Microsoft Is Dead (2007) | Hacker News Reader

Microsoft Is Dead (2007)

paulgraham.com

50 pointssampo82 comments

Interesting to see how Microsoft is resurgent again with their Office 365, Azure, MS Teams, Xbox services and vibrant participation in the open source ecosystem since this article was written in 2007.

I recall somewhere in the 2002 timeframe having a discussion with someone that worked at Microsoft. The anti-open source attitude, and the anti-innovation attitude more generally, made Microsoft a place where you only took the job for the money I was told. It was a dark place to work. Your job was to protect the status quo, and that meant you were supposed to prevent progress. I had a hard time believing it at the time, but it was probably true.

I would love to see Microsoft use the work from Flight Simulator 2020 [0] to create a Google Maps competitor.

[0] - https://youtu.be/EOeDTr1x3XI?t=53

Microsoft already has Bing Maps with aerial view and street view

FS2020 uses the satellite imagery from Bing Maps, so if you get lost or need to find a very specific place during the flight, just open Bing Maps and look around to spot the same scenery as around your plane. I located e.g. Nubian pyramids that way.

Microsoft will buy Facebook within the decade, calling it now. Oculus is a great addition to xbox, and they have the technical stewardship to maintain the social network's hegemony. Then, they can work it in to Office 365 and offer an enterprise social network (badly needed).

I am divided into this one. For one, most people are less afraid of Microsoft now than before, which makes his assertion true.

On the flipside, the demise of Slack as an independent company and the slowness of Google cloud-wise shows that Microsoft is not "dead" in his definition.

I am living in 2 tech worlds that couldn't be more different...

The web / startup / open source / constantly changing world of continuous learning and building solutions looking for problems.

And the enterprise world of minimizing risk, meeting all day to kick real problems to someone else, and staying just far enough below radar to keep direct deposit and 401Ks humming along.

Make no mistake about it: Microsoft OWNS the enterprise world and has for 25 years. These people know little else. They do Webex and Skype, not zoom or slack. They use Microsoft Project and Sharepoint, not Trello. They outsource or purchase e-commerce, HR, and supply chain. If any programmer knows anything not Microsoft, they're either "not a team player" or "a god". Hell, even one CEO once told me, "I do not allow my people to use I.T. as an excuse. They have Excel." That's pretty much the thinking and I don't see any evidence that it's changing anytime soon.

Any enterprise people here have similar experience?

Would just add MS Teams to that list, which at my company replaced Skype and Webex, and has a worse UI than Slack and whose custom, non-OS notification windows close but don't destruct on Mojave

Oh, and can't forget wiring all of Production to specifically target MS SQL Server dialect

That MS security blanket is cozy for somebody though

This 2007 blog post neatly captures the anti-corporate, pro-startup hubris of the early stage tech community in the 2000s. We all believed that big, incumbent tech companies were antiquated dinosaurs surviving on momentum alone, and that it was only a matter of time before agile startups swooped in and stole all of the new business with their nimble methods and superior technology.

This sort of hubris is common in startup and VC circles, where it's a necessary part of the narrative required to convince people that a scrappy startup has a chance at upending an established industry player. There is obviously some truth in it, as we routinely see new startups come out of nowhere to capture significant market share. However, these narratives can also be a trap that leads startups to underestimate the power and abilities of their multi-billion or even trillion dollar market cap competitors.

Keep in mind that the 2000s were a different era in the tech industry. If you had a tech product or service to sell, there weren't nearly as many tech-centric peer companies to sell to as there are in 2020. The big money focus was on enterprise sales, not scrappy SaaS startups with <$100 monthly pricing plans on their generic website pricing page. It was easy to feel that big companies like Microsoft just didn't get it, and that they were going to miss out on the new era of tech.

As it turns out, there's a lot more to being successful than simply building the best technology faster than your competitors. Companies like Microsoft were slower to catch up, but their distribution channels and finely-honed sales processes were harder to compete with than us startup people wanted to admit. It also turns out that with hundreds of billions of dollars in revenue to work with, big companies really can overcome their inefficiencies through brute force to deliver quality products.

This was also before the 2010 lawsuits against Silicon Valley companies for colluding to keep engineer wages down. Microsoft wasn't part of that lawsuit, but high engineering salaries in general weren't as easy to come by in the 2000s as they are in 2020. When this was written, working for a startup didn't necessarily mean you were going to earn less than working for a big company. Funding rounds were also smaller and fewer (less dilution) and engineering teams were generally smaller, meaning your odds of having worthwhile equity in a startup were better than they are today. The modern era of ultra-high FAANG compensation has been very effective at funneling top engineers into big companies and away from all but the most overfunded startups.

Some addenda by the author himself: http://www.paulgraham.com/cliffsnotes.html

> ”If you're the first to call something, you'd better be right. If the monster turns out not to be dead after all—if they can somehow morph themselves into something startups have to worry about again—I'll look like a fool. But I'm willing to take that risk.”

I do think that startups worry about Microsoft. Slack worries about MS Teams I believe. So I think Paul Graham looks like a fool on this one. And that’s ok, he knew the risk, and even laid out the terms which would make him look like a fool.

Slack worries about MS Teams in the same way you should worry if a person trying to kill you is in your house with a shotgun and you're barred in the bedroom.

It's hardly any better since he still writes "But it probably means there is trouble ahead"

A little weaselly since that can be said about any company.

> Actors and musicians occasionally make comebacks, but technology companies almost never do.

Microsoft had one of the greatest comeback. Just wondering if Oracle and IBM are capable of this ever.

> Microsoft is Dead: The Cliff Notes

This title does a great job of setting the essay's arrogant tone. Like the Cliff Notes, this essay is written for people who didn't bother reading the original Microsoft is Dead. It's not a walkback, not even a defense, just a summary.

> When I wrote that Microsoft was dead, I didn't mean it literally. I couldn't have. Companies aren't alive, so they can't die.

Companies can go bankrupt.

> In fact "Microsoft is Dead" was what we in the trade call a metaphor. I meant something else.

> What I meant was not that Microsoft is suddenly going to stop making money, but that people at the leading edge of the software business no longer have to think about them.

To say that someone is dead, if not meant literally, means to predict that he's going to die soon. It does not just mean that people no longer have to think about them. Metaphors are not arbitrary. Metaphors, like words and phrases, are usually based on conventions. You can't just write something and then claim that you meant some other thing. That's not how metaphors work. Note also the implication that people who don't understand that saying "x is dead" means that people no longer have to worry about x just aren't "in the trade", don't even know what a metaphor is.

But even the claim that people at the leading edge of the software business no longer have to think about microsoft turned out to be wrong. Maybe it's time for "Microsoft is Dead: The Cliff Notes: The Cliff Notes" to explain that this claim was also meant metaphorically?

Totgesagte leben länger.

Dead talk lives longer?

Dead-declared live longer.

Not sure why you have been downvoted, yours is the only correct translation.

More like "Those who are said to be about to die live longer", I guess? (Love how compact German can be, by the way)

Nonuniformly so. I wonder if there's a single word that would stand for "live longer", that would expand to "live longer than [the other thing]" in longer sentences?

"überleben [other thing]"?

Yes, you can use "überleben" in this sense of "outlive someone/something". It's not very common though, "überleben" is usually used for "survive [some dangerous situation]".

I think that's mostly just because it's not a sentence you would be saying that often. E.g. when talking about parents outliving their children you'd use it

The moribund live longer.

I think when people talk about Microsoft declining it’s from the perspective of consumer oriented products. The common tech wisdom is they’re as vital as they’ve ever been, but have shifted their focus to mostly enterprise or b2b products (that normal people don’t know or care about)

That is completely true of course, but people said the same thing about IBM too. The rebuttal is always something like “but Azure” and yet they’re undeniably not the primary leader or innovator in the public cloud space. Most of their products are that way, outside of Office and Exchange and AD. Microsoft has gone from leading, to skating where the puck once was, and that isn’t a good sign imo

I think the core assumption of this essay is that innovation can be had only with startups.

This is something I personally believed as well in the past, probably because of my experience with highly hierarchical companies and middle management.

This idea changed after hearing how teams work independently from each other at Amazon.

I'm not sure if this is really the case at Amazon, but I believe that if a company is able to organise itself so that you essentially have an internal market of tiny startups instead of teams (so they have freedom to pursue their products, they can fail and they can be hugely successful) you can replicate most of the startup benefits.

I've been trying to propose this model in the companies I've worked at - but generally there is always some Napoleone on top who vetoes risky bets or force teams to work on non productive endeavours.

You can get approved some practices that give more freedom to individual contributors, but important decisions always need to be directly or indirectly approved on the top.

I'd be curious to know what M$ did internally

> I've been trying to propose this model in the companies I've worked at - but generally there is always some Napoleone on top who vetoes risky bets or force teams to work on non productive endeavours.

You need a separate R&D division that plays by different rules than the established product division.

New product/market R&D is expensive, slow, and risky. If a company swings too far to the sales-oriented mindset, new R&D ventures can look very unappealing under typical models. Sales-minded executives would rather focus on the investments they know, such as hiring more sales people or increasing advertising expenditure.

Having a separate division responsible for R&D allows them to play by different rules and take more risky, longer-term bets. However, this is easier said than done. These R&D budgets are usually the first thing on the chopping block when sales are down and executives need to clean up the budget numbers. It's also extremely difficult to transition new products out of R&D and into their own, sustainable divisions. Some companies, executives, and engineers get hooked on tinkering with new product R&D and forget that they have to turn things into long-term product offerings. This is how you end up with constant product churn like Google's ever-expanding graveyard of abandoned products.

2020 - /s/Microsoft/Intel/g

No one realized how MS would automatically win with office inertia in enterprise. Even now, 13 years later, O365 still struggles with features I was using on Google Docs in collage, but it works well enough.

I’m surprised it was legal for MS to attach a Slack competitor to O365 and grow their monopoly so easily.

Are there specific downsides to Word Online compared to Docs? I imagine collaborative editing might not be as good, but how bad can it really be?

The collaborative part is still quite buggy. Lot’s of time documents we are working on will revert while you are working on them for who knows why. Very annoying when it happens with legal documents.

I understand that they had a harder problem working from where they were and trying to make it cloud, versus GDocs starting cloud from the ground up.

As it was in 2007, it is only "mostly dead". I am reminded of this every time I start work on my Dell machine running Windows 10, which is also maybe mostly dead, but not dead enough? Microsoft seems to still occupy living spaces in the enterprise desktop marketplace. So what exactly is mean by death here? Innovation? Dominance? Marketshare?

And death where? The application space? The cloud space? Most comments I've seen here suggest Microsoft still lingers in these domains. Its tentacles still manage to reach Cthulhu-like into spaces that weren't expected.

Always funny to read predictions that aged poorly :)

For people that think they can foresee the future of "tech" or whatever else: tell me how around 2000-2004 anyone could have foreseen that Yahoo Search will be completely smashed by Google Search. I mean how could anyone come to a facts-based conclusion that would draw today's picture (Yahoo essentially bankrupt, Google one of the most profitable companies in the world).

2004? If you ever get the opportunity to travel through time, go back to 2004 and find one person with any tech sophistication that didn't think that about Yahoo search.

Oh, and the article was dead on. There's a lot of examples, but the only one you need to consider is that Microsoft now owns Github.

So they had poor search results? Didn't know how to sell ads well? What was wrong with it? You haven't provided arguments.

Those here who think that Paul missed the mark on this one either didn't read the article or (more probably) fall into the category of the person he talked to in the story he told. If you experienced personal computing in the 90's and early 2000s, you can appreciate his message as, "Who are these guys and what have they done with the Microsoft that we all knew and feared?"

It’s funny how you can get into a niche and be unable to see out of it. OSX never really had much market share, but if you were in certain circles (programming, design) you’d have a hard time finding someone not using it.

Ironically Apple’s ownership of iOS, which came out shortly after this was written, is now dangerous in the way Windows used to be.

So how is Ms teams doing?

Crushing Slack

My team tried it out at the beginning of Covid. It performed really bad for me. I've heard it's better now, but that experience really turned me off to it.

Very well, it would seem.

I'm on Teams for close to a year now. It's great, much better collaborative tool than Slack - assuming your company buys into MS ecosystem enough to also use MS Office, Outlook and SharePoint. I keep hearing all these negative comments about it, but my N=1 experience since April disagrees. I'm guessing it's the matter of all these integrations.

Teams is much, much better than Slack.

I don't know if there is something better than Teams, but between Slack and Teams, Teams is vastly superior, even without MS integrations in my opinion.

MS Teams sucks like anything.

The calls are bungled up almost half the time.

I get calls on both phone and laptop, and can’t accept for at least 10 seconds after I start hearing the incoming call sound.

And of course, there’s omnipresent “Poor network” notification, even when network is strong, leading to call drops or in the very least highly attenuated communication.

The moment you start Teams, it has to guzzle at least 0.5GB of your RAM.

Poorly designed software (or should I say bloatware)

I meant it economically, in the sense if it is a threat to startups

My company is pretty slow to adopt stuff fortunately. They paid to retain support for Windows 7 up until mid 2020. It's probably gonna be another year or two until someone even mentions distributing teams.

We use Slack at my new job and I miss Teams. Teams is very far from perfect but Slack is worse in my opinion because it just seems to have been designed from the start with the goal of disrupting your productivity. It feels like I'm fighting a constant uphill battle - mute this, unsubscribe from that etc etc. - I guess the main reason why Teams doesn't feel as much "noisy" is the fact that it's based around the idea of teams - as the name suggests.

Most of Paul's essays stand the test of time. I guess this is an exception to the rule.

Microsoft has reinvented itself again and again. Defying conventional wisdom. Multiple times. Under different leaders.

Was it able to do this because of its culture? Or was it just luck? It will make for an interesting case study :)

Wee see it time and time again, the new kid in town get older and slower because its own accumulated mass is slowing it down.

I'm too young, but I wonder if the thought of Microsoft not being the biggest forever in the 90s is equal to Google not being the biggest forever in 2020

AT&T and IBM in 1985 = Microsoft and Intel in 2005 = Facebook and Google in 2025

There are always differences, of course, no analogy is perfect, but this is the basic pattern. Note that AT&T and IBM both fell from alpha predator status, but ended up in very different places, and the same is true of Microsoft and Intel, so Facebook and Google will both fall, but how far is not so easy to say.

What's truly pathetic about AT&T is that all they had to do to maintain dominance was take the literally hundreds of billions of dollars they were given in tax breaks over the past several decades and build out fiber to every home in America.

The mismanagement of AT&T should be a course at Harvard Business School.

Love this: Windows is for grandmas, like Macs used to be in the 90s.

That's hilarious because nearly all the young people live streaming on Twitch are running Windows. I'm surprised when I see a Mac there. Even a wide majority of programmers hanging out in the Science & Technology streams are mostly on Windows although I do generally see slightly more Macs in that directory.

I can't imagine an OS that is very closely associated with pro gaming also being associated with grandmas.

A Mac is very definitely what you'd still give to your grandmother these days.

You were not kidding. I checked the top Science & Technology 18 streams. 14 were using Windows.

This is such a laughable take that it could only be expressed on HN...

Anyone who wants a one-stop shop for doing computing is forced into Windows 10 by default.

Don't believe that? Name another OS you can play AAA games on, then fire up professional software suites (Office, Adobe CC) to do nearly anything creative / office / analytic-related, etc. so forth.

This article is idiotic. That statement is idiotic.

Microsoft should be fucking terrifying to anyone with any foresight. Yeah, they're not Apple... it takes them 3-4 tries to get shit "right", but once they do, its fantastic.

The entire Surface line up is proof of this. Until the M1 MacBooks, Surface Book 3 was hands down the best laptop you could buy. Sleek, powerful, good battery life, and the pen support basically makes it an iPad Pro / MacBook Pro in a single device.

Microsoft's first ARM chips will be garbage, sure. But their 3rd and 4th ones won't.

I'm a little terrified at the walled gardens that are going up, and everyone else ought to be as well. A world where Apple chips run on Apple OSes and Microsoft chips run on Microsoft OSes and Amazon chips run on Amazon OSes and who knows who else is going to jump into this ring, but I'm not sure this is the best direction for computing.

> Name another OS you can play AAA games on

I'm very surprised that Apple hasn't tried to tap into this market. I miss gaming since I'm on exclusively on Mac. Once in a while I get an urge to set up Steam etc., but then the effort involved makes it tough for me to overcome the inertia so I leave it at that.

Why don't EA Sport etc., launch native app on Mac? Would someone from the gaming domain/industry comment?

"Because there's not enough Mac users to justify support," is the official tag line, but we all know that's bullshit.. the answer is, "Because we can't rake in cash hand over fist."

Microsoft might not be the leader that it once was but they've done a very good job of reinventing themselves. It's a little unfair to call them dead.

> It's a little unfair to call them dead.

At least start to read the article.

> They still make a lot of money—so does IBM, for that matter. But they're not dangerous

It's not "unfair" to not match your definition. The author is making the case about a particular definition of a predatory force in software, at large.

There's perhaps an argument that they are dangerous in a way we haven't yet seen, due to their involvement with the US military and the corruption inherent in that terrible choice.

They still have a tremendous amount of power and other companies should still fear them. For example, Microsoft Teams versus Slack.

>involvement with the US military and the corruption inherent in that terrible choice.

Can you elaborate?

Perhaps they meant this https://www.theverge.com/2019/10/25/20700698/microsoft-penta... ?

In a sense, they aren't dangerous because you wouldn't dare to compete with them in some areas. Most notably, every freakin business in the world uses their Office Suite and there's no serious competitor. Think about that... software that everyone from college students to accountants to lawyers to investment bankers uses and there's only a single vendor.

In some of their other businesses, I'd argue that they are very scary. If you work at AWS they are a formidable competitor. If you are in the video game industry, they are a behemoth. But very few startups can really compete in this space, other than for the niches.

Microsoft isn't dead, but Windows is. They've allowed the web, and more specifically Chrome, to take over for desktop apps.

On the other hand, Azure is the closest thing to an actual competitor to AWS, and in that sense Microsoft is very much not dead in the sense that pg means it.

How is Windows dead? How else does over a billion people access tools like Chrome and Google Docs, if not through Windows? iPads? Android Tables? Windows don’t have to be the dominant force, they just have to exist as the glue (or foundation) for everything else to work.

Mobile is getting increasingly important. I'm writing this from an Android tablet.

"They've allowed the web [...] to take over for desktop apps."

Nope. Once the net became fast and mobile devices became prevalent, it was inevitable.

True, they were late to the game, but they still had enormous piles of cash, existing relationships with virtually all corporations, and a strong brand.

Nadella finally stopped leaving all the money on the table by pushing to embrace open source, web-based apps (Office 365), and Azure, and now Microsoft is back to being the third-largest company by market cap.

>"Microsoft isn't dead, but Windows is. They've allowed the web, and more specifically Chrome, to take over for desktop apps"

Frankly while I do use some web apps / services like Netflix, Amazon store etc. for which web app model is a natural the rest of my the software I use on Windows or Linux is native. The sheer amount of high grade consumer and pro level native applications for Windows is insane. I do not think Windows is even remotely close to being dead.

It's irrelevant for 90% of the users that use a computer to read email, browse facebook, twitter, and dozens of connected apps. I don't think my mom has installed an application in the past 5 years.

My workflow includes locally installed IntelliJ, Confluence, Jira, GitHub Enterprise.

>"It's irrelevant for 90% of the users that use a computer to read email, browse facebook, twitter ..."

And somebody on HN told that 90% of the people use Phone/Tablet for those tasks. Not PC.

From what I see in practice PCs and native applications are abundant in way too many work places.

> And somebody on HN told that 90% of the people use Phone/Tablet for those tasks. Not PC.

That too. I also use a browser (or Workday's Android app) to do a lot of things that used to require a physical trip to HR or an unpleasant session with SAP.

A lot of professionals, depending on what area they are in, need to run tools that work best when locally installed (even though I saw some interesting things going on with CAD over web interfaces that'll end up being like Google Docs in terms of collaborating over complex designs). Besides, it makes it possible not to have a ridiculous 17" laptop heavier than most boat anchors and still run the heavy stuff on a beefy server somewhere else. Locally installed software is also a bit of a pain for the user, as they need to jump through different hoops to keep everything up to date, something that is simply not there with web-based software.

When computers got high speed always on connections, distributed computing became the norm. Kind of the same we did with X11 in ages past (and I'd gladly do again - I just love my IBM RS/6000).

When's the last time you read about a hot new Windows desktop app on HN?

There is life beyond HN and Web Development in particular. To me my photo editing software for example is way "hotter" than the one available on web.

And yet we are discussing whether something is "dead" according to pg, which is to say irrelevant to the startup hype train.

The article is from 2007, right in the middle of Ballmer's lead when I think the writing on the wall seemed much different from today. :) Hell, didn't the Microsoft stock take a jump when Ballmer left before investors even knew who'd succeed him?

I think Nadella took Microsoft for a turn no one had expected, especially tackling open source software so differently that they now drive the leading open source hub on the web. Given this, I'd say this article is an interesting historical document that bears little relevance today. It's one that would probably hold true if Microsoft had never reacted and still tried to drive their company mostly through Windows and Office sales.

The one point that I think still stands though is that "No one is afraid of Microsoft anymore" -- this is true especially because they couldn't enter the mobile market like they intended and that's where computing is radiating from today in so many respects (and especially those of which where Microsoft has traditionally often operated in).

But they're absolutely alive and kicking!

I don't think their contributions to OSS, or them buying GitHub or NPM, says much about them.

Sure, developer's perception of m$ improved, but the what I find surprising is then buiding Azure / Teams.

They are the second most valuable company in the world, right behind Apple.

A lot of people here are too young to remember the overwhelming influence Microsoft had on the industry. As such they can't grok the intention of the essay.

If Microsoft introduces a new product today, a new browser, a new toolset, a new standard -- even a new OS -- it Just Doesn't Matter to most people. You can be perfectly content ignoring them entirely and if it eventually rises through the market naturally, maybe then give it a bit of attention.

In Microsoft's halcyon -- smaller revenues but exponentially more influence -- Microsoft's every move shifted industries. You'd sign up to every beta program and read every tea leave to know what they'd do before it kills your product or renders your work futile.

That Microsoft is very much dead and buried.

Paul Graham is dead too then.

Did Zach Tellman post this?

The companies that killed Microsoft: Google and Apple, have more than replaced them.

Reply on news.ycombinator.com