Hacker News Reader

Comment by michaelochurch | Hacker News Reader

Google's problem is that most of the work is shitty legacy maintenance. If you can get on one of their machine-learning or research projects and have a good manager, you'll have a fantastic experience (and I say this as a notorious Google detractor). Take the offer seriously but negotiate extremely hard on project allocation because it's really hard to transfer to a good project if you start in the wrong place, so your first project is the difference between an awesome Google career and shit. And most people get shit. Good luck!

ETA: Google would be awesome if it adopted the Valve culture where employees are trusted to allocate their own time, but I wouldn't hold my breath.

View on HN

Do not, under any circumstances, take this guy's advice or in fact listen to anything he says. Anyone who was an engineer during his brief tenure is well aware of just how much he self-destructed and how much he's living in a fantasy world. The only problem here is that to an uninformed observer some of what he says looks plausible.

As for the OP's problem: a lot depends on your options, your experience, what you're hoping to achieve, what you'd like to work on, what office this would be in, etc. There are tons of variables.

Speaking as a current Googler, it is a fantastic place to work but it's not suited to everyone.

tl;dr it depends.

I don't have time to respond to this ad hominem smear. Yes, there was some blowback related to my attempt to save a failing product. I learned a lesson and moved on.

I'm not anti-Google. I want the OP to have a great Google experience and to do that he needs to negotiate hard on project allocation.

As a random HN user with no connection to Google who has seen what happens when you comment on stories about Google here (it's a shit storm), might I humbly suggest that you consider not commenting on stories about Google? It seems to always turn into The Michaelochurch Show when you do. Not saying your opinions aren't valid, but I think your contributions re: google are unfortunately not useful because they inevitably spawn arguments about your character.

Take one for the team and sit out the Google discussions maybe? That way when you have a contribution about some other topic, people will see you as a person and not "that crazy ex googler whose credibility is highly suspect." :)

I don't think he should sit out Google discussions. He had the experience he had. Others have the experiences they have. We're all capable of assessing his reports for ourselves.

I might add we're also all capable of assessing the responses of Googlers to his reports for ourselves, and those responses don't always speak well of Google.

And we've almost all seen what's happened to Gmail, Google Reader, Google Labs and 20% time.

Mayson, can you be more specific? What has happened to them?

Gmail: I can't think of anything negative that's happened to Gmail in recent times (unless you didn't like the UI redesign - there are certainly mixed opinions on that), and certainly the performance and uptake of Gmail has only improved.

Google Reader: Not much new happening, but it's still running. Are you referring to the changes in the sharing model? Whichever sharing model was better, surely you can agree that having one unified sharing experience across Google produces results in a more streamlined product, and at least this was a step in the right direction.

Google Labs: Was closed down, yes. I still don't understand how this relates to the parent post. My thoughts on closing down labs: Labs was supposed to be a way to get ideas out there in the wild before turning them into a fully-supported core product. This was great for the engineers who had built these things (they got to see them used), and the early-adopters who tried Labs products. However, as time goes on and nobody's doing any development work on a particular thing in Labs anymore, there is still an operational cost to other engineers at the company who have to keep the thing up and running. Since it's a 'Google' product, the brand image is at stake, so we can't just let them wither. So I think it's reasonable to decide to have each labs product either shut down or integrated into a core (supported) product. Perhaps it would have been better to chuck all the old labs and start again, with a published policy that things in labs will either be fully supported or completely gone 1 year after launch (or something). The removal of labs is a signal that there's more red tape around launching new products than there used to be, but I think that's inevitable as a company grows, and more than some engineer's weekend is on the line if a lunch goes badly.

20% Time: Still as strong as ever, from what I can tell. About half the engineers I know have a 20% project, and there's nobody who doesn't have one who wishes they could have one. Most of the time if people don't have a 20% project it's only because they find their core job interesting and diverting enough that they don't have a desire to split their time with a side-project. Despite what some people have claimed on HN, your manager can't deny you from having a 20% project if you want one.

(I work for Google, but not on any of the products mentioned above, and I (willfully) don't have a 20% project. These are my opinions and not necessarily those of my employer.)

Despite what some people have claimed on HN, your manager can't deny you from having a 20% project if you want one.

Theoretically, managers can't block you from having 20% time, and there's no official permission process you need to go through to take it, but if your manager says, "I'll fuck you over in Perf if you do a 20% project", then you don't have 20%-time.

What I'm told is that, before 2007 or so, Google actively worked to avoid manager-as-SPOF, and that this is what made Google great. But then they hired a bunch of executives from mainstream large companies and didn't tell them to wipe their fucking feet off before coming inside.

> if your manager says, "I'll fuck you over in Perf if you do a 20% project", then you don't have 20%-time.

I find it hard to imagine a manager could get away with saying that, but even if they did, I think it would be an empty threat. Perf is designed so that the manager is not a SPOF, your peers' reviews are considered at least as important as your manager's review. It would be pretty obvious if there was a big discrepancy between what your manager says and what your peers say, which would call your manager's review into question. The only way this would be a problem would be if your peers give you bad reviews too (and if that's the case, then maybe, just maybe, you need to take a look at yourself).

See, I have absolutely no objection when people speak positively of their Google experiences. A lot of people have good Google experiences, and they should. I didn't. I had a shitty one.

What I find ridiculous is that any time I say something about Google, people leap into ad hominem attacks without even reading what I have to say.

You know what I did to piss these people off? Absolutely nothing. I've never met most of them. I'm surprised they even give a shit about me or what I have to say.

It's never the powerful people getting indicted who crush whistleblowers most with the most fervor. It's other peons who want to stick their fingers in their ears and say, "nanananan-I-can't-hear-youuuuu!" in the face of depressing news.

Maybe it has something to do with the way you say it. Maybe you meant to say, "I had a shitty experience at Google", but you accidentally typed, "most of the work is shitty legacy maintenance". Maybe you meant to say that you were not satisfied to start out with the other Nooglers in intro jobs and prove yourself by working your way up, but instead you said that most people start out in shitty jobs and never escape.

What I've observed is that most people at Google are working on shitty legacy maintenance.

I was in New York, which is generally considered to be slightly above other satellites on the project-allocation heap but far below Mountain View, because that's where all the players are. Most people I knew were on uninspiring work, although there were exceptions, but you generally had to be Senior SWE or higher before you were even considered eligible for the good projects (and even some people at Sr. got shafted).

Google has a palpable "Real Googler Line". Above it, you're free to transfer and work as you wish. If you want to change teams, just decide to do it. Hell, Steve Yegge quit his project in public. Below it, you have manager-as-SPOF and one-assigned-project... banking without the upside. When I was there, the RGL was moving upward to somewhere between Senior and Staff.

I don't know why you have a chip on your shoulder against me. I never did anything to you.

You are maligning my company.

You could not figure out how things worked at Google, and expected people to start following the new guy who had only been there a few months and hadn't actually accomplished anything yet, and when people didn't drop the 100,000 engineering years of accumulated wisdom to listen to you, you started believing that it must be the system that was broken and not you.

And that was fine; you can believe whatever you need to believe. But when you start spreading this delusion outside of Google, you are hurting Google.

Most nooglers get a legacy job to start out. When you prove yourself, and people like your ideas, you will be asked to join other projects. You were only there for a very short time. In order for you to transfer to another project, that project has to want you. Without any accomplishments...

And you quickly developed a reputation as the new guy who was publicly lecturing everyone about how things should be. You told them that their languages of choice were all crap, when they were the ones actually getting work done. Despite how talented you think you are, can you see how that might be a turn-off for any team thinking of taking you in?

All these things you claim are you projecting: 1) Most work is shitty legacy maintenance. (You didn't like your first assignment.) 2) Google doesn't listen to reason. (Google does not give carte blanche to nooglers.) 3) There is no way out of a shitty first job. (You couldn't hack your first assignment.) 4) There is a glass ceiling. (You were alienating people.)

When you present your paranoid beliefs as objective facts, people who don't know any better might believe them. That is why I have a problem with what you post.

You, sir, are deeply wrong.

You are maligning my company.

YOUR company? You're telling me you get your pick of the best projects, and that you have at least one of triumvirate on speed dial any time you want to hash out an idea? You think those SVPs and VPs and managers-of-managers-of-managers you're putting yourself out there to defend give a shit about you? Ha!

But when you start spreading this delusion outside of Google, you are hurting Google.

Google is chock-full of high-ranking people making bad decisions. They are hurting Google. I want Google to succeed. It had a great run and a really great culture... for a while. It would be good for the world if it went back to such a state.

Without any accomplishments...

Google's problem is that there are a lot of influential but underqualified people who think that if you didn't do it At Google, it doesn't count; you didn't do it at all, and you don't know what you're talking about. Now that is the epitome of corporate arrogance. And it led to some absolutely stupid product decisions, which I won't name but they're relevant to this discussion.

You told them that their languages of choice were all crap

Google's limited language white-list (C++ and Java only, with Python deprecated in production) is a bad decision. C and C++ are the right tool for the job in some cases. I like Go, but I haven't done much with it. My problem is more with the short white-list and the bad justification for it than with the languages themselves. (Although there is no justification for Java over Scala in 2012.)

I disliked C++ then more than I do now. The quality of C++ code I fell into was pretty bad (object-oriented spaghetti) but that kind of code can be written in any language, and so can good code.

You couldn't hack your first assignment.

Wrong. I didn't get into what happened or why I left, but it wasn't a performance issue. My manager was unethical and lied to me and it caused a lot of hardship. Among other things, he promised a 3.4 rating to prevent me from seeking transfer (I agreed to stay on his project in exchange for a decent number) and then gave me something lower (making transfer impossible). If he had been a decent person and kept his word, I'd have stayed on the project for 18 months and things would have been fine. But I lost all faith in him after he openly lied to me in order to keep me on his project. (He also bait-and-switched me into joining his project, promising work on the new system but putting me on the old one. I gave him the benefit of the doubt on that one. Priorities shift unpredictably and it's often not the manager's fault. Lying about the Perf number was unforgivable.)

There is a glass ceiling.

I never made that claim.

That is why I have a problem with what you post.

I doubt that's all of it.

Will you two please just stop this?

Why? I find both perspectives very informative.

Please don't let your personal relationships spoil an informative (if sometimes heated) discussion. That kind of kills the "hacker" spirit.

Please don't misinterpret anything Michael Church says as being informative. The man is crazy. Like, paranoid schizophrenia crazy. He was banned from Wikipedia not just for vandalism, but for trying to blackmail wikipedia admins by threatening more vandalism (http://en.wikipedia.org/wiki/User_talk:Mike_Church). He will, of course, claim that this was all a conspiracy against him by the Wikipedia admins, just like he claims that Google management was out to get him. See also http://en.wikipedia.org/wiki/User:Isomorphic/Minions_of_the_....

At Google, Michael Church is infamous. Practically everyone has heard of him. He is known primarily for having posted many insane rants to a company-wide mailing list. These threads literally consisted of Michael Church writing a wall-of-text rants followed by leagues of coworkers replying to say, in effect, "For the love of God would you shut up already?". He had few, if any, defenders, yet he pretty much ignored everyone and would keep posting unapologetic follow-ups. He would say the same things internally that he now says externally. He is the best proof there ever was that Google employees can say anything they want and not get fired.

To this day, Michael Church is still a popular meme. As in, literally, we have a meme template that is a church with an M on it, and we (his former peers) use it to mock him. He is that crazy.

He's supposed to be the crazy one, and yet you're still arguing on this 4 day old thread everyone else has forgotten about? Will you please just let it go?

I was hired at Google at nearly the same time as a certain famous ex-Googler. Both of us were given difficult intro jobs with legacy systems, the sort of projects that chew up engineers and spit them out. It's a chance to prove your mettle.

One of us attacked the problem head on, taking on tasks that would normally be "below" us. In the end, the bad project taught him what he needed to know about Google engineering. He replaced the legacy system with modern infrastructure, which brought him public recognition from the senior VPs, new opportunities to work on interesting projects, and nomination for promotion.

The other one of us decided that he was too good for his job. He very publicly whined about how it was below him and how the only way to get ahead at Google is to start out in one of the interesting jobs. Rather than work his way up through a meritocracy, he expected to be instantly recognized for his natural genius. He got poor performance ratings from his peers and manager, and eventually rage-quit after about six months.

It was the classic Goofus and Gallant story.

That's radically inaccurate but I don't have the time to set you right.

As the author of more centithreads than God himself, you don't have time to reply to a comment on the Internet?

I'm kind of impressed with your restraint, actually.

Of course you don't. You are too busy writing the shiny new stuff.

Well, I was there for six months over the course of two internships in two different offices, and nearly everybody I met was excited about their project. I know you had a bad experience, but I think this is a little bit of projection. I would be extremely reluctant to, based on such a short time at the company, decide that most of the thousands of engineers are working on shit projects.

Since this comment is currently at the top, I feel I have to add that michaelochurch's overall Google experience was an extreme outlier, as is well known internally at the company.

People who are currently at Google will obviously not be an unbiased sample of all people who have ever joined Google.

My experience at Google has been nothing like what you describe. There's a ton of interesting work being done, and from everything I've seen, a culture extremely supportive of internal transfers.

Every time I see you write, i get the picture of a drunk guy in a bar walking around saying "lemmme tell you about the time i worked for google" in slurred speech to anyone who will listen.

Damnit - you're comment made me snarf coffee all over my laptop. :-)

As Google customer I wish there were more strong professionals at Google who like legacy maintenance.

Today I received email from Google AdSense about change in their system. The only important part of the change was not even mentioned in that email: AdSese lost my payment information (alongside with all other AdSense users I assume) so I have to re-setup payment information again.

Strong team would be able to preserve customers' payment information during system upgrade, but it wasn't the case today.

I don't want to speak on behalf of AdSense, but there are often good reasons for this.

In the project that I did for another product, we required users to re-enter their payment information to revalidate the account and ensure that the user still had access to the instrument. That process was done to protect the user's instrument.

As I said, I don't want to speak on behalf of AdSense and I'm not familiar with their processes to speak with any authority.

In any case, the change and the reason should have been very well documented and messaged to you (we go through a lot of internal user testing even on these mundane processes). If not, please reach out to the consumer operations team, they should be able to resolve your concerns.

I would highly doubt that they dropped everyone's payment information. Think about how much money would be lost from people not willing to overcome the inertia and re-entering their payment info.

It could be that management hopes for that -- due to publishers' inertia AdSense will delay payments them.

I wonder if that delay would improve GOOG Q3 revenue results.

Edit: it seems I guessed incorrect (due to GAAP).

You report revenue and expenses during the quarter in which they were earned, not when they're paid. So a delay in payment wouldn't change their Q3 filings at all.

Beyond this, Google makes a profit from having both sides continue to participate. The more ads they can show, the better. Hampering one side of this equation for temporary gains that only really improve how you look to Wall Street (not how you're actually performing) is silly, and seems completely counter to how Google typically works.

I agree that it does not make sense from long term business perspective.

But then what's left?

I know for sure that my own payment information is gone and it coincided with AdSense system upgrade.

Can anyone with AdSense account share if their payment information is still set?

But even if it's just my payment information lost - it's not a good sign about quality of engineering at AdSense.

That's not how GAAP works.

> ETA: Google would be awesome if it adopted the Valve culture where employees are trusted to allocate their own time, but I wouldn't hold my breath.

I don't see how that's supposed to work. From what I understand it's too big and clunky. A lot of dpts don't even know others exist, some groups are just a product bought off other companies and kept in tact.

The way I see it's too big and diverse to be able to have such a structure.

I had seen your earlier comments and asked Googlers how hard they thought it was to change projects. They said they hadn't ever encountered another company that made it easier to change project than Google does, and wanted to know if you had encountered such a company.

Just for some context. I started a couple months after this former Googler and our work backgrounds prior to Google are pretty similar.

In that time, I started out on an underappreciated starter project, did a hackathon for Hangout, filed some patents for that work, did 20% work for an yet to be announced project, and now building out a completely new yet to be announced product.

Google is a place that really allows you to find something that interests you and encourages you to do that. Internally, people are always looking for people to put in 20% work on their projects; so if you come across something that interests you, nobody is there to stop you.

Personally, if I was only to work on one project for my entire tenure, I'd be bored out of my mind and be looking elsewhere for opportunities. At Google, I can feed by project-ADD within.

One thing you learn at Google is that it really humbles you. You really are working with really intelligent people and you need to have some faith in your fellow coworker. Everyone has something to contribute and you can learn even more than you can teach/preach. Much of the issues with this former Googler is that he may have had some good/interesting ideas but he failed at garnering support. It doesn't help when he publicly on a global mailing list dismissed coworkers that have a significant amount of experience and real world accomplishments in the field that he was advocating.

Much of the issues with this former Googler is that he may have had some good/interesting ideas but he failed at garnering support.

I had widespread engineering support, actually. The problem was that (a) the people managing that product didn't understand the importance of quality in content, and (b) they were under obscene deadlines (which almost certainly wasn;'t their fault).

What really kicked me in the shins was when an executive, pissed off that people on his team agreed with what I was saying being the right way to go, threatened to damage my Perf ratings. My manager promised to protect me (he promised to "peg" me to 3.3-3.4, which is decent) and didn't deliver. After that, I started looking for transfer opportunities, because I'd caught my manager in a few lies but that was a big one, and I just couldn't trust my manager after that.

It turned out to be easier to get a job outside of the company, so I did.

I had widespread engineering support, actually

From what I saw on the engineering-wide mailing list, where you told people who had many more years of engineering experience than you that they were full of sh*t, I'm sorry to say, no sir, you didn't have widespread engineering support.

We're talking about two different times. The RGI had a lot of support, and everyone who knew about it knew that it was the right way to go. That was in late July/early August before management started cracking down on internal dissent (the first Real Names PIPs were getting served around late August/early September).

Eng-misc went to hell about a month after that. By that point, it was obvious that RGI was going to fail due to executive pressure and that higher-ups were cracking down on internal dissent, so only the loyalists were even speaking up.

So what are you doing nowadays, anyway?

I stopped telling people where I work after Google. (Someone from Google tried to fuck up my next job, and Google refused to investigate. That's the main reason I hate Google's management. If they'd done their professional duty and investigated and fired the rat bastard, I'd have some respect for them.) But I'm doing well. Working on a side project to release this card game (Ambition) into the wild. The game rules are here: https://docs.google.com/document/d/1S7lsZKzHuuhoTb2Wj_L3zrhH...

Most companies make it hard to transfer by setting headcount policies that make it really hard to move internally, so people have to change companies to get a better job. Google isn't as bad with the headcount issue but they have this mean-spirited and possibly illegal thing where your "perf" numbers are visible in the transfer process and the result is that it's hard to get a good project unless your political success is top 25% or so.

"most of the work is shitty legacy maintenance"

LOL no. Stop projecting.

Reply on news.ycombinator.com