Steve Yegge on Google+

This is an awesome rant about what's wrong with Google, with a hearty side-order of what's wrong with Amazon.

Stevey's Google Platforms Rant

Google+ is a knee-jerk reaction, a study in short-term thinking, predicated on the incorrect notion that Facebook is successful because they built a great product. But that's not why they are successful. Facebook is successful because they built an entire constellation of products by allowing other people to do the work. So Facebook is different for everyone. Some people spend all their time on Mafia Wars. Some spend all their time on Farmville. There are hundreds or maybe thousands of different high-quality time sinks available, so there's something there for everyone.

Our Google+ team took a look at the aftermarket and said: "Gosh, it looks like we need some games. Let's go contract someone to, um, write some games for us." Do you begin to see how incredibly wrong that thinking is now? The problem is that we are trying to predict what people want and deliver it for them.

Previously, previously, previously, previously, previously, previously.

Tags: , , ,

17 Responses:

  1. Frederick Roeber says:

    Larry was very strongly anti-"platform." Having that word in your product plan was a sure way of getting nixed. Some of us tried to sneak in the idea anyway, but I remember trying to find codewords that would communicate the message to other engineers while keeping it away from the authorities.

  2. It's kinda hilarious that he spends a good paragraph complaining bitterly about the arbitrary, mis-incentivized and unprofessional amazon recruiting/hiring process, which when he finally gets around to describing it sounds largely identical to Google's. (Press-gang interviews? Candidates routinely slotted into the wrong silo? Zero accountability and in fact general cheering on of obstructiveness? Check, check and check, right on down to the use of "B players hire C players" as a catch-all justification for any and all misbehavior.)

  3. Related - I came across this
    It kinda blows my mind that Eric Schmidt uses a blackberry and doesn't have a google plus account

    • Jon says:

      Isn't Schmidt also on the apple board? The bb makes sense to me, totally impartial.

      • MasoudS says:

        He was. He had to leave after android.

      • Art Delano says:

        No, he left Apple's board of directors a couple years ago on grounds of conflict of interest. Also, he doesn't run Google any more. He's got an advisory position at Google and a nice-sounding title. Carrying a Blackberry doesn't help advertise Google's ability to eat its own dogfood but that's about the extent of it.

  4. Ronan Waide says:

    Oddly enough I recently listened to a bunch of Google Developer Podcasts from about 2009 or so (attempting to increase my knowledge of tech I don't know about while using gym equipment) and a huge portion of that is about APIs (for GDATA, Google Maps, Google Calendar, and so on) and how to use them. I wonder what those guys are up to now?

  5. Jeremy Wilson says:

    This whole rant is brilliant. He should get a promotion. I've been directing my company in the direction of platform from day one and it makes a huge difference - we'll be ready for public-facing APIs right away, while our competitors scramble.

  6. Phil says:

    Another "Previously" might be:

  7. gryazi says:

    I think the one thing that might have kept Google from becoming extraordinarily creepy (as in, "fuck this I am definitely packing up as much personal data as I haven't already surrendered and going to play anywhere else") instead of just creepy is that they don't do "platforms."

    So, like Apple has actually worked in practice [despite "fruit-themed-toymaker"'s occasional efforts to transcend it all and get enterprisey], you get A Bunch Of Shiny Baubles On Offer To The World, Some Of Which You May Find Useful, but they've usually managed to skew just shy of creating an """ecosystem""" where you have to buy into Everything As A Way Of Life to participate. Even while releasing a bunch of things that individually possibly deserve to be/can optionally be treated as such; I think the operative word there is 'buy' as in 'immediately offer something in exchange for' - Goog mostly throws open-source at the wall and seems happy enough if it keeps the employees happy (hey, look, 'altruism'!) and anything whatsoever happens to trickle back (free work!).

    If there was a single Google Way Of Doing Things, that would instantly catapult them up to at least Microsoft's level of "dominance," but with possible IBM-of-yore discipline/intelligence and ownership of the Internet (which they already have, but the internal balkanization creates the impression that they're too disorganized to be evil all-at-once).

    So... Defect or sly marketing/business/regulatory strategy? And isn't having a bunch of different toys instead of just the 'OS' more 'fun' (despite the ongoing reiteration of UNIX)?

    I imagine Goog really doesn't want to be in a position where they ever have to consider bailing out MS to maintain the illusion of competition, if you know what I mean. As long as they stick to providing random 'products', and 'services' to help discover their and others' products, they're playing in different sandboxes and the unlikely possibility is completely negated.

    • gryazi says:

      Eh, talking to myself, but - er, the one and only Platform that needs to matter to Goog as an ongoing business is the ability to reliably squirt ads into things, unless they lose track of that as their actual business. Or start thinking things like "Hey, we can devote our idle infrastructure to 'grid computing'^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H^H 'cloud services'!" instead of devoting all cycles to search.

      • gryazi says:

        P.P.S.: If you can access useful data out of a Google service without going through a [human-to-computer] interface Google has carefully designed to optimally connect eyeballs to ads, Google has lost. Why does this Steve guy hate Google?

        • LionsPhil says:

          If it were as simple as that, why does GMail offer mostly-working IMAP (or POP3) and SMTP?

          (No, they don't stitch ads on the bottom of the messages or anything. Not yet, anyway, and they've had years to try.)

          • gryazi says:

            In all seriousness, I expect that resides somewhere between a courting gift and someone doing a spreadsheet and finding out that it was still cost-effective as long as anyone who would otherwise use a real mail host logs into the web client even once / they get to screen your keywords and tune the ads associated with your Google cookie etc.

            I keep forgetting and still can't wrap my head around the fact that even Yahoo! gives away IMAP now. Because that wasn't available to actual paying SBCT&T customers when the 'migration' happened ten years ago, and I only found out when I got a replicantphone.

    • LionsPhil says:

      Well, they're sure ramping up the creepy (and insecure) awfulness with federating everything through a single Google sign-on.

      The horrors are not lost on me (althought apparently don't bother me quite enough to not do it) that if I want to allow a game---say, Team Fortress 2, which made quite a big thing about this capability---to YouTube, I am passing it the same credentials needed to get into my e-mail account. I can imagine that Google really quite love that people who use the web interface to GMail are therefore also providing logged-in usage data for search and YouTube viewing.

  8. Alex says:

    My personal theory about this is that out of the three most rated Google APIs - Maps (and associated KML), Wave, Android/Dalvik - two came from the same gang, the Google Aussie engineering group which originated as Keyhole, the company they bought for the mapping tech. The other, Android, was an acquisition as well, come to think of it.

    Conclusion - it's not a core Google capability and the institutional knowledge is concentrated in a couple of peripheral operations.

  9. gryazi says:

    Oh crap, I'm never going to shut up in this thread, but someone just came up to me at work and brought up Raskin while I was trying not to go there, so... maybe I'm getting my HCI pantheon confused, or maybe they both said that there's a value to non-customizability.

    This applies both to 'the only way to get data in and out of Google should be through a Google-designed interface that shows you ads if Google is an advertising business and wants to succeed as such' and, in one of those delightfully insane ways, the misnonfeature Yegge bemoans in Chrome: In some ways, it's better to make having to thwack CTRL-+ a requirement consistently across all instances of the product (and in a browser you can even argue that it makes a perverse sort of sense to push the problem back on page authors' choice of stylesheets, as long as the browser is doing its now-HTML-is-back-to-being-'layout-centric'-for-a-while job and rendering text in the size the code specifies)...

    Now, in practice, there's some difference between let's-use-the-good-old-automotive-analogy cars having their pedals and steering controls in roughly familiar places versus only giving a choice between shooting forward at full speed or backing up to the left (weren't we just talking about GNOME 3?), but... if a 'product' only comes in one shape, your familiarity is applicable to all instances of the product, not just the special customized never-leave-the-house version you've set up.

    Now, in the particular case of browsers and websites - if the browser is mostly just sitting there to render the interfaces for various services you log in and out of, then doesn't it actually make a sick sort of sense that your font prefs should live associated with your login credentials 'in the cloud' rather than just the one local instance? Except for the part where you can't read the login page's text, but then maybe the damn thing should be inheriting text scaling preferences from how blind you've told your OS you are rather than perpetuating the theory that each application should take on the burden of making you figure out how to tell it that yourself.

    And then if it still sucks you can .. consider whether an entirely different product should exist to meet the needs of your demographic, or if we should just swing the fad pendulum back to semantic markup next year.