Hacker News Reader

Comment by pgt | Hacker News Reader

Given the growing Google graveyard, why would anyone who is serious about building product consider working at Google? Guys, you are destroying your hiring pipeline.

I have stopped using Google search (DuckDuckGo is great), Google Chrome (Firefox is great), Google Domains (use Gandi / Route53). The most difficult products to shake are Gmail, Docs and Sheets, but I am gradually weaning myself off.

The only signal that seems to get through to corporates is to vocally withhold your funds. PMs at Google probably underestimate the % of people on HN who control the company purse for software spending. Personally, there is no way I can risk locking myself or my employer (that just raised a Series B) into any Google products at this point.

If I were in charge at Google, I would put engineers and PMs / POs on call immediately so they start talking to customers again.

View on HN

> serious about building product consider working at Google

my view is that if you're serious about "building product", then you wouldnt work at google, or Facebook, or for anyone else. You would build and run something yourself.

If you like the "building" part but are indifferent about "product" then somewhere like Google seems great.

Maybe google lets you focus more on the product part? If you run something yourself, most of your time will be building company, not so much product.

  > if you're serious about "building product" [...] You 
  > would build and run something yourself.

Perhaps as a side project, something useful just for yourself. Lots of developers scratch itches they can't scratch at their day job that way; new tech, greenfield ideas instead of maintaining someone else's garbage, etc.

And sometimes those devs go into business for themselves and realize that they don't get to spend all their time building product -- now they've got to build the marketing page and write a bunch of copy and take a ton of screenshots (of features they know they're going to change), and figure out how sales funnels work, and all the other things that go into running a business.

Building and running your own company is the fastest way to get out of pure product development I've ever seen.

"building a product" and "building a company" are two different things, but i agree that for some kinds of products and scale, you need a company to make it really successful. I think about all the "lifestyle businesses" that are examples of products that don't need full on "companies" to exist.

I think that if you are passionate about creating a product for people to use, then you will see through on all the other aspects of making that product survive and be used.

Otherwise, what you're interested in is "development", and big places like Google will give you plenty of experience in doing that.

This is not a value judgement. Personally, I have no problem working at a company where I'm paid to engineer products from the ground up. Maintaining things is boring. But that's just me.

Generally agree, but building product for a wealthy established company can give you an opportunity to reach a wide audience of users with your work.

Why, if you want to keep building an already successful product, like Kubernetes, or Golang, or Chrome (or React, PyTorch, or VSCode, etc) you can join a megacorp and go on. These are going to be around for probably a decade or two.

If you want to build an original product, then yes, it can happen not to align with a megacorp's priorities, even if reasonably successful. Found a startup, try to sell it to a megacorp (if it's Google, it will likely discontinue the acquired product anyway), or keep it to yourself if you can afford it.

Actually, let me put my money where my mouth is. Just signed up for Fastmail. Email and privacy is worth way more than than $5/month. I know not everyone is in that position, but I pay more for Spotify and GitHub already.

Getting a Fastmail was my first step toward de-googling myself. Have been super happy with them. They are really stirring email in the right direction. Their effort on JMAP is also very much appreciated.

Get a fun custom tld while youre at it. Fastmail has very good and convenient custom domain support.

I did that when getting rid of gmail (have firstname@lastname.email), it still makes me feel cool and defiant every time I enter it somewhere.

I use surname.io and I can't believe the amount of people that ask me if .io is not a mistake. Makes me realise how much I live a bit in a tech bubble.

I have me@firstname.lastname -- so many double takes when I give it out in person or over the phone

My surname ends in "io" and I've been trying to get my <partial_surname>.io domain for years. It's a shame that it's being squatted, especially considering the only people with that surname are directly related to me (great grandpa legally changed his last name to something made up when he left a certain criminal group) and don't do anything with tech. I've been relegated to using a <surname>.me domain.

If they are related you should find them and ask for it! Have you tried to whois the domain?

About the tech bubble, if you are on HN, you are already in the bubble :D There are so many people that have never ever heard of any domain that is not com/net/org/country

Unfortunately, they're not related. My surname is also pretty short, so the first part (e.g. minus the "io") is set up to be a decent acronym...for something.

Want to be really cool and defiant? Go with gmail@firstname.lastname.com - and then you can enjoy explaining to people that no, you didn't make a mistake... ;)

Fastmail is amazing. I have like 20 domains I have from old side projects I can easily setup email + catchall addresses from and go wild with them. So much customization + UI doesn't look bad as Gmails (I switched after Gmail's horrid design update to include all that padding) and a ton of functionality with Apple/etc. devices.

And you can host websites directly from fastmail! Can't ask anything more from fastmail.

This was the first and best step I have taken in getting off Google services. And, as a plus, Fastmail is _freaking awesome_. I cannot believe I allowed myself to have ads in my _email_ client.

Switching email addresses is somewhere between very hard and impossible. I have a Fastmail, which I now use for new sign-ups, but I have lots of accounts on my Gmail address which I don't want to lose. So effectively I'll have two email addresses forever. Worse things have happened, I suppose.

1 - get yourself a domain and use that as your mail address going forward. Then your mail address can be independent of your provider.

2 - Point it at your mail provider, say fastmail.

3 - Tell gmail to forward all your mail to your new address.

4 - Ask your friends to use your new address. Make sure your mail client uses it as the from address.

4 - <wait a little while and then>: Put a rule in your mail client to highlight (flag or colour or whatever) all messages that were forwarded from gmail. This will help you figure update people, companies, mailing lists etc.

That works fairly well for most needs, but you will more than likely hit an issue with that one website or company that uses your email as a primary key, and it either can't be changed, or you have to contact them to do it manually on their end.

Well you can leave the gmail account around for these stray cases or, you know, contact them.

You can (almost) always change your email address. Just connect to the website and it's usually an option.

On email providers - I moved away from Gmail for my personal email address a few years ago, and now use iCloud. I've had very few issues with spam, and there are no ads. My personal email has much less volume than work, so I haven't really missed the advanced features of Gmail.

From a privacy perspective, how does iCloud compare to others like Fastmail?

The disappointing thing about iCloud for me is the nonstandard 2FA.

No QR code you can scan into an app of your choice. No FIDO/U2F. No printable backup codes. You use your iPhone and if you want a backup, you'd better buy a mac.

As far as I know, Australia forces email providers to give them backdoor into email:

https://fastmail.blog/2018/09/10/access-and-assistance-bill/

https://www.theverge.com/2018/12/7/18130806/australia-access...

I don’t know how it affects Apple or Fastmail’s non AUS customers. I still pay Fastmail just to support their mission.

https://www.homeaffairs.gov.au/help-and-support/how-to-engag...

You're not getting any privacy with Australian services. They require by law backdoors into everything.

> you are destroying your hiring pipeline.

why do you think everybody working at a software company cares about building products? I'm in this field to make a living. The only thing that would turn me off from a company that compensates like Google is a bad work life balance.

As a user though, I'm definitely turned off. This year I've started detaching from google as much as possible, and it's great. the main sticky point for me is how to replace Google Photos.

You say you're turned off enough as a user to divest your personal life from Google.

Who do you think funds the salaries at Google that feeds the compensation? It's you - the customer. Some of these customers are slow-moving businesses, but as an individual you are living in the future.

A decline in revenue will follow a decline in product - the question is by how long for a Titanic the size of Google.

I gather that your argument is if I don't have confidence in google's products, I shouldn't have confidence in their success as a business either. If that's correct, I think that's a reasonable argument but I'm not really concerned about whether Google as a business will fail 5+ years from today. I would honestly be shocked if they failed in less than 5 years, so I wouldn't hesitate to take their compensation.

This is all hypothetical of course. I have no idea what their work-life balance is nor have I tried applying to them

> Guys, you are destroying your hiring pipeline.

Not really, as long as the pay is there, devs will come.

Spending a lot of money helps less than you might think. You can definitely get well qualified people, but ‘qualified’ is only a proxy for what you want. For example if people leave after 2-3 years you lose out on massive amounts of institutional knowledge. But, more importantly you select for people that are very good at hacking your processes rather than your problems.

Which I suspect is why their search for example has gotten so bad. Google is doing a great many things that look good internally that make search less useful.

Some people are willing to take a principled stance. I have personally turned down significant offers at a cost to myself, including:

  1. Building an HTTP proxy that inject ads into open WiFi networks, e.g. hotel lobby WiFi.
  2. Forking a browser project for a gambling company that would hide their competitors' ads in return for providing free data to East African gambling addicts.
  3. Building a website for a health quack with completely made up medical claims.

None of which seem in the same league as "turning down a little used print protocol."

It’s not about any one project, it’s knowing that your work will either sell more advertising or be tossed aside arbitrarily harming anyone that dared to use your service.

The things that make Docs and Sheets and such attractive aren't really the ability to create and edit documents -- we all have perfectly adequate versions of that on our laptops and desktops already -- so much as the availability of those documents on a shareable networked infrastructure. "Here's the URL to see my doc, here's the URL to edit the same copy in place" are the most powerful features.

If, say, Backblaze or Dropbox wanted to put a higher-value service on top of their storage infrastructure, I think that would be a good way to go.

Like onecloud or nextcloud with libre office plugins? Or office365?

Have you used Office 365 and OneDrive? You actually get a nearly full blown Word and Excel with the same file sharing features :-)

You can link the desktop versions to OneDrive and get the actual full blown word and excel.

Is buying desktop versions still a possibility?

Office 365, 5 desktop install (at one time), ~$99 a year. We do one subscription for our family per year. Gives 1T of onedrive space as well.

I personally have had very bad experiences with collaborative editing on Word 365, with it locking whole sections of the doc because someone else is editing them, while in Google Docs even editing the same sentence (e.g. adding items to a list) at the same time Just Works™.

How long ago was that? I've been thoroughly impressed with the installed o365 apps' ability to live multi-user edit these days.

Ah, that's good to know. I last tried it about a year ago, will try again

Dropbox does have Dropbox Paper.

My understanding is that Google tends to have internal metrics around all of this stuff, if the data is coming back that usage is declining enough - they usually re-prioritize their efforts. Not necessarily saying I would do the same, but I believe that's how it's supposed to work

Sure, but even if every product closure they only annoy 1% of their users, a hundred product closures in, that's everybody.

Their data-based model fails to account for the fact that they're getting a reputation for not being dependable.

This is the single reason why Google Cloud will never be successful, at least in the next decade. A huge number of software engineers have been burned by being treated as Google statistics. That works great for ads or search, but Google has a reputation as someone you would never, ever, ever, ever build a sustainable business around.

There is a critical mass of people who simply won't use Google in a business-critical setting, and it's growing quickly. Now, that adds any GSuite office which relied on Google for working printers. Is that a lot? Probably not. Multiply that times each time this has happened before, and times the never of businesses wacked out-of-business by Google's automated algorithms, and at this point, I think it's quickly reaching the point where most teams have a Google horror story or two.

It's fine for purely transactional businesses (ones with zero startup/shutdown costs -- for example, one can arbitrage adwords and referral fees).

My guess is that at some point, Google will realize this, either from data or otherwise, and sunset GCE and related services, probably with very short notice, leaving many businesses hanging.

Short of giving a 10-year guarantee of keeping services running at specific cost and service levels, with real liability (not externalizing costs to customers again), I don't see how Google could address this.

The biggest cultural weakness at Google seems to be the No Human support policy and outsourced regional distribution which fuels a growing divide between engineers and customers.

The myopic salarymen who shut down entire product lines based on usage metrics that seem to pale in comparison to Google Search metrics do not understand complex systems, do not understand startups (and the fragility of growth) and certainly don't understand branding or hiring. If these nerds had ever run a business themselves and spent time talking to the customers who pay their salaries, they would spin out these companies instead of sunsetting them.

A large part of leadership/strategy in tech companies is ex-consultants and MBAs.

Engineers don't call all the shots with sunsetting projects.

If anything, we have a healthy (?) skepticism of product managers too but they definitely do talk to customers, look deeply into metrics and think about branding.

I have totally replaced Drive/Docs/Sheets with Nextcloud [1] + OnlyOffice [2]. You still get online cloud storage and online editors with the bonus of using the widely popular (OOXML standard).

1. https://nextcloud.com/

2. https://www.onlyoffice.com/

As far as Gmail goes, I have heard really good things about Fastmail (as others have mentioned) or if you want to host your own, I highly recommend checking out mailu.io.

What do you use to host nextcloud?

I just have it running in Docker on a DigitalOcean droplet (alongside several other services).

But there are quite a few other options, from renting a managed instance to buying a hardware device with it preinstalled. https://nextcloud.com/yourdata/

For anyone who is interested in self-hosting stuff like Nextcloud but is not ready to dive into the deep end of server management, I would highly recommend checking out https://yunohost.org/. Basically it is a Linux "server" OS that tries to make it as simple as possible to install and manage services like Nextcloud.

It's not just Google:

https://www.versionmuseum.com/history-of/discontinued-micros...

It's $GIANT_CORPORATION value drift.

I switched from gmail to mainly using protonmail. I can highly recommend them so far.

I hope they don’t end up killing gmail one day. I was really surprised when they killed there inbox app and it was killed after few years. I was basically using it as a primary email app for the phone

I don't necessarily think the Google graveyard is that much bigger than other companies, it's just more visible to the public. Any large software company is going to have a lot of projects that were sunset early or canceled before the first release. I don't have statistics for this, but Cloud Print lasted for 10 years, and that seems like a fairly typical lifespan for any product.

It can be painful seeing the fruits of one's labor killed off by an unsympathetic executive, but it is the nature of the work.

It's not about bigger, so much as about more sudden, arbitrary, and capricious. In most companies, products which consistently don't show a profit are phased out, usually with transition paths.

At Google, products that don't show exponential growth are wack-a-moled, offloading costs on customers and dumping them on the ground.

> Docs and Sheets

Try out Office online, insanely more featureful. Scold anyone mailing Docs share links by saying you can't open it, insist on a file instead

> Guys, you are destroying your hiring pipeline.

I actually had a former college approach me about an opportunity to work at gcloud and feel like I am a great fit for the position. The work seems fun and the pay is right...but my single biggest concern is the future of the product in terms of longevity. I am strongly considering rejecting any possible offers just for this reason.

Is gcloud = GCP? think GCP is a longerm product. It would be CEO suicide to announce you are not going to compete with AWS or Azure.

gcloud = GCP in this context, yes.

GCP could definitely be a good product and a real competitor to AWS/Azure given Google's technical expertise. Given Google's recent track-record though it isn't clear that the vision is there.

I love Google sheets but I also work with large enough data sets that it regularly freezes up which is always associated with a frustrating “lost connection” message. I’ve been shifting more and more to Excel because of this. It’s pretty funny to move away from Google because my data is too big.

Well, working at Google has clear benefits - insane money + Google on your resume.

> Given the growing Google graveyard, why would anyone who is serious about building product consider working at Google? Guys, you are destroying your hiring pipeline.

Given that Google is known for using filtering methods with a high false negative rate because they have a great surplus of qualified candidates that they need to weed out, even if it was true, that's probably not a concern for them.

> If I were in charge at Google, I would put engineers and PMs / POs on call immediately so they start talking to customers again.

SWEs at Google are certainly on call for pretty much every product. Not really sure how you expect PMs to be on call considering their general lack of technical knowledge. However, PMs and T/PMs are constantly grooming and organizing bug hotlists.

Disclaimer: I work at Google opinions are my own

Given the growing Google graveyard, why would anyone who is serious about building product consider working at Google? Guys, you are destroying your hiring pipeline.

Actually it sounds great - build whatever you can dream up, but never have to worry about supporting it.

Nextcloud integrates with Collabora or OnlyOffice. Can anyone here tell if either of them is usable and compare them with the Google offerings?

As long as new grads can continue to make $200k out of college, Google's hiring pipeline will continue to be fine.

You're not using YouTube?

For business?

Reply on news.ycombinator.com