely orthogonal to the rest of the world (except inasmuch as the rest of the world needs to interact with it).

A lot of their product line boils down to implementing MBA newspeek in disguise: clearly; writing good web software is difficult and a good business-orientated app features database, GUI and user-interaction parts – thus Dynamics CRM is born. It's not the business's "core competence" to redevelop those skills -- they're hard. Tie it in with their other B2B offerings and you have central control over both employees' performance, and customer's offerings, in a very tailored, swish way. Sure, other companies can do this (looking at you, Salesforce) but this is just one example. Salesforce can't also do, say, long-term archival storage on Azure for less than a cent per GB, and won't provide you with a calendar-and-meeting-filled video-chat app that is "free" with your existing subscriptions.

All of these things tick boxes. They provide "return" on the "investment" of paying the microsoft tax. The broader their product offering, no matter how shit it is to the poor saps that have to use it, the cheaper it effectively becomes. That's the true success of microsoft: they've always been big enough to do things well enough, conglomorate enough stuff, and amortise its effective cost over lots of different offerings. The net result looks good on an Excel spreadsheet, features that managers care about are there - and they keep getting business. (And end-users be damned!)

>GPC can deny a user the ability to do something with their computer that other users can do.

Yes, this is for security and management. Pretty important stuff. We don't just set GPOs to be dicks.

I wish more people in corporate IT thought like you did. Sadly, many seem to forget that availability is the first pillar of security.

Of course! Always invoke Hanlon's Razor.

Ah yes, IT is big dumb dumb I forgor. durrrr

If you want to make the equivalent of "You are stupid" comment, go back to /g/

I mean if you are incorrectly locked out of something, then don't assume malice.

>We don't just set GPOs to be dicks.

That's not how it is perceived by users. I don't think I've come across a big organisation where IT wasn't seen as an obstacle to getting things done.

Yeah but that's because "getting things done" for many people involves stupid stuff like installing malware-infected warez copies of Photoshop because they can't be bothered getting approval to expense a copy.

I used to think IT departments were dicks, until I worked at Google and went to a tech talk by their WinOps divison (what's called IT in every other firm). They were explaining why they were transitioning Windows users to binary whitelisting - literally not a single EXE runs unless it's whitelisted by IT. I thought wow, how tyrannical, that's surely a Dilbert-esque IT power trip.

And then they told us about all the stupid stuff people did with their Windows desktops. There was literally nothing so ill advised people didn't try it, and even worse, those people were sometimes very senior engineering executives. You might think such people would know better but ... no. Also, engineers aren't any more immune to phishing than anyone else, it turns out.

Pretty sure they did the same to MacOSX systems, too, with custom kernel module even

Isn't the whole drive towards zero-trust network configurations to allow BYO device to work, i.e. to assume that every device will be compromised and plan accordingly? Seems much better (to me) than crippling the desktop environment of your employees and hobbling their productivity.

Not sure how you could accomplish that safely without crippling users in different ways

You don't override people's home pages for security.

Just FYI that the Microsofty way to say this is “You have failed to provide sufficient evidence that the GPO policy set as a result of work units executed by my business group were to be dicks”.

> they treat employees like crap

The local Seattle area is full of Amazon employee millionaires.

Somebody has to be at the top of the stack ranking.

I don’t think that disagrees too much with the hypothesis. There’s plenty of jobs with good pay but will double the rate of your aging — Amazon is one of them. If you’re someone who has the skills and can put up with the culture then collect that paycheck.

Right now today I could switch companies and make about 30% more if I was willing to take a corporate job. I don’t because I have a great work life balance and the money just isn’t worth it right now.

> will double the rate of your aging — Amazon is one of them

Amazon workers have a life expectancy of 50? At least try some plausible arguments.

hy·per·bo·le /hīˈpərbəlē/

noun

exaggerated statements or claims not meant to be taken literally.

People work for Amazon from age 5 until they die? If you're going to be pedantic about a hyperbolic figure of speech, at least think it through first.

Normal life expectancy is 80y. If you start work at 20y, you can expect to live another 60 years. Cut that in half to 30, and you die at age 50.

> at least think it through first

Back to you.

The life expectancy of people with similar living conditions to Amazon's high-paid employees is higher than average. The correct comparison is with people who have similar living conditions, except significantly lower stress levels, which is higher still.

Which is besides the point. Who works for Amazon for 30 years? Amazon hasn't been around for 30 years, and their turnover is high enough that I expect it'll be a while before they have any 30-year employees, excluding the owners. So Amazon workers probably wouldn't have a life expectancy as low as 50, regardless, even if working there did make your body degrade twice as fast.

Do you have any evidence at all that Amazon workers have any reduction in lifespan, let alone 30 years?

Amazon has been around 25 years now. If their workers had their life expectancy cut in half, the vanguard would be dropping like flies today.

I have no evidence that the hyperbolic figure of speech is literal truth, no. I do have evidence that people in high-stress environments have health complications.

Is it really that high stress?

My dad flew 30 some missions over Germany. The cohort he went over with had an 80% casualty rate. Some missons had a 10% death rate. He could do the math, and assumed he was going to die.

(Obviously, he did not die, since I exist, but he came literally within a couple inches.)

When he arrived back in the US, he was bemused by all the stuff back home people worried about. He thought "what the heck are you worried about, you're going to live another day!"

Stress is a physiological reaction to certain stimulus. You can stab people with blood monitors and conclude that yes, it is a high-stress environment.

Nobody hired into Amazon now is going to become a millionaire. Their total comp packages are the smallest of the MAGMA and their stock has only 3x'd in the last seven years.

To become a millionaire, you have to get into a company where your equity will 10x or more. You'll have to pay cost of living, capital gains, and you'll probably want to diversify a bit along the way.

For this to happen with Amazon, inflation will have to be wild, or they'll have to enter and win a lot of new markets. I can't even image a $10-20T Amazon.

Take your $120k a year job at Amazon, buy a $800k house in Seattle, hold for a decade and your a millionaire with the equity in your house alone, nevermind the $1.2 million you earned at Amazon during that time (and what you chose to invest it in).

That's assuming 100% of everything you earn is saved. Cost of living is high in Seattle. You're also in a higher tax bracket. How much of that income will you actually be saving?

If prices remain fixed, after ten years, you have about 25% equity in your home (not including the down payment, which makes it closer to 40-50%). Assuming the housing market keeps going up, then maybe you'll be a millionaire by this point with just your home equity.

But this assumes you aren't renting. You first have to put up $200k liquid for the down payment. New hires won't have this right away, and the housing market will keep getting more expensive in a world where "home equity makes you a millionaire".

I'd still wager that only the most frugal will become millionaires within ten years.

Amazon isn't some magical entity making it happen either. You could attempt this strategy with any engineering job in Seattle.

To become a millionaire, you have to get into a company where your equity will 10x or more.

Maybe if you want to be a millionaire at 25, but a career is longer than three years.

If you start at $150k, get a raise to $250k within 3 years, and $325k by 6—-$1 million net worth should be very very achievable before retirement.

I started at $50k, never made more than $200k in a year. My net worth will be a million well before I'm 65.

A million isn't what it used to be.

Right. The term millionaire lost its original meaning and nothing new has replaced it. Even if we go back to the 1820s when the term was coined, the inflation rate hasn’t been nearly high enough for billionaire to work.

Decamillionare is probably about right but it’s a mouthful.

A million won't do much in maybe 10-20 cities across the world. Outside of that, a million dollars liquid is still life changing, retirement money

1 million at a safe withdrawal rate of 4% is $40,000 a year. Portland, Oregon in the US is somewhere around the 20th most expensive city to live in. You could survive in Portland on $40k/year but not exactly in some enviable lifestyle.

> AWS has been rock solid

AWS is a hodgepodge of stellar, middling and some downright crap services. I think their strategy is to make a million flowers bloom and see what takes root. Unfortunately it's hard to tell at first sight which services offer a good experience and which are half baked. In my past experience teams tend to learn this through trial and error.

This. Read the docs, do a tutorial, start implementing, run into obscure issues, performance problems, or missing features on Stack overflow that are unresolved years later.

I’m curious how confident you are in this assessment of each company, and how you came upon it. I’ve worked for several big tech companies, some named here, and my insider perspective is these entities are so huge that you can’t really boil it down to a simple explanation.

Put another way, is this a well researched opinion held after discussions with a variety of roles and time periods? Is the sample size enough to think it’s representative?

You’re on hackernews, so the answer is likely no.

out of curiosity, where would be the answer yes?

> Google ... reached the end of the growth line.

I think they still have some things set in motion during their old culture that have a chance grow up big.

But it does seem like their best bet would be to move all the good parts of Alphabet far, far away from the decaying center.

having their war chest means they have a chance to 'grow up big' (not sure what means given they're already one of the biggest companies ever). They could fund entirely new businesses without issue, so there's always a chance.

This is a good and thorough article. The author got down to brass tacks pretty quickly and brings up interesting hypothesis about $MSFT.

That said, I do have one gripe:

> To oversimplify Notion to its demographics, it is Office 365 for people below age 35.

I recognize this is an oversimplification, but even so, it seems like a stretch. Notion is a decent product, and I have used it for a few small-scale team projects in uni (mainly for Kanban-related stuff) - but to call it a replacement for O365 is an exaggeration at best.

Yes, you can have pretty, nested documents in Notion and that's great, but a tabular database in Notion is by no means a replacement for Excel or even Google Sheets. The velocity that is afforded by Excel in terms of formulas is unmatched and there's a reason it has yet to be unseated as the kingpin of modern finance.

Most young people I know use a combination of Discord + Google Suite to collaborate. I am aware this is slightly anecdotal, but I am also having a hard time imagining myself as a founder and then asking my CFO to use Notion to prepare investor pitches.

Source: Am 23 :)

> Discord + Google Suite

I use shitton of MS products, but I cannot switch from Google's gmail, docs and drive to anything.

I hate MS login page and endless redirects between their services

I cannot explain what puts me off, but Google's (gmail) login page and way better switching between apps feels way better.

Also I have feeling that MS account is more "formal", idk how to explain it.

God, if you have personal and work accounts under the same email bring on the insane pain. There is sometimes confusion on the MSFT backend (or was) in these cases and you'd get jammed into weird corners (ie, reset would do just your personal so you couldn't ever get a work account reset etc).

It's something about their federation tech vs google. Google, if you are in wrong service, nice switch account, one redirect it feels like.

Exactly. this is such a pain, especially if you happen to be apart of one organization that's AAD Connect and one that's full AAD (Azure Active Directory).

As far as I can tell, the easiest way to switch users is to go to https://www.office.com/login?es=Click directly (?es=Click is important) which will present you with all the accounts you're 'Signed in' with, even if office.com has you locked into one. Then you can use the app switcher to switch domains, which ensures the correct account.

I thought this was bad, then my work changed from on-prem AD to Azure AD. Now my account is "linked" to my computer, and the only way I can log into a different Outlook inbox is by opening an incognito window.

It really does feel like something is broken on the back end of Microsoft's SSO, but I'm sure it's just so complicated with decades of cruft that it's hard to dig out.

Same here but with the added pain of an early "onmicrosoft.com" account that splits our users between their domain and ours, doesn't let me use my work domain address, and yet denies the onmicrosoft.com one access because it's not a "work or school account".

It's seriously bemusing to be logging in to an organisational account with admin access only via a personal Gmail account username but there's literally no migration path for these early adopter accounts.

I wonder if this is a generational thing. If there's a decent app for a thing, I use the app¹. In this case Outlook. I have 4 email addresses to manage my personal email, throw-away registrations, website comms, and website SSL/admin.

I never access any email ever from a browser - Outlook is unbelievably good, and with the .PST file size restrictions no longer being restrictions I have my entire email archive of ever in it. Been doing things like this since 2002.

¹ Social media is the exception. Where I have multiple social media accounts (Instagram) I use seperate browsers (Edge for hobby account, Firefox for personal account). This is not ideal but the workflow is easy.

Have you tried multi-account containers in Firefox? I've found it less painful than multiple browsers.

Ye there is something fishy about Office365 login. Sometimes the account switch for no good reason. Even if I just entered the username and password in a prompt.

It is a great mistake login in on different accounts on the same computer ...

> I hate MS login page and endless redirects between their services

I dread logging into any Microsoft product for this reason, I just know it’s 30 mins of BS trying to find different passwords for different accounts, none of which relate the thing I’m actually trying to do.

And it carte blanch refuses to work in lynx or links too, even if the site behind an SSO login would work perfectly. Bah.

This is getting downvoted quite a lot. I don't understand why. There are lots of legitimate uses for a text-based browser, from quickly looking something up on an SSH session to trying to script something. Servers are often placed behind an SSO wall and Microsoft's authentication regime means that you can't, for example, curl server/data without passing a token coming from a graphical browser and a whole world of pain. This is explicitly a design decision from Microsoft: you need to support javascript to use their SSO login.

A password manager and the Authenticator app makes it dead simple

It's really not. I have no problem remembering or calling up the passwords or emails for my personal, Xbox, current work, previous work, or previous student (!) accounts. And yet Microsoft somehow gets them confused; I haven't ever signed in with this PC or from this ItoP address to my student account, the email no longer works, and it obviously can't access my work document...but something is tied together there and MS insists that must be the account I want to use. Private windows don't always fix it either, there's something shady going on behind the curtain.

Contact support and let them know you have multiple accounts and these issues. I thought they were going to be clueless but they were able to resolve some of these issues quickly behind the scenes.

It does not. I have some combination of my work, school and personal microsoft accounts signed in across a couple different computers. It seems to randomly pick one based on some really poorly inferred context. For example, going to outlook.office.com will (usually) load my personal Outlook email. while going to outlook.office365.com (which redirects to outlook.office.com!!!) always loads my school email.

All of my accounts are meticulously password-managed. The ONLY Google property with this problem is Forms. With Microsoft, the entire authentication system is a disaster. Did you know you can sign into your microsoft account with a security key if the moon is waxing crescent, your computer is at least 3 OS updates behind, and you recite a secret incantation?

I got locked out as an admin of a Microsoft 365 workspace while using password manager. I still haven't recovered the account. You can only recover your account by contacting their support on mobile but they are clueless.

I have a feeling they truncate or santise password somewhere.

I switched to gsuite immediately.

As much as I hate using Azure, the integration of DA makes it so simple. The login screen just lets me select which account to use. That's it.

Flashbacks to when i started a new job and just got the credentials for microsoft 365 but for some reason even after logging in through my work email the service seemed to be tied to my college email had to open an incognito tab for it to work

Anecdotally, O365 products always felt like disjointed attempt by various engineering and sales teams to put a modern spin on the client side offering. In these contexts, it never had to be good - but just similar enough to the old products that excel, word, and outlook power users would use it and find value.

GSuite, Notion, Confluence, and others appeal to an audience that either never required MSFT office nor felt the need to become power users in it.

I use gsuite for my own needs, but unfortunately whenever I deal with anyone on the non-technical side (marketing, support, etc) I always need to switch to word/excel/powerpoint.

I have been waiting for this to change, but I can't yet free up the gigabytes of space on my work laptop that is currently occupied by MS software.

I don't know why people like gmail so much but I found gmail's design, or Google's Material language in general, really bloated. Outlook.com has a flat and simple layout. Of course it's inferior to gmail in search and I don't understand why they makes me type out the "@outlook.com" part at login.

Another reason I use outlook is that I already gave everything else to Google so just diversify a bit.

>I don't understand why they makes me type out the "@outlook.com" part at login.

Probably due to that's not the only address they support still - I've got a hotmail.co.uk account and outlook login page for me.

Another thought is universities and other groups use outlook as their MX signing in via outlook.com which can then redirect

Yup - my old LSE e-mail was via outlook.com

gmail has the same thing, but username will autocomplete to @gmail.com unless otherwise provided

If only they had different websites or domains to use for those different login pages.

> but I cannot switch from Google's gmail, docs and drive to anything.

Until you:

- Use Google Drive and its subpar desktop experience of sync'n files (OneDrive is far worse, but Dropbox+MS Office beats it out)

- Experience Gmail's hostile user stance against non-Google calendaring. (1) Automatically creating Google Meets for every invite (or did they fix this recently?) and (2) the terrible formatting that gets sent to non Google-based accounts that gives me the "fingers on the chalkboard" feeling every time it gets sent to my O365 based email.

> I hate MS login page and endless redirects between their services

100% agree.

> Also I have feeling that MS account is more "formal", idk how to explain it.

Agreed. And I feel like anyone who sends me email/cal invites from GSuite are essentially amateurish. Can't explain it either.

Desktop experience? This isn't the 2000s grandpa.

There is no web substitute for Excel yet, especially when you get into bigger files.

Because at web scale you don't use individual files.

> I hate MS login page and endless redirects between their services

whenever I login to Gmail it redirects me to YouTube and then back to Gmail

> I use shitton of MS products, but I cannot switch from Google's gmail, docs and drive to anything.

Here's my libre stack:

  - Thunderbird for e-mail (whichever provider you like, i also self-host a mail server for automation etc.)
  - LibreOffice, which i use as a local piece of software, because web office apps rub me the wrong way
  - Nextcloud, which i also self-host and which has both desktop software and mobile apps

Of course, that's not something that everyone might want to use for themselves, but personally it has worked out pretty nicely for me, allows me to keep my data private (for the most part) and is a rather cost effective way of doing so!

It is a running joke in our team that MS cannot even implement a login page without problems.

To be fair, the authentication for Microsoft is far more complex than any other comparable system. They have to deal with their own Microsoft accounts (Outlook, Live, Hotmail...), federated domains (via ADFS or third parties) and domains hosted in 365 too - Some of those being synced passwords from AD, some of those using pass through authentication from local servers...

Google just has to deal with accounts with Gmail domains and third party domains that they host.

I totally get that. This is why it would make sense to implement login a way that it is obvious to the user what is happening. I am not sure how much time I wasted on trying to figure out which system I am logging in to and with.

Yes, the article is good but that's where it goes off the rails. There's lots of talk about how MS has 96,000 "talented engineers". Yeah? Where are they then?

My experience of Microsoft products in the past decade, both Azure and Windows, as a developer, has been extremely poor. The decay inside the Windows org has been very sad to see in particular. Windows was never exactly bug free but it had a certain robustness about it simply due to the sheer weight of apps using it. The Win32 docs were extremely verbose but mostly did tell you what you needed to know. But, those apps have been evaporating for 20 years and it shows.

From the perspective of a developer trying to do stuff on Windows in 2022 is an exercise in frustration. Nothing modern works right and their developers don't seem to know or care. Trying to do absolutely basic tasks using their recommended approaches will yield an un-ending stream of stupid, impossibly basic bugs. To name just a few bugs I've hit in recent times: they've managed to screw up things as simple as taskbar icons, downloading files from web servers correctly and restart apps after an upgrade. I never thought I'd find myself actually liking Win32 but it does at least tick all the boxes and the standard code paths are bug free.

Nothing about this experience radiates experience or talent. It leaves you with the constant impression that everyone working on the Windows team is a new grad who learned C++ 3 years ago and is shielded from the reality of what their customers experience by a wall of even less talented program managers whose primary job is to post vague reassurances and empty promises to (badly implemented) web forums.

That's Windows. Azure was little better. Again my experience was one of incredibly simple bugs in basic functionality, like holding TCP connections open for more than a few minutes. The fact that people keep finding "root@azure" bugs is also indicative, because usually these reveal that there's no defense in depth of any kind.

What you’re experiencing isn’t the result of bad engineers, it’s the result of business decisions that prioritize new features of testing. There are many talented engineers in Microsoft (as well as some terrible ones) but they don’t have full control over the goals of the business.

Exactly this.

For my professional stuff I’ve actually dumped O365 and Google.

I’m using Apple stuff only (pages, numbers, keynote, mail, calendar, reminders, icloud). It’s completely different to anything else and there are a few small compatibility issues but it works really really well and doesn’t stab you in the face with complex problems. Add zoom and slack and you have enough interoperability.

I set three simple standards that I will not budge on which ensures this solution will remain working. Documents via PDF only. Copy via cut and paste on slack. Data via CSV and JSON only.

I also only use spreadsheets internally to my “partition”. They are shitty for everything else.

I’m not going into the long time rants with my extensive experience with O365 and GSuite and LibreOffice here but this feels like I’m being shafted the least hard at the moment.

I think Excel is the main moat on people/company keep using Office. Even the spreadsheet product on macOS is unmatched. I think it's the defacto tools in the salaryman-world for any backoffice work requiring some spreadsheet app.

I know an avid iOS fanboy that uses Excel on Windows for their main job, it's on a different level than its macOS counterpart.

Yes that’s true. There are better ways to solve all the problems that spreadsheets solve but no one way of solving them all in one place as badly that’s quite as good :)

That hurt to write.

Zoom and Slack are defacto standards which have similar issues as O365 and Google.

That said, I love Numbers. It has some features like easy to use formulas right out of the box. I'm not able to do it that quick/easy with any other spreadsheet product.

However, given I don't use Mac on work as now, I'm bound to Microsoft Office and Microsoft Excel.

> I use shitton of MS products, but I cannot switch from Google's gmail, docs and drive to anything.

This is how other people feel about the Office suite FWIW.

I'm usually in Teams all day. It's our VoIP, file share, and chat. We have scheduling, notifications, interactive web cards, and multi-user document editing all in there as a single interface.

I get that that it's a duck-on-water front end to OneDrive, SharePoint, Skype-for-Biz, etc... but it works well enough to get things done.

I can't stand google docs myself, or using gmail for actual business emails.

I'm stuck on Notion since my much older CTO thinks Notion is 'hip'. I think it's more fair to say Notion is just the counter culture solution to the same problem O365 and G-Suite try to solve. I suppose Notion kind of made sense back when were tiny, but these days Notion is just clunky and has one of the worst search functions I have ever had the displeasure of using.

At least it apparently isn't doing well in our security audit and the security team is telling us G-Suite is in our future.

I’m there with you on the search function. I have no idea what its result ranking algorithm is, but I’d love it if the top results were ones with the exact search string. Instead, we get “You searched for AWS. OK! Here are pages with awl, Australia, House, lasagna, [8 pages], Google Wave, and AWS.”

I haven't used Notion much but if the search is worse than G-Suite, that is bad indeed. G-Suite search is so bad I sometimes wonder if I'm being trolled. Google is supposed to be king of search, yet I can type in one of the words in the title of a document that I was editing a few days ago, and it won't find it, but it will find some obscure and completely unrelated string in a csv file from 2014 that I haven't opened since I uploaded it. To work around the horrible search, I literally keep a google doc that starts with AAAAA so it will always be at the top that has a bunch of links and keywords in it. When I need to find a document or spreadsheet or whatever in google drive, I Ctrl+F in there and it's a life saver. Also seems ridiculous that I have to keep my own index file in order to find anything.

Try cloudsearch.google.com. It works much much better than the GSuite search.

That's pretty nifty, thank you!

(But isn't available for G Suite Legacy, for anyone idly curious.)

I can't imagine search worse than teams (which often fails on the literal exact text from earlier that day). I'm so sorry.

Don't even get me started about problems in Notion. It is so depressingly bad that it makes me hate one thing I really like about my work: writing. The search is just broken, but even more broken is the user interface for writing. All kinds of weird things happen if you click your text or start a code block and move your cursor to the wrong way.

What I do nowadays is I iterate my texts in Org Mode, get the comments on GitHub and when we're all happy about the text, I import it to Notion and lose the writing to the black hole of Notion search forever. Now when they have a public API, I'm thinking could there be an Emacs plugin that could sync the text between Notion and Org Mode, that would let me keep my texts in version control? I guess that's just a matter of time...

I feel like Notion and GSuite (Docs mainly) solve two very different problems.

The thing that Notion brings (which I found hard to work with in Gsuite) was that shared documents aren't owned by any one individual, and copies are far less likely to proliferate. So its good for shared documentation that an Organization needs, as opposed to two people collaborating on a document, which google docs handles fine.

The basic tables and cards and stuff are nice for many things but are not a substitute for sheets by any stretch of the imagination. And Notion has no analog for Slides.

It depends on what O365 means to the user.

A lot of people (me included) aren’t using Excel beyond opening the files to look at the graphs or entering data in predefined fields. It’s the usual 20/80 split between the people actively creating and managing the data and the ones on the receiving end.

For a long time Microsoft pushed Office adoption by forcing the “viewer” to have licenses to access the data. That’s why a company would buy Office for every single employee. But now, you can have Office for the “20” part, and have them push their result to Notion for everyone else to view, do simple things with it.

Notion can effectively be a O365 replacement for the proverbial “80” part.

(Basically O365 is becoming a “power tool” and Notion can skoop the “casual users” slice. I see G\Suite inbetween)

No, because Office 365 is also the email hosting , MS Teams , Windows 10 enterprise license and management of the computer with Intune.

So you still need Office 365 for each user.

And then, why pay for 2 tools when you can pay for 1.

Don't forget optional Power Apps!

This is the biggest reason MS remains a huge player though. Whether you're using Microsoft's cloud e-mail or an onprem exchange, they'll be more than happy to figure out a CAL bundle that is compelling.

Email is pretty important. You really think a company would pay for Office suite, Notion and Gsuite? You are insane.

And CAL is not used for Office 365 sweeite

I'm saying the exact opposite, that they won't.

> And CAL is not used for Office 365 sweeite

No, it's not, but they are used for exchange last I checked. Perhaps my verbiage was a little inaccurate, but thank you very much for the downvote and condescension.

You are right in that Microsoft is the giant pillar sustaining a ton of big and small companies, I'd randomly guess more than 90% of them.

Macs and GSuite are more of a long tail that anything, but these would precisely be the companies that would go to Notion I think. They will have way less incentive/lock in to push Office365 on the bulk of their employees.

> For a long time Microsoft pushed Office adoption by forcing the “viewer” to have licenses to access the data

There have been free, no-license viewers for MS Office files since the early 00s, if not earlier.

https://support.microsoft.com/en-us/topic/supported-versions...

I'd wagger these were just excuse programs aimed at releasing regulatory pressure. Employees working on them might have put their best efforts, but they were inherently flawed and limited beyond just the "viewing" aspect.

A long time ago of my employer tried to skimp on a license and we tried those instead, just to give up after a day and buy the whole package.

I wonder how many people just had the same course, and we were visibly not alone: https://www.mrexcel.com/board/threads/using-macro-enabled-wo...

They couldn't enable viewing of truly crazy files without shipping the rest of the elephant with it.

And Word users never listened when told to only share RTF files instead of DOC files, except maybe early on when dealing with how Mac and Windows version files were mutually incomprehensible (RTF was kept at feature parity and documented all the way to Word 2003, precisely as interchange format)

A lot of people didn't understand there could issues at the base level. They were exchanging files with their colleges on IT provisionned machines and didn't have to mind what they were using, on a day to day level they had no issues, it was the people "outside" that were a PITA always boyhering them.

Also biting the bullet and being on the latest version of Office with everything from Access to Outlook installed sadly covered all the bases. We worked on an intranet for a smallish company and they had web components that only worked with the specific Office DLLs installed. Took us an afternoon to figure out why it didn't work on our machines.

Lotta people make this mistake. They think Notion, or Google Sheets, or Slides is the only tool they need. In truth they need all three, depending on the task at hand.

Lots of product managers seem only capable of thinking in slides when Notion would be better for documentation. I've received 500 pages of content for review written into CSV file for some insane reason. And yes, I've seen people struggle to run mathematical models in Notion when Excel is right there.

I try and remind people not to use a hammer when you need a screwdriver, but sometimes that's all they know.

For what it’s worth, Loop[0] is an experimental Notion compete that seems to be targeting its demographics.

[0] https://www.microsoft.com/en-us/microsoft-loop?ms.url=micros...

IMHO the most important period for microsoft is the 90s. There is nothing about it in this article. 1990 is windows 3.0. In 2000, it was the start of Ballmer's era with the erratic management.

There's a pattern underlying the claim:

The reasons why the modern productivity suite is winning (Notion, Coda, Airtable, etc) is because they've embraced the web and collaboration.

- ability to share the original source to people

- tag people right inside the document.

- set tasks to people and track them from within the doc.

- new age collaboration features - likes/comments/hashtags built-in

- ability to tag a document within another document

All these are small steps towards embracing the web's nature deeply inside product instead of just shipping the native features to run over the web.

But there's a catch, these modern solutions tend to lag a bit over the powerfulness of what native offers, like

- powerful charting capabilities

- the calculation engine of excel

- pagination of documents

- powerful set of formatting features

Interestingly it's much easier to build the new age features into the already powerful editor than vice versa - which is why google docs pivoted towards smart canvas features instead of launching an all new app.

For me, as a developer, Microsoft is a company with deep roots in compilers and operating systems. All of the business empire started from these two foundations. Since its inception till today, Microsoft has been producing some global products every decade. And then making these products unbeatable in the global marketes. Be it Windows, Office, Exchange, Developer tools & Compilers, Web Servers, XBox, Azure, Teams etc. You just name it and Microsoft is right there in almost every field with profitable products. A lot of people don't like the traditional Microsoft products but all of the products just work and are being used for tens of millions of customers around the globe every day. Any Software company with ability to create profitable products every decade is a killer company. That is the secret souce Microsoft has.

As a developer just imagine about a company which gives you a developer tool to make a web application, using the company's provided compiler, which then can be deployed on a company's provided webserver and can store some data on company's provided database which is running on company's provided operating system.

The company also happens to provide end-to-end tools for running a company of a 5 people to a company of 500,000 people.

This is Microsoft.

Then there's other ethical developers. This is a company with a deeply scummy past, who did everything to throttle competition when they had the upper hand.

They might be trying to change their image lately, still a hotbed of scumbags who learned from the best.

>This is a company with a deeply scummy past, who did everything to throttle competition when they had the upper hand.

This applies to literally any billion/trillion dollar big tech company ever, Intel, Nvidia, AMD, Dell, Facebook even Apple and Google. They all abuse their market dominance at the expense of their competitors when they get there. It's literally the M.O. of any major corporation.

Microsoft was just the first major successful big tech software company to make it there.

Maybe companies shouldn't be that big then?

Welcome to Europe.

Don't tempt me

If you come to Europe can we please exchange jobs and wages? ;)

Considering I work in academia it won't be much of a change, but the work is meaningful :)

Haha, nice. Though, a work colleague's brother was doing laser research in academia in Europe and got offered a 4x pay increase to continue his research in the US academia. Make of that what you will, but it seems like the US stil has no issues attracting top talent in some fields.

VW, BP, Deutsche Bank, Airbus and especially LVMH and Nestle, would like to have a word with you.

You're making it sound like the lack of major European SW megacorps comes from some benevolent voluntary decision on our side, when the truth is that Europe is full of unscrupulous, corrupt and exploitative megacorps like the ones I listed above, except in SW, since we missed the SW bus entirely due to reasons I will go into detail below, and so the US dominates that field entirely.

Firstly, don't have the capacity to pump trillions of EUR into our stock market like the FED does and we also don't have the military capacity to invade any country that would threaten the EURO or our oil/energy markets.

But most importantly, the US is a single market with 300 million consumers with a high purchasing power, making scaling of SW products and services much easier than in a fragmented market with about 24 different languages and many more different cultures and differing regulations, with some EU member countries having conflicting interests on many important topics, making scaling of SW products across EU members a nightmare if you're not flush with cash. Which is why we're full of thousands of small local companies that don't have any international leverage.

And before anyone brings up ASML for the thousandth time as some silver bullet example for EU tech dominance, please note that ASML's golden goose, EUV, is a product of US Cymer wich ASML bough, and of Sandia labs research which ASML licensed, so the US has veto rights on what ASML can do with the EUV tech and who they can sell it to (spoiler alert, not to China).

> VW, Deutsche Bank, Airbus and especially LVMH, would like to have a word with you.

Sure, but we're talking about tech firms (this is an article about Microsoft's place in the tech world). All the FAANGs are American. Big tech is American.

edit: I see you edited your comment (many times) to align with this part of the discussion.

>But most importantly, the US is a single market with 300 million consumers with a high purchasing power, making scaling of SW products and services much easier than in a fragmented market with about 24 different languages and many more different cultures and differing regulations, with some EU member countries having conflicting interests on many important topics, making scaling of SW products across EU members a nightmare if you're not flush with cash. Which is why we're full of thousands of small local companies that don't have any international leverage.

And what stopped Japan from achieving what Europe couldn't?

>And what stopped Japan from achieving what Europe couldn't?

Why would you assume Japan would be different? They have most of the issues Europe has in the SW scene.

Japan has (or at least had) several advantages that Europe still lacks. Single language, high and concentrated population, highly technological consumer markets, massive government and corporate investment in semiconductor and software.

For a while, Japan's was beating the US in semiconductor advancements. If the Japanese software scene failed, I'm not convinced of how it would have happened the way same Europe's had.

Germany also has one language and a high and concentrated population. As do Britain and France.

Germany alone is a big and rich enough market to scale. Add the UK and France and you're very close to the US population.

SAP?

Well if everyone else is doing it, I guess that makes it OK...

How did you come up with that conclusion from my statement?

I explained why I don't trust any major corporation richer than God, as they're all guilty of bad practices and why you shouldn't trust any of them either.

Is there anything Microsoft could do to gain your forgiveness or do you just plan on hating them forever?

They wouldn't even if they could, so the question is meaningless.

They would probably be willing to parade Steve Balmer through downtown Redmond and yell "shame" at him every few steps. I think I could totally start overlooking IE6 if they did that.

in their most powerful days, they didn't dictate what software could run on their OS and tax developers 30% of their revenue to allow their software to run.

> A lot of people don't like the traditional Microsoft products but all of the products just work and are being used for tens of millions of customers around the globe every day.

I have never spent a dime buying a MS product (directly) but they definitely don't "just work" based on my experience with my work computer. Windows is awfully slow, Office and Teams are sometimes just unusable because they take so much memory on my machine and keep freezing randomly. I do agree that MS used to be the company where products just worked (windows 95, 98, office 2003 and prior, hotmail etc.), but these days pretty much all MS products are awful.

People say the same thing about apple products, but both Microsoft and Apple products never 'just work' for 90% of what I'm trying to do, and if they do the process is convoluted and unintuitive from my experience

Apple and Microsoft products aren’t even on the same dimension when it comes to poor design decisions, cruft, stability and performance. Microsoft fell behind a while ago.

And apple seems to be catching up

Honestly this is my experience as well, they “just work” till they just don’t work. Than getting either the Apple or Microsoft product to work is usually hours of time wasted.

I guess you could say the same about open source or systems like Linux where they also have their issues, but fixing those issues always seems to be a fraction of the time.

If your PC is slow, and and running out of memory for apps. It may be time for an upgrade.

Your Core2Duo machine is a little old now there buddy.

You have windows 95 as a OS that just "worked"

Windows 7 was the first Windows OS that "Just Worked"

One thing that may be worth considering is the quality of corporate laptops that are handed out versus, say, a top of the line custom-built PC.

Outside of any gaming rig I've built I've never used a good Windows computer for work. Not a "top of the line" laptop, not a virtual desktop client, nothing. Every single one has absolutely sucked.

They give non-developers at my work Surface devices, and those people have nothing but issues.

Core2Duo might be old but apparently MS also considers skylake i5s "old" as well.

It might be time to downgrade office instead.

tbf, skylake is 14nm and 6 yrs old now. It is objectively old, though probably good enough for another 5+yrs.

you mean windows 7 sp2, right?

> This is Microsoft.

This is not freedom. As a non-technical client that just wants to make a spreadsheet and share it with the rest of the corporation, that perhaps doesn't have to matter (as long as you can afford the ecosystem). But from a technical perspective, you give away the ability to learn from and develop your own software, to compete and innovate, to combine different parts and come up with something better.

As a developer, I don't understand why a developer would side with a company like Microsoft, unless your product ties deeply into their ecosystem and you are very optimistic about your relationship and the future.

If you want freedom, you are free to use Linux.

Good luck getting wifi and graphic card to work.

> No singular power law product defines Microsoft like Google’s Search, Apple’s iPhone, Amazon’s e-commerce, or Facebook’s social network.

Umm windows and office?

Great article, I still think that with thousands of amazing engineers in microsoft, they need to simplify 2 things. Focus on using 1 thing, then reiterate to make it better, i.e: why is there teams and skype? then there's teams for business, teams for personal, skype for business. Also teams app is built on angular instead of xamarin? I heard that teams of engineers within MS are free to choose any tech that they want. If they had chosen xamarin since the beginning (now MAUI), wouldn't it made xamarin much less buggy? It seems like every teams are going on different directions with different managements. you don't see that on apple (pushing swift everywhere), facebook is also using react for almost all their internal apps.

Second thing is it's so hard to make your voice heard in MS that most people just gave up lol. I have this bluetooth issue on windows 11 where my bluetooth speaker produce no sound after receiving call from teams. I have to reconnect every couple of hours. I've had it for 4 months, and I don't know where to give feedback at all (feedback hub is useless) other than asking strangers on reddit. If you have 96k engineers around the world, can't you assign at least 5-6 PM that focuses on consumer satisfaction, engineering excellence, or at least have a bug bash once in a while.

Again i'm not sure what's happening inside the company, maybe i'm to judgy, but at this point a lot of people are feeling the same way

> Umm windows and office?

So it's two, not one.

I share the sentiment of your reply though, e.g. I have a hard time associating Apple with the iPhone only also hard, would actually associated them with iOS products.

For a company that had a dude yelling Developers on stage - how in the world did they blow their dev stack so badly?

XAML, WPF, UWP, silverlight etc etc.

They owned with WinForms back in the day. There was nothing close to market share / productivity for LOB apps. Then it was like they just dynamited repeatedly, and kept on dynamiting?

I can't even imagine the wasted dev cycles, and now the wasted time using janky juddering online apps (even Vax/VMS green screen LOB apps were actually FASTER -> keyboard driven, no lag). If you tracked medical billing from vax/vms days (a fast typist could crank through billing slips and a tech forward clinical staff could checkin a patient and go with a few keyboard keys (including the good old F keys)). Now its wait wait wait, mouse click, mouse move, click, wait type, submit, wait.

I dunno, a Martian surveying the 2022 development landscape might believe the company that owned GitHub and VSCode was doing fairly well.

When the GitHub purchase was happening, there was a lot of noise from both parties about how the companies would remain separate and GitHub would keep the things which make it appealing to developers; it therefore seems disingenuous to me to consider any of the positives of GitHub to be attributable to MS.

But they don't capture them to develop against their frameworks. I liked WPF, but their APIs were clearly influenced by business decisions like forcing software into their store, especially in UWP.

I am not yet if they are a reliable partner to invest time in. Apple uses such a model, but then why would I use Windows if it just morphs into a worse clone of it? I dislike that model and believe it is extremely bad for digital/technical education.

Those that use open systems with software freedom will shield others from the worst exploits MS or Apple might come up with in the future.

VSCode is indeed nice. Github is ok.

WinForms is still available and has the same productivity.

dotnet has successfully transformed itself and is quite popular.

XAML is ok, I guess.

They have stumbled in the Mobile app dev frameworks, focusing on Xamarin, but seem to be moving to MAUI now.

I like .NET Core a lot and use it for my back-end services, but when it comes to giving things to end users there is just no good story for it at all.

I'd be more than happy to try .NET Maui but I have to keep a VM around to try it out on my main dev machine. I figure being a .NET developer on Linux you are always going to be slightly 2nd class in their eyes, but the experience with docker-ce and VSCode on Linux is streets ahead of Windows and WSL2 with that single glaring exception. It just can't be that hard.

what's wrong with WSL2? I've had a good experience with vscode running on windows with WSL remote connection.

There is nothing really wrong with it, I used it for a while but in the end I felt like Linux with extra steps was too much.

WSL2 allowed an easy migration path to VSCode but once I had that working, I found myself learning all the CLI for dotnet, docker and AWS to replace what I was doing in Visual Studio.

At that point I had nothing to lose and switched to Debian, which had the great side effect of making Android Studio better (the emulators on Linux seem faster and more reliable).

I don't hate WSL2 but if all your tools are supported on Linux, removing Windows from the equation makes your setup much simpler. At least it does for me.

I do work for myself though, so no corporate requirements held me back.

Not really true, I develop .NET Core apps on Linux only. Migrated from Windows to actual Linux because of the simplicity.

Have you tried Avalonia? It's not from Microsoft, but it's a .NET GUI solution that works on Linux and is really pretty good.

I'm fine with XAML, but I don't know what XAML platform is stable and futureproof. Maybe WPF, still.

imo. WPF seems like a good framework that will last a while, while not being a pain to work with like WinForms. I know there are better options nowadays but WPF is a trusted and established technology that will probably not fade a way for quite some time, simply because "it works" and there is no real reason to get rid of it.

I agree but they don’t care because they dominate the market.

Yes. They have the power to push whatever they want on captive audience (engineers forced to work with MS stack). That means they don't need to care whether what they're pushing is actually good.

So it mostly isn't.

They lost the mobile war and web tech creeped in. Their tech stack is quite good but they stuck to Windows only and that turned from an advantage to a disadvantage. They're turning it around at least.

> XAML, WPF, UWP, silverlight

Aka bunch of mediocre walled garden tech. I'd much prefer QT

You seem to suggest that it would be a good thing for everyone for those thing to succeed? The only people who benefit from those are MS and their developer base.

I'd argue the users benefit from having something that integrates tightly with the OS rather than an application that targets the lowest common denominator of all platforms. Qt applications always feel a bit shitty to me, I only use them if there is no alternative.

the users benefit by having an app at least. It's unlikely that apps written in the walled garden will be ported to other OSes. If they make those apps cross platform somehow, it would be the lowest common denominator in other OSes.

Windows isn't a walled garden. It's a bit shitty but anybody can write and distribute software for Windows.

I don't know if you mean the same thing but I feel it's KDE that's shitty. When I played around with QT Creator on Windows it seemed great

> walled garden

https://github.com/dotnet/wpf

> how in the world did they blow their dev stack so badly?

Did they?

With TypeScript, Visual Studio Code, npm and GitHub alone they control by far the biggest share of development stack out there in 2022.

I would add on top of this some other libraries that I consider huge sleepers but I think will even further spread and conquer market such as pnpm, playwright and rush.js.

There's also C#, Visual Studio, DirectX and many other Microsoft technologies devs use every day.

Step by step they will keep integrating many new generation developers into their ecosystem which, if they keep improving on their azure, office and teams integration will considerably drive up their revenue.

As long as Microsoft keeps working hard in this direction and doesn't start alienating the customers I can see them cutting more and more shares of the dev and release markets.

Microsoft's biggest failure was Windows Mobile. If they got Mobile right hands down they would be on a highway to $10T company but otherwise they are riding the wave of cloud, AI and gaming. Like Steve Jobs[0] said they are strong opportunists, they have no taste, they don't have original ideas and they don't bring much culture into their product but they keep on coming. Microsoft uses its big cash pile to either acquire companies or to copy them.

[0] https://www.youtube.com/watch?v=KSg3fU9XWow

Apple was quite happy to get some of that cash in the late 90's.

I feel if Windows Mobile had just kept at it for ten years straight they would finally have had something - continually changing strategy killer any hope they had of being in a solid third place.

That compatibility break between Windows Phone 7 and Windows Phone 8 was brutal to their momentum. I had multiple friends with Windows Phone 7 devices, and only one with an 8.

I may be confused about this, but I think there was also a compatibility break from WP6.5 to WP7, and from WP8 to WP10.

It seems that eventually, mobile app developers had enough of this farce and abandoned Windows Phone.

It's funny that you can still run 1980s visicalc.exe on Windows 10, and CMD.exe supports 1980s DOS command syntax, but for mobile apps they decided that apps should be rewritten from scratch every year.

Yes, there was a break from early Windows Mobile (pre WinPhone7) because that UI system was really more or less a classic Windows GUI squeezed into a small screen and it sucked (unless you had a pen, but it was still a tad cumbersome).

I bought a cheap WP7 Nokia to develop on and the thing i remember most was that the UI was as smooth as Apple and not a stuttery one like Android of the same age.

Don't remember exactly why they left WP7 owners in the dust, but the biggest issue with it was that they fell for the Osbourne Effect and left quite a few mobile merchants pissed off about the amounts of WP7 devices left in stock that nobody wanted (and since those mobile merchants were often not the same as those with MS brand trust from the computer industry they felt no reason to give them a second chance).

While customers might've been a tad cautious, the biggest reason IMHO why WP8 died because few carriers,etc wanted to stock or sell them to begin with.

The number one rule of software and hardware should be never give your customers a reason to re-evaluate their original decision - and a breaking upgrade does just that. I had a friend who loved his windows phone 7 - but ended upgrading to an iPhone because of that.

There was also a break between 8 and 8.1, yep.

Ah interesting. So compatibility issues from 7 to 8, and 8 to 8.1. But not 8.1 to 10?

I would really like to learn why these breaks were considered necessary at the time. In a practical sense it's irrelevant since WP is dead. But on the other hand, with the huge mobile market at stake, why did MS make these decisions that doomed it, and that seem obviously wrong to outsiders?

Also between 8.1 and 10.

To avoid repeating myself,

https://news.ycombinator.com/item?id=27412065

The answer to that is Windows proper either ran on top of DOS or ran a DOS-compatible layer (Wow32). Windows Mobile 1.0-6.x and Windows Phone 7.x were based on the Windows CE (an embedded OS from the 90s) where as Windows Phone 8.x and Windows 10 Mobile were NT based with modifications made for ARM and battery life. Different kernels are the reason for the breakage.

To be fair, they were at it for about 6 years before giving up (2009 to 2015); which is a really really long time to spend trying to get momentum going.

I suspect the real problem was it was hurting other parts of the business. They needed to abandon it to free themselves to accept the reality of the dominance of the iPhone.

Dominance of iPhone? Do we live in the same planet?

Android won the masses.

>Android won the masses.

Nobody cares what poor people buy.

What planet are you on...that's ALL corporations care about.

There is a reason everybody targets the iPhone before Android. That is iPhone users spend more than Android users.

https://www.dignited.com/48795/why-do-ios-apps-generate-more...

I loved my Lumia 920. By far the fastest and cleanest mobile OS I had used.

App support was lackluster though. It always felt like companies and developers had absolutely no intention to reimplement their applications on Windows Phone, even when it had a bigger market share than iPhone, especially in mid/low-income countries.

So I was stuck in 2015 with no authenticators, no mobile banking, no public transport apps. Even some of the biggest apps out there like Facebook where not releasing all of their products, e.g. there was no Snapchat and many other big names.

Windows Phone ultimately was killed by lack of third party applications and Microsoft imho should've kept burning money for 3/4 years more into the ecosystem bringing more applications on their mobile platform.

Same here. That Lumia 920 Windows phone was a breath of fresh air compared to the mess of laggy icons and screens you have on Android and iPhone. Shockingly for Microsoft, the OS didn't get in your way. Snappy and very functional like a phone should be. I have Google's flagship pixels on Google Fi. Bloated slow shit.

We have Android apps running on Windows now, and devices like the Duo which are not quite phone-like but pretty close and we know they've been experimenting with running Windows 10 X on those, so it wouldn't be far fetched to go back down into mobile space with a Windows 10 device running Android apps in an emulator (alongside the limited native app catalog).

I maintain, the Windows 8.1/10 mobile OS was superior to iOS and Android. At least in terms of UX. Microsoft should have tripled down on app support and brute forced their way into the market, as they're doing now with Game Pass.

MSFT tried to brute force their way into the mobile market. App developers wouldn't take their money.

> It is not a beloved consumer brand like Apple, Facebook, Amazon, or Google.

the reason for this is very simple: very few people choose to use Microsoft products willingly (except maybe the xbox)

their products either come with the computer by default or it's installed on your work machine

and the software is at the very best mediocre and somehow getting worse (trying to figure out how to save a Word document locally these days is NOT easy)

[1]: even for the original xbox Microsoft initially deliberately kept their name off of it

>and the software is at the very best mediocre and somehow getting worse (trying to figure out how to save a Word document locally these days is NOT easy)

This is a great example of hostile UX. It appears designed to encourage people to use their cloud services, since they offer OneDrive as the first option in the list, and something like 90% of users choose the first option in a list.

IMHO, what Google and Microsoft are sorely lacking is an obsession with UX. It is what drives Apple to maintain the hardware integration, and it works. Google and Microsoft are making inroads here but it just feels very half-hearted by comparison.

> very few people choose to use Microsoft products willingly

My gut feeling is that, for PERSONAL use, the US has shifted to favor Apple products, over something running Windows. Even in my backwater neck of the woods, I'd say that Apple vs Microsoft is running at least 2 to 1. For 25 years, Gartner ran the same report saying that Microsoft "owned" the computing device landscape. Now their schtick is that Android rules the world. Fine. But what I want to see is the operating system market share data, by country, with the corporate purchases factored OUT. It doesn't seem that hard, but, naturally, that data is NOWHERE to be found online. If Gartner is actually using real numbers, then this would be "internal polling" data which I'm sure they consider their trade secret. If it were possible to get the data for the US, for personal use only, I think it would show that Apple is doing even better than most trade analysts begrudgingly give them credit for. With the demise of most Microsoft stores, and the way Microsoft is cannibalizing trust with the direct commercialization of Windows on factory-installed machines, I don't see how Windows on personal devices can be the "thing" it has been traditionally considered.

At the large retailer I work for, AWS is off limits. Strategically, they don't want to base their IT operations in a cloud of a company that is also a competitor in online retail. I can understand that. Here Azure has a great advantage.

Couldn't agree more. If regulators ever attack the Amazon stack and split up the company, AWS could become one of the most valuable companies in the world.

> If regulators ever attack the Amazon stack and split up the company

they'd have to have justification for doing so - which i don't see any.

Profits from AWS props up Amazon losses on retail.

Amazon started breaking out operating income separately in 2013, and the rest of the business has made a profit every year since then. The US retail operation now brings in as much profit as AWS, though some of that is offset by losses in their rest-of-the-world retail business.

This is very common for my clients. Another common variant are B2B focused clients where their customers contractually require them not to put data onto AWS. IE, Walmart doesn't want any of their data to end up on AWS, and if you want to do business with Walmart you need to similarly protect their data.

> Here Azure has a great advantage.

Why not GCP?

Yes it would work too. We are very Azure-heavy but lately there is a move to multiple cloud vendors, basically GCP in addition.

Does anyone else find it boring to kibitz on how an absurdly powerful company could become more powerful?

Wouldn't it be more interesting for Microsoft to actually improve its products, e.g. stop Excel making a nearly irrevocable assumption that something is a date just because it has two numbers and a hyphen between them? Make a successor to MS Access that doesn't suck? Fix the shitshow that is figures, tables, and cross references in Word? Make a nice new product inspired by FOSS projects that are way ahead of them, or failing that, at least some decent craftsmanship on existing products?

At some point, doesn't the endless lust for monopoly power become boring? Why not just actually do a good job at the thing your company is supposed to be about?

How many yachts do Satya and the MS board of directors need? Why not do something beautiful instead? Who needs to spend their life replaying the modern equivalent of an ambitious feudal lord?

Dont' forget *its* Microsoft.

Github never asked for login until Microsoft acquired it.

Broken windows help page. Try installing drivers and navigate to knowledge base.

When some thing does not work on windows you reboot and hope it works or just reinstall operating system.

Try tweaking all the settings in visual studio only to be reset on the next update

You have no control over updates. It happens especially when you want to give presentation. Also you can not poweroff whenever you like, because it chooses when the update should happen. No amount of changing setting can fix this.

These are few problems.

Don't forget *its* Microsoft. It won't change.

Can Microsoft please make a Windows phone after all? Both Apple and Android ecosystems and hardware together are terrible in various ways and don't work well with Windows.

They did. It was called Windows Phone... It came out too late, offered too little.

It was pretty good. Seemed better than Android at the time. I'm not sure why Microsoft didn't put as much oomph into it as they did for, say, the Xbox, but if they had it might be a major player now.

I had a HTC windows phone and it was not good in my experience. The app store was non-existent, the OS forced you into using its own social media experiences which were half implemented and janky vs. native apps, and the web browser was slow and clunky for the era (2010s). The home screen was enormously painful with all kinds of different sized icons strewn about and blinking, changing, etc. randomly--I could never find anything in the same place.

The only good thing I could say is that the settings were much simpler and easier to navigate than apple or android phones at the time. But that was really more because there was far less you could change or do in windows phones.

As far as effort put into it, Microsoft spent years pouring money and resources into trying to grow their phone platform. Remember they basically invented modern smartphones with Windows mobile phones--I was using 'real' web browsers on my windows mobile 5 phone as far back as 2006, years and years before the iPhone came out. They spent billions buying companies like Nokia, trying to get more developers on their platform, etc. There were at least two or three complete restarts to the whole platform. There were internal competitors even from the Xbox/game division (remember the Kin?). It just churned and churned and churned despite all resources and money thrown at it--all of it wasted in the end.

Oh, thanks for the context! I never owned a Windows phone, I just played with ones other people had, same for Android. And from a total beginner perspective, Windows phone seemed a lot easier and more intuitive, but I can believe that with more use its annoyances would become more clear.

That's interesting about Microsoft really trying for mobile. My recollection is that they didn't put much effort into it, but clearly I was just unaware of it. Which is, perhaps, a sign? For all its investment, perhaps it didn't have the right marketing, as I was never aware of Microsoft's efforts for Windows Phone the same way that I am for Android and iPhone.

There was a point where I saw a decent amount of buzz, but I also live in an area that had one of the bigger "Microsoft Stores." That said there was a point Microsoft more or less lost the plot on Windows phone, it was a few things that added up to them giving up IMO:

- Being unable to negotiate with companies like Snap to bring apps to the platform. Often this was because of Microsoft's level of sandboxing which prevented certain types of low level access.

- Other app developers never bothering to even try, despite the perfectly servicable APIs provided and ability to write in multiple languages.

- Google was purposefully obstructive with Youtube's APIs IIRC, which eventually made the Youtube app useless.

I bring up that last point, because it seems like the closer to competitive WP -got-, the more google turned the screws. Thinking very specifically about WP10 here, which really was a Swan Song for the platform; using a Widi receiver to get an ad-hoc windows desktop was pretty freaking cool in 2016.

as far back as 2006, years and years before the iPhone came out

iPhone came out in 2007, at least in the US.

Sure, but when it came out it was pretty barebones : no 3rd party apps, no 3G...

It also couldn't connect to AD or be managed by System Center.

Windows phone were pretty good. Always been happy with them, of not for the lack of developers interest.

How are they terrible?

Having used cellphones since the 90s, my experiences with Android/Apple phones have been nothing short of remarkable and to be frank, life changing in terms of day-to-day utility.

Android hardware is weak. UI is good, app ecosystem is weaker than Apple but more free at least.

Apple hardware is top notch, obviously. But I wish the software side was not so strictly controlled, look at the browser picture for example or in-app platforms or the shenanigans with the 30% cut.

Then small nits like:

- the alarm doesn't tell me how many hours to go when I set it

- the orange silent button is an abomination that needs to go. With Apple's focus on not having any buttons I have no idea why it has lasted this long

- the volume settings make no sense in the way that Android's make sense

So basically I'd like some wizard to take both of those phone systems, throw them in a cauldron, stir them together, throw away the slag, and sprinkle in some extra goodness. Ahhh some day maybe ...

> Android hardware is weak

Is it ? I'm using a 4 years old $200 phone which suits me just fine, if not for stupid apps requiring more resources for a worst experience. The only reason of upgrade is battery wearing off.

Any Google related issues might be mitigated switching to LineageOS as the hardware is relatively open.

There's a certain level of truth to it, as none of the Android phone makers appear to be able to push for a customized SoC, even the ones with own designs (Samsung) appear to keep a "wall" between mobile and SoC divisions.

This means that relatively easy wins like throwing a lot more cache on high end phone, are unavailable because instead they are just one of the many customers for the SoC with different priorities resulting in a compromise design.

I've stuck to the Google product lines and the hardware has been awesome. Galaxy Nexus, Nexus 5, and Pixel 4a. Each has improved on the previous as well.

My pixel 3 shows "Sun, Jan..." Instead of the much more useful and same length "Sun, Jan 30". There's a bunch of other minor annoyances like that. But pretty egregious that their own phone can't display the date anymore

I'm surprised you say the Apple ecosystem is terrible. It's arguably the best, most complete and integrated ecosystem out there currently.

Of course it's not going to work well with Windows. Apple wants you to buy a Mac.

It's hard to call the iPhone complete when it makes egregious concessions in it's capabilities in the name of "safety" (which turns out to be a moot-point anyways, as sufficently funded state actors have proven to us over the past few months).

Security is not moot just because state actors can break it, that's basically expected. The idea is to give everyone below the level of a state a harder time.

If all that effort didn't come at the expense of user agency, I'd be inclined to agree. Instead, it looks like an increasingly toothless justification to prevent people from undermining the App Store's profitability. The security theater isn't as convincing when their puppets are falling apart in their hands.

But it's not security theater. Sandboxing and strong isolation is one of the oldest techniques around for increasing security. iOS has the least malware of any OS. The security measures it contains would appear to be extremely good at keeping out threats that plague many other platforms.

Sandboxing and strong isolation are not mutually exclusive with allowing third-party sources to install apps. If their sandboxing is as good as you say it is, there should be no security threat imposed by allowing other apps on the device. The two should be distinct, separate threat models, but Apple conflates them.

> If their sandboxing is as good as you say it is, there should be no security threat imposed by allowing other apps on the device.

No, because it's part of a defense-in-depth strategy. Sandboxing and isolation is meant to catch what app review misses, without app review people would be attacking those mechanisms directly at a much higher frequency.

Allowlisting is also one of the best security practices you can implement, because allowlisting significantly reduces the problem space of executables you have to consider (and dangerous executables, if found, can be removed from the allowlist). There's always a chance that a piece of malware which somehow breaks the sandboxing and isolation can find its way onto someone's device; allowlisting greatly mitigates this possibility.

Endpoint security is everyone's problem now, because everyone has one or more always-connected devices. The future of computing is a signed, remotely attested path from power-on to user application code, all checked against an allowlist of approved binaries. For most, this will be a good thing.

What state actors? They are compromised by a small, private company.

After 7 years of android I switched over to iPhone. Largely due to the Google hate I experience here on hn. There is so many small annoyances on iPhone. Some things like hide my email, password management are brilliant, even if I need to use Safari, which I hate because I want to use Firefox for ad blocking. But iOS is objectively terrible. Notifications are so horrible. Support for my Garmin watch is awful. The jarring system sounds and don’t even get me started on the awful alarm sounds. Siri is unusable and the fact that I can’t change to another assistant is terrible. Apple Maps doesn’t even have biking directions in one of the most bike central cities of the world, amsterdam, lol!! Useless. Dutch is not supported in system wide translations feature, again I need to revert to Google translate. Keyboards for multilingual people are really really terrible. I need to switch between 4 languages every day. The fact that I need to switch between keyboards and then still have autocomplete fight me whenever I dare to use a word form a different language ( even simple ones like “yeah” which I want to use in German or Dutch as well ). The keyboard is so terrible as well. And then just things that make no sense at all. Why can I not see a battery percentage on the icon of the battery? I need to swipe control center down to see that. Why can I not see seconds in the time? Why can I not set two timers?????

I really really really hate iOS. It has things that are nice and better than android. But for anybody to think that it is a “complete and integrated” ecosystem makes me … laugh. Like it’s just objectively terrible.

> Why can I not see a battery percentage on the icon of the battery?

Settings :: Battery :: Battery Percentage

Siri, Keyboard, and Maps gripes granted, but not really what people mean when they talk about "ecosystem", I'm very happy with how I can copy paste between my phone and laptop, and how my homepods and airpods and appletv all hand-off to each other.

> Settings :: Battery :: Battery Percentage

Yeah, that would make sense, right? But Apple decided not to include that setting on iPhones with a notch (which is basically all of them these days).

> I'm very happy with how I can copy paste between my phone and laptop.

If you use Linux and Android, you can do that with KDE connect. Pretty sure there's some solution for Windows too.

A lot of the annoyances that you cited are valid. I speak three languages (one that the keyboard does not support), and my accent when speaking English trips Siri a lot (But I prefer not to switch to French). But the ecosystem really works together when using Apple products. Anything else and it becomes jarring. And you can appreciate it when you accept THEIR way of things. I used to be a heavy customizer (Arch, Custom roms for Android,...), but this day not so much.

And it does work for people that can accept that. Anyone else? Not really, as you will feel frustrated by the constraints and the tall wall of their garden.

That's just saying it works together so long as you define working as what it does.

> objectively terrible

I think notifications are far better on iOS, love using Siri for controlling things/the phone (search is mediocre at best), I like the system sounds and alarms, hate when I have to use my Pixel dev phone's keyboard, etc, etc. It's rather subjective.

The two timers thing though, holy crap that annoys me. I swear they said they fixed that...

> I think notifications are far better on iOS

Except you have to swipe-then-tap to dismiss them. It's so fucking annoying. That's the biggest thing I miss about Android, just a simple swipe dismisses them.

Here is some tip if that can help : if you continue to swipe after the button are shown, you'll feel a vibration and releasing your finger will activate the default option (which is "Close"). In fact, you just have to do larger swipes.

But yeah, I also uses iOS because I don't want anything Google. In terms of UX, i don't like iOS but I also loathe Android. I currently lack courage, but I think that in some point in my life, I'll just throw smartphones. I'm sad that Firefox OS became this closed KaiOS thing. A hackable dumbphone is just what I need.

There still is Ubuntu Touch with slowly growing device pool support. Dunno how polished or performant it is right now, though.

And for those of us who don't want to live exclusively in their walled garden, it's an ecosystem rife with friction and sludge.

Right now due to my lack of self control a lack of apps is a bonus. I just need a maps, camera, uber that's about it.

I know I noticed that iPhone hasnt really made my life better. It only has wasted a TON of time. Sometimes I wish I just had a flip phone, but FOMO is strong to give it up. Maybe one day.

The OS is in the middle of an internal refactor which will make a modern mobile edition (with support for all chipsets/apps on launch) more feasible in the future.

In an odd turn of events the hardware has exceeded the software in the case of Surface supporting Android on the Duo.

> The OS is in the middle of an internal refactor which will make a modern mobile edition

Could you elaborate? Is there any info on that?

Personally I wouldn't be surprised if they take another crack at it and succeed at some point. Microsoft is in it to play the long game.

They did. I had one. It was not the worst. They gave up because precious few of you also did.

The one idea I question here is advising Microsoft to acquire Zoom. Microsoft has almost as bad a track record with chat/videocall apps as it does with cell phones. Plus, Zoom has perhaps already seen its best days. It was in the right place in the right time, but the public is fickle. If Apple ever gets around to putting enterprise features into FaceTime, the entire product space will be disrupted with unpredictable results. Better I think for Microsoft to spend a couple of billion at figuring out what went wrong with Skype (free hint: reliability and call quality) than to throw $150b at Zoom and just repeat the same mistakes. BTW Microsoft, why does your Skype website still feature a "Shang-Chi" movie tie-in? Is anybody paying attention to keeping your landing pages up to date?

They already have Teams. Zoom would just languish as their 2020s Skype. Waste of money unless they just want to kill another competitor

This. It is truly amazing how many in this thread have 0 clue about MS products. I guess GSuite is amazing if you are a 2 person company. Everything above that is on the Office 365 train.

Disclosure: I used to work on Google Cloud (but never on Workspaces née GSuite itself).

There are plenty of companies with tens of thousands of employees using GSuite [1]. Also various governments, agencies, regulated industries and so on.

A few years ago, GSuite added improvements to Office importing (primarily Excel and PPT, IIRC) that assuaged a lot of folks. Companies still might have departments using Office (e.g., Legal still using Word for its redlining or Finance using Excel) but because you can shove them all in Drive or import them into the GSuite equivalent, most of the company happily uses GSuite for everything else.

[1] https://workspace.google.com/customers/

I've worked in 3 companies with the classic Silicon Valley starter pack consisting of Slack, GitHub (not part of Office 365), and GSuite, and found it overall pretty good! What features did you find missing? Personally I shudder at the thought of using SharePoint, it's the worst designed product I've seen.

Never seen a silicon valley tech company on MS office products. Airbnb, Coinbase, Dropbox, Lyft, Uber, Facebook etc use Gsuite, slack, zoom.

MS teams only comes up when you talk to lawyers, accountants etc more legacy or enterprise companies.

I’d think there is something wrong if I see a tech company on MS office. Either they are clueless or they want to save costs.

Facebook has an O365 license; managers definitely use Outlook e-mail & calendar to its full extent. ICs rarely check their e-mails.

> Zoom would just languish as their 2020s Skype.

I wonder what was so wrong with Skype that many abandoned it? Now, i've been told that the enterprise variety is a different product that's way worse, but i use the regular Skype occasionally and it's okay.

Personally, i think that it was best before its many redesigns (e.g. the versions that you could get from sites like http://www.oldversion.com/windows/skype/ at least when they worked), but even nowadays it remains usable and does most of what i'd like for the basic use cases from a chat application, or even for video calls or group calls.

Then again, in my eyes many of these platforms are just reinventions of IRC in some capacity, with the occasional nice feature (e.g. Slack/Discord/... having threads, deep API integration with bots/apps etc.) that gets tacked on.

Skype for Business was "Lync" rebranded, not Skype right? It was so terrible I think they wrecked Skypes reputation.

Video is actually the only part of teams that doesn't suck

As someone who hasn't used Teams much—what sucks about it?

High sound latency. Can't copy paste code properly. Code formatting doesn't.

Electron app for chat... Try using narrator with teams.

"It is not a beloved consumer brand like Apple, Facebook, Amazon, or Google."

They're not a beloved consumer brand, yet.

Microsoft has made an enormous investment in consumer and developer sentiment.

VS Code, Microsoft Flight Simulator, WSL, and Xbox are huge, long term plays that are a rocket ship in their brand's sentiment.

Office has a good reputation with most white collar professionals. Windows is still ubiquitous outside of programmers

Going by the StackOverflow yearly survey, Windows is still the most used among professional developers by a good margin. But yeah maybe 40% doesn't equate to ubiquitous.

https://insights.stackoverflow.com/survey/2021#most-popular-...

It's super easy to convert an average user to libre office tho.

Just wait for the next major office update and tell them that there is a piece of software that has a UI that still reflects what these people are used to.

I agree that Microsoft has put tons of resources into developer evangelism. Wsl, vscode, and terminal have been a major turn around for them. But I’m not sure they’re trying to be anything more than “the only option that was presented to me” with anyone else other than executives that sign the office365 contract.

I'm probably in the minority here, but I'm a software developer and my sentiment regarding Microsoft is negative and growing more so as the years go by. VS Code, Flight Simulator, WSL, and Xbox are fine, but they have destroyed goodwill with me through mismanaging Windows.

Windows is now sprinkled with dark patterns like not allowing local accounts w/ windows 10 unless you turn off your internet during install. Windows 11 now requires a MS account. The number of advertisements in the OS continues to grow. The enormous and unforgivable amount of e-waste that they will soon create with the Windows 11 hardware requirements / the end of life date for 10. The sense that everything I do on their OS / software is being monitored through telemetry. The UI regressions w/ 11. Hard to quantify this but the I have encountered a lot of bugs with Windows in recent years and some of these bugs are infuriating.

I could go on, but I'll leave it there. Like I said, I'm sure I'm in the minority here, but I suspect that the number of dissatisfied users is growing.

Yes, it is getting more and more on the horrible side. But something’s gotta give at some point. I stopped using Windows for my personal computing but continue to use it for work because I have no other choice. Give Linux a try if haven’t yet

My counterpoint:

More than half the population doesn't use ad block when they browse the internet.

The mind boggles.

There is, clearly, an enormous gap between the Hackernews perception of Windows, and actual non-ad block using end users.

Microsoft is intrinsically following a pathway I've seen before with Reddit,

Reddit has two completely different value propositions to power users and the great unwashed.

Reddit has two modes, "new" reddit which is available by default, and the "official" Reddit mobile app. And "old" Reddit, which is buried in the settings menu, powered by a third party browser extension on desktops, alongside third party paid for mobile apps that use a Reddit API.

Microsoft to me is following the exact same strategy. They are screwballing "the internet explorer users" with Windows, and meanwhile they are creating a parallel, WSL enabled value proposition for power users.

It'd be nice if they went all the way with the second one. I need to manually wireshark and add to my blocklist after every major update.

Just want a Windows 10/11 "Power User" SKU with completely disabled telemetry, "suggested content" and any "home phoning" not otherwise authorized (I still want Windows Updates, MS Account / OneDrive / Xbox Game Pass). Heck, I'd even pay a subscription for that.

Oh nobody is ever disabling telemetry by default, system wide.

Telemetry is too damn useful for developers. Nobody gives a damn.

Is it really?

In general I feel spying is used to cut longtail feutures (most users are dependent on one of them) and justify changes without reason other than flawed interpretation of statistics.

Moreover, their software quality has gone downhill ever since their pervasive telemetry. I am not claiming that it is the cause of the decline, but it did not prevent it. It's curious that the version of windows which most people liked best was 7, the one without all the telemetry.

I think it's also hard to overcome the feeling that everything you do on their software is being monitored and evaluated. Even if we accept the argument that telemetry improves software through allowing them to focus on more used features, the improvements need to offset the feeling that you're being spied on, and that's a tough ask.

I agree. Telemetry is mostly used to figure out things which you don't need to do because nobody uses them in my experience.

Still, it's a tool of big corporate to provide transparency to VPs, and VPs would love to cut costs in Windows since it's a captive market.

You are not alone at least. The only reason I've used windows in the last 10 years was to compile some release codes.

I have zero joy using any of their products. I even stopped using GitHub when they brought it.

I don't get the hype for VScode either when we already have Atom with a bigger eco system and sublime with a better editor. Except several windows specific things it does nothing better or especially great.

I couldn't care less about WSL it doesn't feel like a real Linux, which is all I want. I don't game so meh Xbox, meh flight simulator.

Everything that's left is horrible either way. Comparingly at least.

The main reason I hear people using Windows is because they want to run Photoshop but don't like the apple ecosystem.

Agree. It boggles my mind how bad and downright hostile windows has become in windows 10/11. I'm supporting a group of ~50 people and most of the time I cannot even give reasons for some of the baffling design decisions and/or frustrating behaviors that I see on a daily basis. I feel like every solution I'm giving is a work-around at best.

And don't forget the new mandated corporate messenger: Teams. After burning messenger, killing skype and linq and all the "for teams" versions, this is their absolute worst yet.

The sad part for me is how you _do_ actually get used to this crap and workarounds. I was able to do a linux-only stint for a few months, which served me as a reset ground. Coming back to windows support was a real shocker at how bad doing _anything_ feels like.

Don't forget GitHub and LinkedIn !

I was on the board of a company with a product offering on Azure (because one big customer used AZ for part of their stack). Expanding outside was tough: few prospects were interested as few used Azure (and we had a lot of uptime issues with Azure). This was in 2020 and 2021. The devs complained but porting to AWS saved the product.

MS has a nice lock in with the big companies’ IT departments, which gave them nice recurring revenue but pointed their attention away from where the puck was going several times (most notably missing the Internet and missing phones and BYOD).

They are huge, but in my technical life I was never really exposed to them (except Excel and some Powerpoint). That’s why this board experience was such a surprise to me.

Did the company end up dropping their big customer or did the company use both Azure and AWS?

I work with customers who insist on Azure as long as they know it's an option, however they also have other vendors who are only on AWS and they seem to be okay with it.

Turns out customer had switched to Azure after Microsoft made an investment.

This reminds me of the 2000 dot bomb: hardware companies were paying their customers to “buy” the product. That didn’t end well.

What the article misses is the essential core: Microsoft is the Big Evil. Only Google really competes in that space. The others might like to compete, but Apple is too stand-offish, Amazon too random, Facebook too inept, Russia too old.

Microsoft and Google are always on the lookout for the next bigger Evil. Google eagerly sheds anything that turns out not to be It. Microsoft is slower at that, not giving up as easily, and often late to the party, but sometimes things finally work out for them better than could have been expected (Xbox). The goal is literal World Domination, and odds are one or the other will achieve it. Younger generations won't even remember when they didn't already have it, or be able to imagine a world where they don't.

Amazon just little ago killed some employees to keep them working during a storm, makes people pee in a bottle, Facebook ignored and hid the burden it puts on teenagers development, and you say they're not as bad? I would say the only not as bad, despite hating it, is Apple

I am saying Amazon is not in a position to try to rule the world, and is not evidently seeking that.

Is Apple actually less bad than Amazon, or do you just know less about what happens at and on behalf of Apple?

Thinking about it, probably you're right, I forgot about the chinese workers who suicided in chinese factories partners

https://www.theguardian.com/technology/2017/jun/18/foxconn-l...

At rates lower than the general population.

as long as they keep track of it on the KPI.. .-.

Why not ? It's not like Amazon is only an online library these days...

> Facebook ignored and hid the burden it puts on teenagers development

Here, read the leaked document. Does it say what you said? https://about.fb.com/wp-content/uploads/2021/09/Instagram-Te...

You forgot about Oracle.

I try to. Likewise ransomware gangs, and Goldman Sachs.

And Apple is not evil? Charging 30% on top of a subscription w/ no value add isn't evil? I think according to your definition every company would be evil.

30% is the standard cut for platforms though, going as far back as the Minitel. And yeah, any platform is evil by definition, but Big evil takes more than that according to the OP.

There is no "standard". There is only what the market will bear. 30% seems to be an amount just barely low enough to avoid the attention of regulators in an environment just now absurdly forgiving of monopolists.

But, yes, Big Evil takes more than a walled garden.

Not in the technical meaning of the word, but for some reason, (app delivery) platforms (and software retailers) do seem to gravitate to a 30% cut and/or to treat it as some "baseline" ?

https://www.ign.com/articles/2019/10/07/report-steams-30-cut...

That is generic corporate greed.

Certainly Apple is deeply evil, and I would never, ever buy an Apple product, but the competition for Big Evil is stiff. Apple remains wholly avoidable. Besides Google and Microsoft, we have Cargill and Archer Daniels Midland. And Russia.

> Even to avid Silicon Valley historians, Microsoft is hard to define succinctly. No singular power law product defines Microsoft like Google’s Search, Apple’s iPhone, Amazon’s e-commerce, or Facebook’s social network.

Are you kidding me? Why can't or won't you admit it's Windows? What did MS do to you?

Windows and Office. These were power law products before the other mentioned products were still a twinkle in their founder's eyes

I think 90% of the time I use MS products, I am happy enough but there are some things that definitely would make them suck less:

* Sort out the single-sign on thing as others have said. Why do I have to tell you whether my account is work or personal? Why do you let me have 2 accounts for the same email address anyway?

* Endless tinkering with things like menus, control-panel settings (usually takes me 10 seconds to find the "Apps option", mostly pointless eye candy like in Windows 11 when most people would rather you just fixed the myriad of minor bugs that have never been fixed

* A proper support system, not just 1000s of call-centre types telling you to trying re-installing windows

* As a dev, not always making massive changed in .Net before sorting out the bugs they introduce far to easily. An example, I recently updated an MS extensions library and because they had added a load more constructors to a class, my dependency injection started failing but, of course, there were no obvious helpful errors.

* They have never sorted out the licensing. It is horrifically complicated and very expensive and would be easy enough to fix if they cared. Yes, we get that you are trying to make sure we don't avoid licensing by purchasing large multi-core machines or using VMs but you could do a better job

* Harmonising their hundreds of customer-facing sites like MSDN, Bizspark, Outlook, Office 365, Product Feedback etc. Even the design is all over the place but in some places you can login with the "wrong" account and it still lets you into a new account. Not cool. If I used the wrong login, tell me it isn't registered so I can find the correct one.

On the other hand, I do like their more open culture and you are more likely now to have a conversation on github with real devs who can explain some of those crazy choices they might have made.

> Sort out the single-sign on thing as others have said. Why do I have to tell you whether my account is work or personal? Why do you let me have 2 accounts for the same email address anyway?

Because the classic use-case is that people in corp environments have set up MS accounts on their work email addresses dating back many years, while the company had on-premise AD/Exchange and actually bought Office licenses. Now, the company switches to O365 + Azure AD because the CEO is following the hype train... and there comes the problem ahead: Microsoft is in a difficult legal position as they can't just blindly "merge" old accounts with corporate accounts (as the creators of the accounts may have created them on private devices, the accounts hold private data, ...), so the cleanest solution is what Microsoft has done: keep the "old" accounts around and don't contaminate them with corporate.

What most analysis if Microsoft misses, and I think this one does as well, is that Microsoft is mostly not an innovator. What they have always excelled at is execution. They were late to the internet, phones, and now the cloud. But given their foothold (in the IBM days, this used to be called Account Control, but we don't whisper that anymore). So when it was clear that cloud was the thing, they came up with Azure, and as usual, are executing on this very well.

Because of this, I don't think it is a good example for startups to follow. I think that successful startups are innovators.

Microsoft is an innovator that is very bad at execution. Tablets With Tablet PC, smarthones with windows ce.

When Microsoft tries to innovate it fails : it's not an innovator

How about https://joeduffyblog.com/2015/11/03/blogging-about-midori/

I think this is a really solid analysis and strategy. I want to briefly expound on the data side which is a huge growth area: mSFT would do well to SHOW the rank and file analyst and developer machine learning in context of their daily work, not TELL CEOs about its transformational possibilities. They own the world’s most popular surface for interacting with data. Why not make Azure’s machine learning be keystrokes away? Give away credits and training, suggest use cases, provide actual value. Sell a $0.000000000001/unit cost that makes it safe for anyone to try in their daily work. It doesn’t need to do much, some cool forecasting, predicting the value of the next cell inline, etc would be delightful. I’m a but disconnected from their ecosystem at this point but I see a window where the can captivate the long tail with pragmatic ML and assert their centrality for the coming decades.

You might be looking for Excel's "Analyze Data" feature, introduced in 2019 to O365 with natural language integration in 2020.

>Acquiring Replit is also worth exploring given its strong positioning among young developers who are still malleable.

Microsoft already has top compiler / programming language / tooling engineers / people

Why would the need Replit? Is the product good? Why not JetBrains?

Replit is an online IDE, it'll fit well with Github/VSCode/Azure devops/NPM/Typescript. Something big and integrated is coming I think.

Ultimate thin client. No more code on your machine. Give employees a $100 netbook.

Sounds awful

Absolutely awful. Doesn't mean it isn't coming.

My company is already here. All code is built and tested on our cloud. Guess what, everyone still gets a $2000 MacBook even though a chrome book would work just as well. Developers don’t like to feel like their employers are cheap.

VSCode already works in the browser and is already integrated in the GitHub website. Just open your repo and press “.” on your keyboard.

I really don’t think they’ll acquire a IDE company when they have the most popular one already

No surprise, given that it started its life an an Azure editor (Monaco editor).

GitHub codespaces I think it's called (I was researching source hosting recently).

"There's a horse in Redmond that always suits up and always runs, and will keep running." - Tim "Apple" Cook

Microsoft needs a CEO with "Microsoft" as his/her middle name.

Mike Rowe might be available...

That was beautiful

"No singular power law product"

Uh, what?

1) owned x86 OSs for... 40 years and counting

2) office

I'm not MS historian, but those two effective monopolies (the first was leveraged for the second) have been the core of MS profits for decades.

Basically-free windows as I understand it fueled Azure to the third pillar only in the last... 5 years?

Not even sure the data is correct on Azure, seeing how O365 is allegedly shrinking in revenue.

Questionable in my opinion and might be very dependent on how you separate here.

Are all these AD users that use Azure AD users of Azure? What if their MS 365 licence counted for Office or Azure? I believe revenue is trimmed to make it look good.

> As the surface area of new software markets plateaus, tech will transition towards a consolidation era.

Do you think this is true? We are seeing more tech companies than ever before. But that could just be due to the pandemic and insanely high amount of money in the market.

If consolidation happens, innovation will plummet. This always happens. Any oil-based company hasn't innovated in 20 years. Most of the consolidated industries have little or no innovation.

As any product gets more complicated, consolidation does happen and innovation does plummet.

Think: 1) Web browsers: Ruled by Google and Safari now (+ Firefox so HN won't be offended, although they have 3.6% and it's going to keep declining over time). We have not seen any innovation in browsers in such a long time. 2) Operating Systems: Windows, Mac, and Linux haven't really done anything new in the last 5 years. Win+Mac just keep adding more "telemetry", but no innovation 3) Cloud: it's obviously really complex. It's consolidated. It's innovating today, but seeing the trajectory of browsers, OS and other products, innovation will stagnate there as well.

Should we just take it for granted that tech is supposed to be consolidated?

New fields in tech like crypto, mobile are still very active but are they also just moving towards consolidation?

Well, yes, it's the eternal cycle...

Funny that you would take the example of oil companies : it was the small tight oil startups that did the latest round of innovation (in hydraulic fracturing, though it was a "take already existing innovations and integrate them" kind of innovation) and captured that market (even though they never became profitable as an industry), and some of the "7 sisters" themselves come from the antimonopoly forcible splitting of Standard Oil into 34 companies... (when the forced splitting of the GAFAMs ?)

Doesn't windows have an ~80% market share in desktop os? Is it really less of a monopoly than other big tech?

Sure, but the desktop market has been shrinking. Unless you are in business, all you need is an iPhone, android and maybe a companion tablet are you are good computing wise. There was a time when your grandparents used Windows, but now they all use iPads/phones.

Microsoft is only relevant to me because my employer hasn't noticed all our needs can be met by cloud services and Libreoffice. Seriously, if there were ever a time for Linux on the desktop it's now - in the office.

Or, better than LibreOffice (albeit closed source but cross-platform and 99% identical to the MS Office suite): https://www.freeoffice.com/en/

The only viable threat to MS Office I see is the Google docs ecosystem.

Interesting article on Microsoft. But: in what world are they forgotten??

Yes it read really weird to me. Someone who only uses Macs, iphones and only worked at startups?

Every company I worked for has been using Windows. From POS terminals to C# dev machines.

In a world that thinks developers (any kind of developer) only uses Macs.

Hindsight is 20x20, but Microsoft share price has increased about 5 fold in 5 years, while only doubling the 5 years before that. It had to play a bit of catch up with other FAANG type companies.

Microsoft may have too many internal competing interests to recognize that it has become the Azure company, but it should be clear by the late 2020s.

Any divisions that yell "Developers!" at events already know this. The rest are probably aware.

My only problem is how everything is going to subscription based pricing and requires a cloud account.

This is a prescription for a complete lack of focus. For an article that theorizes that execution takes a second seat to riding an S curve, there's not a lot of focus on acquisitions that aim at another S curve. Instead, it focuses on a somewhat odd competitive advantage - antitrust - and considers scattering bets all over mostly on the basis of demographics.

$10T is treated like a target which is naturally good. But so many acquisitions would be financed in significant part by stock swaps, so all you're doing is getting to the $10T by agglomeration rather than growth.

It's a fun business strategy essay but I didn't find a compelling strategy and the sustainable competitive "advantage" is odd, because it only empowers agglomeration instead of growth and actual competition. Needs more focus on why the agglomeration would be more efficient.

Author doesn't give Ballmer enough credit. Ballmer built out Microsoft's enterprise business - primarily SQL Server, Exchange. Not only did this give Microsoft deep in-roads into enterprises (as author notes), but it gave microsoft insight into just how fast AWS was growing.

The rumor is Ballmer saw the rapidly growing license fees being paid by Amazon for running Windows Server and SQL Server. When he didn't see enough traction being made on cloud initiatives, he replaced Bob Muglia (then head of the Microsoft's Server and Tools Division) with Satya Nadella.

> Undoubtedly, Nadella has been an exceptional CEO.

> With the right strategy and execution, Microsoft can become the first $10T company.What would I do if I were running the company?

I guess unfortunately they don't have OP has a CEO

Even though I moved on from using Windows years ago, I am still a happy Microsoft customer. My wife and I really like the Office 365 family plan. She likes having up to date Word on her MacBook, and we both like the OneDrive cloud storage for backups. A very good deal.

I don’t know if they still do this, but about ten years ago I got accepted I in the BizSpark program and was given $150/month for three years to start a business. I ended up giving that up in a year, but really appreciated the support and thought it was a clever way to get people on Azure.

Having been stuck using Azure due to client's poor decision making, I can honestly say Microsoft need major changes before it comes close to gain support from developers at large.

Number 1 should be documentation that accurately reflects the state of Azure services and tools. I should not be expected to translate names of things, or figure where things have moved since the documentation was last updated. Nor should I have to figure out the "microsoft way" of doing things. They should be simple, not needlessly obfuscated.

Resource groups is amazing though.

Microsoft is a legal & sales company. Not a tech, not a marketing, not gaming. They can sell condoms and every corporation CIO will eventually get microsoft condoms operation manager user license for every employee, a million dollars pack of condom server, as well as condom office and will ensure that payment to Microsoft is most important task of his company and that users should suffer and waste as much time as possible. I don’t know how they do it. It must be a deal with the devil. Or they are weasels & bribers company.

Here's a bad joke: with how many security holes Windows Server has historically had, maybe the line of business that you're suggesting is not the best fit.

On a more serious note: i doubt that they are alone in pushing for marketing and sales, since most large corporations do that: some are just better at not being very obvious at doing that. They are still a ways away from the likes of Oracle, though, at least in my experience.

I wouldn't (always) call doing sales effectively in a corporate environment bribery, though, that's a bit much. Now, personally i'd opt for libre software, be it a database or an operating system, but i understand that for many settings that's a non starter and oftentimes the larger entities will be willing to pay for support and also licensing.

Show on the doll where MS hurt you.

> Without a clear growth driver, Microsoft went on the defensive against the rest of FAMGA in the late 2000s: Bing, Skype, Surface, and Windows Phone were reactive moves.

Eh, Surface shouldn't be in this list, its from the 2010-2020 decade:

> Microsoft first announced Surface at an event on June 18, 2012, presented by former CEO Steve Ballmer in Milk Studios Los Angeles. Surface was the first major initiative by Microsoft to integrate its Windows operating system with its own hardware, and is the first PC designed and distributed solely by Microsoft. [1]

Zune, however, belongs to this list. Along with a plethora of other products. They were each ways (or 'hobby projects') to move profits from Windows/Office in different markets. Many of these failed, or had only marginal success at best.

Windows Phone is just the (failed) successor of Windows Mobile. It was an attempt to develop a capacitive touch-based OS and range of devices (Lumia) but it flopped because of two dominant market players with no market share gained.

[1] https://en.wikipedia.org/wiki/Microsoft_Surface

Unfortunely as names get reused, Surface was initially something else at Microsoft.

A screen based table which one could place tags with IDs that would trigger tag specific menus around the card and a .NET SDK to go along with it.

Ah yes, I remember that. Kind of like the command table in Command & Conquer. That was a cool concept, but it was just a concept AFAIK (nowadays, it wouldn't be so novel anymore). IIRC the one they released was the mega touchscreen for presentations. Or did that have a different name?

They have the same name, hence the confusion when one doesn't know about that MSR work.

I really enjoyed this article. Can someone recommend a book about Microsoft history? I'd love to read about the internal through the ages...

Older book but

Show Stopper!: The Breakneck Race to Create Windows NT and the Next Generation at Microsoft Hardcover by G. Pascal Zachary

There are a lot of insights on this page. Start with chapter one. Written by BiiGs personal technical assistance back in the old days https://hardcoresoftware.learningbyshipping.com/

While not Microsoft history, try "In Search of Stupidity: Over Twenty Years of High Tech Marketing Disasters". Has many stories on what made MS successful.

I’m somewhat surprised Azure is doing that well. I work for a Fortune 500 company. Before jumping on the cloud computing bandwagon we examined the AWS, Azure, Google, etc. and settled on AWS. So far we have been very happy with AWS.

There are a lot of assumptions about Microsoft in this thread and they basically all boil down to people not understanding what it is Microsoft does.

They sell solutions to non-IT enterprise organisations, and while it’s reasonable to assume that a lot of HN simply aren’t in touch with this world, it’s one of the largest markets for IT software there is in the world, and the only real competition Microsoft has in this area is AWS.

With several decades worth of experience in enterprise organisations in the European public sector, and quite a few years in the European financial private sector, Microsoft has been the sole business partner in terms of IT that has always been a solid pick. There seem to be this notion that Microsoft sells shitty products by manipulating IT managers, and that is just wholly untrue. They sell the software people want, and they sell it in a package that includes real world telephone support with their actual headquarters. Not some chat-bot or forum like Google, not some outsourced call Center like Apple, but real phone lines directly to Seattle. When something big goes wrong on any of our user facing solutions, which is basically anything office365 including sharepoint Microsoft will call us, on the hour, every hour, with updates unti things get resolved. This is essentially the most valueable thing to any IT manager in an organisation of 20.000 employees where maybe 50 are IT related, because it lets you tell the organisation exactly what Microsoft is doing to resolve the issue that is currently stopping your business.

The reason Azure was capable of sneaking past AWS and securing itself a healthy market share wasn’t only that it makes sense to live in Azure when you already have Office365, it was that Amazon didn’t realise how much of a deal phone support meant to the European Enterprise market. They very quickly picked up on it though, and are now in some areas like HDPR a better option than Azure.

Mean while a company like Google had office365 before there was office365 and have some arguably interesting services in Google Cloud, but they will never sell anything to enterprise because Google still doesn’t understand how to sell things to enterprise. It doesn’t seem like Google really cares in terms of gsuite or Google Cloud, but they do care about education, and, still they struggle with delivering what we want from them even though they ask us and we tell them. Apple is in a somewhat similar boat, except they don’t really care what we think. They do things the Apple way and never reach out.

Anyway, the result is that Microsoft is a great business to business partner. Especially in recent years where they have inhoused more and more of their services so you almost never have to rely on some 3rd party “gold partner” or whatever they call themselves that essentially all suck and always have sucked. The only partner we have currently is for licensing, and even this is an area that I hope Microsoft inhouses because it’s just such a stupid mess.

In my eyes a lot of what Microsoft sells to private customers is know how. They don’t give Office365 to students because they are nice, they do so because it means that every hire we onboard already knows excel. This means it’s incredibly hard to compete in the office space. Similarly everyone we hire knows windows, some of these people can’t tell the difference between an android or an iOS devices when they call IT support (no I’m not kidding) but they all know windows. I know that a lot of techies want Linux to be a competitive choice for users in areas like the public sector, but those are the people who would need to use it, you can’t imagine how expensive retraining 20.000 employees who can’t tell an Android and an iOS device apparat to use Linux.

So as private users, we’re not really Microsoft customers. I mean, we are, but not really their primary customers. Because Microsoft makes their money in enterprise, and that position has probably never been more secure than it is today. Because what is the alternative to office365? Nothing. And when you already have your AD and licensing tied up to AzureAD and all the other integrations between Azure and windows + office365, then the business case to not use Azure as your cloud environment dwindles. It sometimes does make sense to use AWS, as I stated earlier, but not often.

As such Microsoft along with Apple (who have the private market of non-techies locked down) are probably some or the safest stock in the world of technology. I’d still rather invest in green energy though. Who doesn’t need energy?

Agreed and also a very fristrating thread to read. Seems like most HN people are Silicon Valley or startups?

It goes like this:

- Company is has IT in the 90s and 2000s. They have computers and a LAN and WAN. They use Windows. To manage Windows, Active Directory is needed/used. So Windows Server is also used. For email, the use Exchange. There is just no alternative because you want users to use the same account in the AD.

- Company starts virtualizing their DC with VMWARE later on, but stuff is still on-prem.

- In 2010 Office 365 comes out. Makes sense to have Office 365 host your email and Office as you are already on on-prem Exchange. So they migrate over the emails.

- When Azure comes out, it makes sense to move to that too. Your IT admins already know Microsoft. And you can use your Windows Server licenses etc.

Now if you are starting a brand new company in 2020, sure, use gsuite, notion or something else.

But its gonna be hard to switch over.

And this is not Tech debt. The stack actually works really well.

The only thing is that Teams is pretty heavy as a client, but its amazing in how you can just use Teams for an entire day as its integrated with SharePoint and OneDrive so all documents etc are there.

Second-mover advantage is underrated and often confused with first-mover advantage.

I think of Microsoft as a legacy, business oriented tech company. Other than the second string XBox product, they don't really have any consumer juice anymore, they are all about locking in businesses and forcing technology top down via company leaders. My nephews and nieces use Google products at school and iPads at home, by the time their generation is in the workforce, Microsoft will be like IBM, still rakes in money but pretty much a has-been. The only reason I still use Windows is I'm forced to on the desktop at work, but Linux has already taken the server market.

> Notion: Office 365 for people below age 35. At a minimum, Office should aspire to the ease of use that Google Docs has (Office 365 for people below age 45)

What? Employees use what the company bought. Every company I have worked for has used Office 365. Teams is fine, but Discord is the best. Yes Slack sucks too compared to Discord.

Office 365 now has all the collaboration features of Google Docs and has had for some time.

Trust in MS is also huge in enterprise. Nothing can also touch the management of computers and servers with AD/GPO or Intune. MacOS only has janky MDMs and Chromebooks are not used in real jobs.

This reads like a defense of Bill Gates and Microsoft. Comparing Microsoft to the Roman Empire is hubris. I don't trust anything from Microsoft and the blue screen of death is still too real. Many parts of Azure, especially its data offerings, are vapor ware and/or cr*pware. In return for the initial low low price, the number of developer hours required to keep all the moving parts moving is ridiculous. The fact that Azure doesn't cap you off so you don't get billed for runaway processes screams Azure is an untrustworthy platform.

I still try to avoid Microsoft, although their opinions on Linux have changed (or atleast, they make it seem like they have), them paying analysts to make Windows look better than Linux, is quite scummy imo.

Their attitude towards things like Linux only goes as far as it profits them. It's pure facade. They still tried to shove their own ODF format (which was garbage) in the UK not too long ago. Will twist themselves in a pretzel to prevent (local) govs from switching to Linux, etc Github desktop client for Linux that was almost finished is no longer worked on now they own it, skype was promised to be maintained on Linux only to be ditched a year after, etc, etc

> Airtable

Microsoft does have a play here. It's Microsoft Lists. The consumer version was even announced yesterday[0]. Sure it's not at the same level as Airtable yet, but the value of that market space is not lost on people at Microsoft.

[0] https://techcommunity.microsoft.com/t5/microsoft-365-blog/tr...

Microsoft is so diversified it would take some truly monumental changes or a spectacular collapse for it not end up a 10T company. The question is when not if in my mind.

If people stopped using their weird OS (which definitly isn't one of their strengths) most tech would render irrelevant as well.

However not even Windows 8, or ads in a paid OS made that change happen.

This is a good article, but I'm surprised an article about how Microsoft can become a $10T company doesn't have a single mention of VR.

If I were charge at Microsoft, I would orient the entire company around enterprise VR Computing.[1]

[1]https://simulavr.com/blog/why-vrcs-are-better-than-pcs-and-l...

Microsoft has never been good at building things, period. With the possible exception of Xbox, their products are almost universally annoying to use.

Excel?

He just said it's not good at building software, how can you say that they excel ? :look:

I don't understand. Excel is one of the most utilized pieces of software ever created. Entire business run off of it. It's amazing software by any definition.

https://dictionary.cambridge.org/dictionary/english/excel :D

Well played :)

xbox had its own issues. The red ring of death, for example.

> Microsoft’s identity became murky. Was it the Windows company? Office? Xbox? Developer tools?

Tea became murky. Was it water? Tannins? Milk? Sugar?

I mean I don’t think it’s wrong to say that MS isn’t a cohesive entity and it makes the branding weird at times.

When you have MS as the amalgamation of the success of 100 mostly isolated little divisions which might as well be separate companies then what MS actually does is as murky as what Alphabet does. I think the sanest statement is that MS acts as a big pile of money and shared infrastructure for their “subsidiaries.”

It's an odd selection of divisions to make that point though.

Looking to compare Micro$oft to Roman Empire I see Gates more similar to Caeser than Romulus. Gates/Caesar transformed the company to an empire from a little republic (Allen(Octavian) initial period). Satya looks more like Costantine the great, who became emperor on a crisis period and rebuild the empire around a new credo: Cristianity/Azure

I would absolutely love it if Microsoft developed a line of no-code (tiny-code?) tools for companies trying to build tech solutions. It would be great if we just had generic tech for generic business solutions so we didn't have to keep building software engineering teams to build the same god damn simple components over and over again. Imagine Terraform or ITTT using pre-made plugins and applications and some configuration language. You could download pre-made configuration for common solutions the way you load Resume templates in MS Office.

Imagine never having to actually build a microservice ever again, but just downloading and running some generic-microservice-app that loads plugins for whatever you need the microservice to do, and the only thing you'd have to "write" is 10 lines of business logic. Now imagine downloading a "template" of an entire B2B SaaS, and being able to tweak a couple lines and then run it in any cloud provider. You could launch a company over a weekend.

I am pretty convinced that Microsoft is about to acquire Twitter. Dorsey leaving Twitter is basically setting it up to be run by big tech. Would add 3-4B$ instant ad revenue to Microsoft's business. Without a mobile OS, Microsoft needs to own significant mobile properties in the consumer mindshare and Twitter is one they can easily grow to be much bigger than it is today. They can build an Xbox gaming social service on top of Twitter spaces (add video streaming to spaces). It just seems to make sense given they've been at the table to talk with Pinterest, Discord and Tiktok. Twitter seems like a deal they can get done and it's not unlike a Linkedin. In fact there are probably synergies there too. The stock dropping is also a sign that something could happen. Usually it gives extra room for a premium bump on an acquisition offer. If it's not Microsoft, someone is coming in soon.

Let us pause for a minute and remember that, if it were not for Linus and Linux, we'd all be paying a computation tax to Bill Gates to cross the street.

if there were no linus we'd be in bsd paradise

“FAMGA” seems forced.

MAGMA is the new acronym

MAMAA

Microsoft is a legal & sales company. Not a tech, not a marketing, not gambling. They can start selling condoms and every corporation CIO will eventually get microsoft condoms operation manager user license for every employee, a million dollars pack of condom server, as well as condom office and will ensure that payment to Microsoft is most important task of his company and that users suffer and waste as much time as possible inside this condom paradise. I don’t know how they do it. It must be a deal with the devil. Or they are weasels & bribers.

Would would the advantage be for Azure and AWS to, as the author puts it, “sandbag [conceal] the breathtaking growth of the segment?”

Perhaps to avoid encouraging potential rivals from attempting to build rival products.

Yes this was my first thought, but wouldn’t the potential stock rises from revealing the true crazy growth override that concern?

In gaming MS still has heavy lock-in mentality. When they'll start supporting Vulkan things might change.

I am sure they are afraid, the majority of every game released on steam today just works on Linux.

One huge reason less to stay on Windows for a lot of people

That still doesn't remove the tax they put on developers who release for Xbox by not supporting Vulkan there. But I agree, their lock-in was eroded pretty well in the recent times through tools like Wine, Proton, dxvk, vkd3d-proton and etc.

Compared to what? MacOS/iOS doesn't support Vulkan either. Graphics driver support on Linux is still horrendous, and is a tiny part of the market - most things use OpenGL still. The Xbox uses D3D. The PS5 uses Gnm and Gnmx which are proprietary. The Switch uses NVN which is proprietary.

> Graphics driver support on Linux is still horrendous

Your information is outdated by close to a decade, so check your sources.

And how is usage of lock-in by Sony and Apple excusing MS? If anything all that just highlights who the bad players in the industry are. And MS is one of them.

If you want the latest NVidia graphics drivers on for e.g. Ubuntu, you have to install them yourself by downloading them from the website, and then when you upgrade your kernel, tty1 into the machine and reinstall the drivers because no desktop will start.

I maintained a cluster with 200 compute cards until about two weeks ago. I’m well aware of the flaws in the Linux driver model.

I’m not excusing it, I’m saying that they’re not doing anything notable. Most graphics engines can

If you are gaming on Linux you aren't going to use Nvidia these days (or if you are using it, your next GPU won't be Nvidia most likely). AMD and upstream drivers with Mesa have no such issues.

Nvidia was and will remain a mess, I don't doubt that. But they never cooperated with the upstream kernel properly, so it has nothing to do with Linux, it's Nvidia's problem and a reason to avoid them to begin with.

And compute should get unstuck from CUDA too eventually.

> And compute should get unstuck from CUDA too eventually.

This is just detached from reality. They're so far ahead with library support, and other vendors are not investing in that. AMD's roc* libraries are terrible, aren't documented well, etc. etc.

It's clearly necessary, so it will happen. It's not detached from obvious need to do it. There is work to be done for that, sure. But lock-in is not a good thing for anyone (Nvidia not counting).

Nvidia was very dominating on Linux in gaming too. Now they are not because good effort was put into the open stack. Compute will be next.

This article seems to admire capitalism? And focus only on economics.

Aside from that. It is impressive how it missed the MS-DOS deal with IBM, subsequent dropping OS/2 and finally WinNT. How Microsoft was never forced to compete with UNIX - but Linux - because the Reagan administration allowed AT&T to leave monopoly control and ruin UNIX. It doesn't talk much about developers and actual ecosystems. Buying and investing is a necessity but the core is information technology.

Lot of people want to be and write like Stratechery, but they are obviously not Stratechery.

It's somewhat surprising that Oracle isn't even mentioned in the article.

It's not terribly surprising: https://news.ycombinator.com/item?id=18442941

Oracle is rich for shrewd business arrangements. I'm not sure its considered a good company technically?

With only very little snark, I'm more than happy to nominate many Microsoft products for the same category of top down B2B sales that Oracle occupies.

You don't chose to use MS Teams. Someone sells it to your employees. On a technical level, Teams technically works.. some of the time.

No one picked Azure because they have better tech. You pick Azure either for business reasons, or because they exclusively offer something you need.

I have encountered more bugs, incorrect documentation, and operations mysteriously hanging for 20 minutes in a couple weeks of Azure usage than in years of GCP and AWS use. It is not even close.

And I will refrain, with great effort, from speaking about the confused inconsistent mess that is Azure Active Directory.

Microsoft Azure Active Directory is a great example of Microsoft's habit of using buzzwords to indicate chronology rather than indicate anything about technology.

That's four words, you've got lots of opportunity to tell me what exactly this does, but nope, Microsoft at the front tells me this is from an era when the Windows group wasn't ascendant at Microsoft, in this case the recent era. Azure tells me this is from Microsoft's "cloud" era starting in about 2010 (but the early part of that is "Windows Azure"). Active though is from an earlier incarnation, it's a turn of the century idea. Huh.

"Microsoft Core Directory 365" would tell me this was a newer product, maybe from the last 3-5 years or so, "Windows Directory.NET" is from maybe 2005, while "Windows Directory X" would most likely be a late-1990s product and plain "Microsoft Directory" suggests maybe it's from the late 1980s or early 1990s.

Notice how if you don't know that Microsoft names their products based on chronology you might assume Azure Active Directory is Active Directory but in Azure? Big mistake. Thinking that way can cost you a lot of time or money.

Hilarious. I had no idea Azure AD wasn’t actually AD, but actually uhh, more like OneLogin? lol

Oracle is floundering: their databases are no longer competitive and nobody uses their cloud voluntarily (except perhaps the free tier). The only reason they're still around is that they intentionally make it very difficult to migrate off.

I've had the displeasure of working on a lot of OCI migrations from AWS over the past few months. Unless your needs are very basic (compute, network, storage), you are going to hit a lot of sharp edges and managing your own solutions.

Our products are built around Postgres/Elastic as the backend. On AWS we make heavy use of EKS, ALB, S3, CloudFront, managed Elasticsearch, Postgres RDS, managed Redis, etc.

On Oracle, we have to build and manage our own Postgres, Redis, Elastic, etc.

The only real managed solution Oracle has is Kubernetes, which apparently isn't even roadmapped to automatically cull nodes that enter into a Not Ready state.

I was generally OK with the Oracle migrations. I understand them from the money perspective why the executive decision was made.

The general uselessness of Oracle support when things do go wrong is going to bite us hard.

I nearly lost my shit last week when I had to explain how their own yum repositories work to their support rep on the Zoom call. TL;DR, whenever they come under heavy load from, say, three VMs in the same VCN all installing the same package set at the same time as part of an Ansible playbook, their yum/identity servers start throttling connections and handing out 401 errors. In the end I just put a retry loop on the yum portion of the playbooks, they eventually get enough of the RPMs.

I spent 10 hours last Saturday leading the troubleshooting effort with their network team for what I quickly identified as a region wide issue between OCI Mumbai and AWS us-east-1. There was a router somewhere in the path between the two data centers that was just dropping packets when the MTU needed to exceed 1478. It wasn't Oracle's fault, per se, but the process of getting their team to recognize and understand the issue was frustrating. I needed to get them to understand it so that they could then file a complaint upstream. BTW, if last Saturday your assets in AP-Mumbai-1 went dark on DataDog or you were unable to pull containers from Docker Hub for about 15 hours, this is why.

I don't know, maybe we're too small to get useful support, we're only spending $200k/month on OCI.

I mean, I use microsoft tools even for compile on Linux nowdays.

It's hard to forget. :D

I always felt Azure was costlier than AWS. My perspective is of a solo developer. Does Azure justify its higher prices in the enterprise space with other value adds? Or is it the "no one gets fired for using IBM" thing?

They mostly cater to big enterprises who are used to getting huge volume discounts.

Pretty much that. If you are an all Microsoft shop and want to migrate on prem infrastructure to the cloud, it’s the easiest way to go.

One main obstacle for their cloud business in general and specifically Azure is that the latest GDPR rulings in Europe basically says that it is illegal for European companies to use Azure (or AWS or GCP) if you store any personal information. This may be a hindrance for both MS and Amazon for the next 1-2 years.

Search for "schrems ii ruling" to read more about this.

This seems rather easy to solve if they approach it like they approach China. In China, US companies don’t run their own datacenters, they’re run by a Chinese company. If there’s a problem in that datacenter, the US on calls can tell the Chinese employee how to fix it, but the US person never gets access to the environment.

Microsoft teaches us how to design bad software

I think this article drastically underestimates how hard it is to build and acquire new businesses. It reads like a caricature of a first-year management consultant’s recommendation—step 1, acquire adjacent businesses. Step 2, profit. (Edit: former management consultant)

Set aside the culture piece, which is well described in the thread.

First, MSFT struggles a lot with keeping innovation going and a big piece of this is that in order to get the full value out of vertical integration with your existing product portfolio (and customer base), you need to build full integration with other products, which often handicaps the new product with old protocols.

Teams is a great (organic) example. The concept of slack+zoom+Dropbox is awesome. I think if it was executed well, it would be the market leader (or the other three would consolidate). But my experience doing anything with files in it has been very poor (it eats my files constantly!!!), and I think that is because it’s stuck using SharePoint as the backend, which was great for its time but was not intended for modern use cases.

On the flip side, MSFT has also done a great job of acquiring / building _noncore_ products that don’t struggle with this. GitHub is a good example—there aren’t a lot of legacy core dependencies. But I find it hard to believe that if they bought airtable they wouldn’t try to merge it with excel and get stuck with the .xls/xlsx limits. By the way, if you don’t… how do you sell to legacy clients, which is their advantage?

Second, Microsoft’s history has made it quite fearful of regulatory intervention. Yes, positioning themselves as “we’re not evil like Amazon, stealing your ideas” is a strategy, but it is also because they are terrified that they will be caught doing something like Amazon because everyone there remembers the ‘90s. It is amazing how often the phrase “we have to stay neutral” comes up in my meetings with my (multi year relationship) Microsoft sales reps.

Third, related to this… Microsoft probably would get slammed with antitrust action if they started buying every company with a $100mn market cap!

Fourth, I think this article drastically overestimates network effects. Actually, I think it misunderstands them. Zoom does not have network effects. If a client or vendor (or my boss) sends me a WebEx link, I am going to (begrudgingly) download the WebEx client and use it. There is no additional value to having more users (which is the definition of a network effect). This article is conflating scale with network effect.

As an aside, I thought some of the advice to startups was funny. Capital may be a nice moat, but it seems hard to action on (“ah, I realized what I was missing… I will go raise $5bn for my series A SaaS business!”) as does the concept of cross-selling, which requires multiple (compelling) products, a large and very capable sales force, and years of trust with the client. I think the core insight here is spot on—it is Microsoft’s sales team that is the special sauce here.

Finally, another aside—I think share buybacks are extremely appropriate for large tech companies. I get that people think about them in % terms, but we have _trillion dollar companies_ now. It’s very hard to imagine that the last marginal dollar of profit goes as far in innovation at a trillion dollar company as it does as a billion dollar company. I would much rather a firm like Microsoft return some of that capital to shareholders rather than spending every last penny on innovation, at least in its current form and structure.

The writing is so sloppy.

Just from the beginning:

> Despite its scale, Microsoft is one of the most overlooked companies in tech.

What's the standard of overlookness? I never for a moment feel MSFT was overlooked, by any measure of sampling.

> It is not a beloved consumer brand like Apple, Facebook, Amazon, or Google.

In 90s, MSFT used to be thought as one of the 2 pillars of THE PC industry as a whole. And MSFT was commonly thought to be the more powerful one of the wintel dual.

MSFT lost its glamour when the anti monopoly suit hit them hard.

Then very loosely speaking MSFT lost a lot of battle in the 2010s.

But still, MSFT has a much longer history than FB and Google. Theese 3 enjoyed probably similar scale of love from their users. And FB is the one with the least amount of love in their hayday.

Apple is a different story. They always have a particularly cult like following that skewed every commentators' perspective.

> It was not a venture capital success story: Microsoft was too profitable to raise real VC money, so the founders owned 70% at IPO.

What? BC has been much smaller in the days of MSFT. You should say BC was not favored not MSFT was not relevant to VC. It was after MSFT created the PC market, and enabled Internet, then that the entrapeneriship becomes much cheaper through online economy. Then the VC becomes a central force of the high tech industry.

You are asking the father to be judged by its grandson for greatness...

> It is the oldest of FAMGA, hidden away in a different state.

What?... Why not included HP, DEC, Fairchild then... Of course some is old some is young. The fact that MSFT lives for so long is a symbol of success itself...

Thus mumbling of words are unbearable...

The OP has a conclusion in his mind and constructing all his arguments to reach that conclusion. I mean Microsoft is Microsoft.

Great article but please use GAFAM not FAAMG or FAMGA

Why? Of the options you list, to me GAFAM is the least recognizable. Why pick that one?

It's the one I've seen most often in France, fwiw.

Hmm, interesting. I hadn't thought of it being a language- or country-based issue. Still, this is an English language site (and article), so if that's the reason for the request, it seems unreasonable.

Like the other reply to you mentioned, it is the one most commonly used in other countries to my knowledge. I first noticed it in an article from FT talking with Japanese investors. It was my first notice of it because I had always seen the many others that were replied to my first comment. And upon my cursory search after that article it seems like the agreed upon big tech shorthand in the rest of the world that isn’t the US.

Why not M-FAAG which you can pronounce Em-Fog

honestly I would prefer people just used FAANG seeing as Facebook's name change didn't really do much to the user facing side and the ticker is still FB, but then you have to make the case for Alphabet.

apple alphabet amazon meta microsoft?

MAAAM?

I think MAGMA is the least controversial alternative and works in many languages

The next wave is 3D. Not just games, but in enterprise, shopping, communication. Satya and team realize this, and the Activision-Blizzard buyout was just setting the stage for what's to come. That's why Microsoft's next acquisition MUST be Epic Games, due to their market leadership in everything real-time 3D.

This purchase would cement Microsoft firmly in the upcoming metaverse race, and provide a real set of legs to stand on to face off against Meta and Apple. Microsoft could also go after Unity, but this would make less sense as the acquisition price would be about the same as Epic, and with Unreal Engine they get a software suite that's getting scarily close to photorealism with Nanite, Lumen, and Metahuman.

> The next wave is 3D

Are we back in the 90's?

3D interfaces sounds cool on paper, but a good 2d interface is almost always gonna be a better long term experience.

Have you ever tried VR..?

It's sub-par; the whole expirence is meh. I have to wear glasses so I had to buy lense extensions for my Valve Index because you can't put headset over my current glasses.

I miss out because my internet is so rubbish. I'm awaiting 5minutes just to download a 200mb world, 2Mb ADSL is all my apartment has.

You look like a idiot waving your hands while strapping a big box to your head. Sure, it may get smaller and smarter as tech envolves on, but you still look goofy. You need a large spacious area otherwise your going to mash your hand on your television or other furniture.

Don't get me wrong it can be fun but eh, it's not something I would jump out of bed to get on a company meeting for.

Yes, just games and some novel VR experiences. Not for enterprise, shopping or communications (unless you consider in-game chat)

To clarify - I'm not against VR, I just think it's mostly gonna be use for entrainment, not so much productivity.

Enterprise seems like by far the best use for vr to me. I wouldn’t be surprised if companies start sending all their wfh employees headsets made by Microsoft for meetings.

VR is a fun gimmick. Nice to visit, wouldn't want to live there.

> The next wave is 3D

I wasn't the last 5 times I heard it, what makes you think it will be this time?

Reply on news.ycombinator.com