Do not panic. We accept late submissions for TechCrunch50, but please submit soon. »
Oh No He Didn’t
by Michael Arrington on May 30, 2008

Blaine Cook, the former Chief Architect of Twitter, takes a shot at his former company today complaining, of all things, Twitter downtime.

The only trouble is, the feature he’s complaining about hasn’t failed, it’s been taken down by Twitter along with other measures to reduce overall stress on the platform. The same platform that Blaine built and that is occasionally seen live and functioning on the Internet, that is.

I’m all for trashing Twitter when they go down (it’s turning into something of a hobby for me), but its just plain weird for Cook to be doing it. Even more so when he gets the facts wrong.

The official on the record story between Cook and Twitter is that he left on good terms, everything was amicable, needed to move to the UK for his partner’s career move, etc. The unofficial off the record story is that he was shown the door so that Twitter could get down to building something that could scale. Many more comments like this, and Twitter may fire back.

On a side note, Twitter’s doing an excellent job of actually communicating with users all of a sudden. Good for them, it was much needed.

Comments rss icon

  • Google Talk to FriendFeed is freaking fast {seesmic_video:{”url_thumbnail”:{”value”:”http://t.seesmic.com/thumbnail/V3MS62BQCf_th1.jpg”}”title”:{”value”:”Google Talk to FriendFeed is freaking fast ”}”videoUri”:{”value”:”http://www.seesmic.com/video/TL5nVqae98″}}}

  • Oh JFC Scoble, will you please shut up about friendfeed? or at least do it somewhere else. :-)

  • Tomorrow FriendFeed will be on the Gillmor Gang. Why can status messages go from Google Talk to FriendFeed instantly, but they haven’t yet gone the rest of the way to make it possible to do what Twitter used to do? That’s my question for the founders of FriendFeed. See ya at 10:45 a.m. It’ll be live streamed at http://www.ustream.tv/channel/techaura

  • Michael: I was pitching Gillmor Gang, a show on your own network! :-)

  • Heheh. If you think a video show is gonna hurt TechCrunch 50, then you probably think the puppets are hurting me. I know the truth: that we’re both laughing all the way to the bank.

    Nice to know that you aren’t like Gillmor, though, and that you are watching FriendFeed. :-)

  • I find it kind of annoying that theres a blog post for each time you find twitter down.

    Leave silicon valley one day and ask anyone what twitter is.

  • RIAA: I have one answer to that: http://www.twittervision.com — most Twitter messages are from outside of Silicon Valley.

  • Robert - don’t talk to my trolls.

    they’re trolls, but they’re my trolls.

  • oh man, I hate to see twitter down. Oh well gotta keep in touch with the chaps on Pownce then.

  • I always find it kind of funny when you post about it. This time even more so.

  • Well, I think he may give comments like this. At least this guy knows much more about twitter architecture than all TechCrunch writers put together)

  • I still can’t understand how come twitter are having such huge problems to scale. They don’t have that many users (numbers that were thrown around were around 1.2 million and just 200K active).
    I can understand difficulties on the first days, but they have been here for a year now.
    In the end, twitter is similar to mail. It’s a queue system. And they also don’t need to send huge and heavy HTML but just small messages.
    Put a tier of web servers on a cluster, with a good queue system like MSMQ behind, and you can scale it as much as you want. Maybe you’ll get some perf problems here and there, but definitely not shut downs…
    Maybe the joke about going down just in order to make scoble and the rest keep talking about them is true… :) It definitely work. I never seen one app stay in the news for so much time (half the message on twitter seems to be on twitter :))

  • Let’s all switch to Jaiku, I have a few invites left if anybody is interested. :) Contact me via my website.

  • Pavlo, yeah. I’d hope so. Too bad all that knowledge didn’t equal a Twitter that could stay live.

  • Arrington:

    Are you seriously running out of good material to write about? You make a post out of one silly twitter port? Are you Valleywag?

    I hope you change your ways before your blog takes a turn for the worse. Maybe Wired was right about TechCrunch.

  • @Shahar, sure it’s just like email, it’s just a queue system. But how about you write me a email/queue system whereby Scoble can post a message that is expected to be viewed instantly by 37k out of 1.2m users, while at the same time 200k other people are posting similar messages, and another 50k or so are real-time searching via Track in IM.

  • can someone create twittercrunch to discuss this topic, save the mind-numbingly boring posts. That goes for you too scoble.

  • Hmm, presence is hard to scale, but would it this much?

    I doubt so: I think google talk in gmail - which architecturally is the same technology - wouldn’t work simply.

    So, Blaine has right: IM may be not the base of the problem of twitter. It’s only there’s so few of us who can speak XMPP (and even less of us who understands the deeps of erlang, the mystic sanscrit-like programming language of ejabberd:)

  • Doug: if Twitter posts are so boring, why do they generate so much traffic for Arrington? Boring=the new exciting.

  • @Shahar - if it was that simple just like putting some clustered webservers and running a microblogging service, than you would see 50 new twitter equivalent companies tomorrow morning.

    I think that twitter architecture and its scaling were analyzed to the last bit and byte in so many blogs and forums. Twitter idea might sound simple at a first glance, but it has a very challenging performance and scalability issues.

    Again, with so many people are excited about such service, you can be sure that if it was so simple, than you would see tons of services like that immediately.

    But you don’t.
    :-)

  • just focusing on the speed;
    Twitter has 1205 Alexa, and Friendfeed has 40,707 Alexa rank.
    I don’t think you could compare both with such different loads on them. Let’s wait and see, how FF is doing when they have a 1000 Alexa rank.
    btw I am sure they will do much better than twitter. However, this is the rıle of nature. The later species are always far more superior than the former species.

  • Robert, everywhere I look (my twitter feed, RSS etc) you are there and now infiltrating the comments with FriendFeed! haha

  • Tom: I’m a social media fungus.

  • guys, take my word: 2008 will be remembered for “scalability”.

    and yep, no matter which service “killed” or piggy backed another service, web 2.0 services and new social graph challenges are making the best architects to wake up and move from the armchair to the drawing board…..

  • Robert: they generate traffic because they’re negative. Even Michael noticed once that whenever he trashes someone he gets loads of comments and trackback. And whenever he writes positive/neutral posts - silence.

  • Uh, what facts did he get wrong? He just said that the IM interface isn’t up, and it’s not difficult to get up.

    I think you’re reading into what he said, and obviously he wasn’t clear, but it’s hard to be clear in 140 chars.

    For example, it might be that he is inferring that it’s not difficult to get enough bandwidth that it’s feasible that they want to put it back up.

    It all depends on what emphasis that you put on the word “restore”, ie is it that the service has suffered enough, or broken, or is it a matter of “restore” meaning committing enough resources/time to make it worthwhile bringing back.

    You could maybe ask what he’s getting at before posting a negative story about something he’s said.

  • Michael,

    Scoop time, why does twitter up and down time show that most long time outages occur during middle of month?

    We are not being told the transparent full story form twitter..

    Now, I will not say twitter staff is fibbing but one has to wonder as that above stat is somewhat strange given their explanations.

  • The thing is, Blaine is probably right. Unlike most other features Twitter has turned off, IM doesn’t dig deep into the database, it just does provides another interface to services that are still up and running.

    It looks like Twitter is currently operating in ’safe mode’, leaving only the most used features. And it’s getting annoying.

  • On the record vs off the record. Who cares? What matters (if anything about this stuff matters at all) is the truth. Off the record stories about things like this aren’t any more or less likely to be true than the on the record ones. Often, the people involved in situations like this are so deluded and/or dumb that they don’t even know what the truth is. If the off-the-record story is that Twitter’s problems are due to this one guy getting in the way, then I smell BS and an attempt to divert blame.

    Here’re a few *facts*. Twitter doesn’t scale. The problems have been known about by everyone inside (and outside) Twitter for a long time. The people that are accountable for this are the people the own and run the business. End of story.

    For sake of argument, though, let’s just say the technical problems are all down to Blaine Cook, and that he was getting in the way of implementing a solution. Then the question is: why didn’t they fire him a long time ago?

  • How did he get facts wrong?

    He said it’s a difficult service to restore. IM (jabber specifically) was one of the reasons twitter went down in the first place after the database crashed, if I recall.

  • As a software engineer, I am embarrassed every time twitter is down. It’s hard to scale software, but it’s not that hard!

  • @32: “How did he get facts wrong?

    He said it’s a difficult service to restore. ”

    Uh, no, in fact he said the opposite: “because it’s not a difficult service to restore”

    What does it really do? I haven’t used the IM, but doesn’t it just post statuses? In which case, isn’t that as intensive as any other post mechanism, and therefore, why the heck can’t it be on?

    Unless it needs a third system to pull updates and then push them into twitter, in which case, okay.

    But yeah, sure, if we want to blame all the scaling issues on Blaine, then how come they haven’t made great strides in making it work since he’s been gone? Surely it’s their prime goal, and the reason they fired him, so why isn’t it getting better?

    I haven’t heard Blaine say much negative about the company, he has made comments skirting the issue, sure… but he’s always sounded like he wants them to succeed and is a bit embarressed he couldn’t do more to help scale, saying it’s more a rewrite to have a system that scales better - so I’m not going to take one comment as a negative snipe, when he’s saying less bad things than Mike does every time twitter has a hissy fit.

  • Maybe he meant to stay supportive but was inspired by Scott McClellan.

  • Anyone else find it ironic that he’s bitching about twitter…..using twitter??

  • I’m a little teapot

  • I’m not a little teapot. I’m a post-it note with the word ‘peanut’ written on it that is then stuck to a frisbee that is then flung over a rainbow.

  • I don’t see what’s so weird, he’s trying to be funny. What’s weird is TC turning the whole thing into a soap opera, Arrington are you better than Valleywag or not?

  • Ohhh I love the Twitter drama. I was planning on watching Sex and the City tonight but I think I’ll just stay here and wait for updates on Techcrunch ;)

  • Preferring Duncan Riley’s take on Blaine: http://www.inquisitr.com/twitt.....-on-uptime

  • Twitter has been receiving free advice from technical people around the blogsphere. Any average B. Sc. with technology experience can solve simple problems like this.

    A good post about this topic:
    http://arseneault.ca/2008/05/2.....sphere-mq/

  • What Cook said makes perfect sense, when you take something offline you eventually have to bring it back. In a realtime system you never know what will happen when you do that.

    Have you ever seen the movie Amadeus — you know the scene where Mozart demos his work for the King of Austria and the king says it’s pretty good just one problem — too many notes.

    That’s what reading TechCrunch is like these days as you venture into the land of computer system architecture.

  • As for the stupid “no one outside the Valley uses Twitter,” I live in rural Iowa and find Twitter essential for communicating to collaborators and customers.

  • Ken Sheppardson - May 30th, 2008 at 6:54 am PDT

    Wait a second… so Arrington is Mozart, TechCrunch is “Die Entführung aus dem Serail”, and you’re Emperor Joseph II, Dave?

    Anway… TechCrunch has enough readers that it’s a little disappointing to see a off-hand comment like “Frustrated, because it’s not a difficult service to restore” get misinterpreted and spun like this.

  • It’s an odd thing to come out of the mouth of someone who is responsible for causing the problem…

    AO

  • ha ha… ha Ha.. ha HA HA.

    In one word: sucks.

    In two: sucks, period.

  • Ken, Arrington is the king.

    Obviously! :-)

    Calacanis is Salieri.

    Scoble is the dwarf jester.

    Gillmor is the mad scientist.

    I don’t know who Mozart is, probably Evan Williams.

  • Michael,

    If Twitter is so bad stop using it and spare us all the bs about the service. It’s a ridiculous, narcissitic product anyway. We know you’re addicted but, for god’s sake, spare us.

Leave Comment

Commenting Options

Enter your personal information to the left, or sign in with your Facebook account by clicking the button below.

Alternatively, you can create an avatar that will appear whenever you leave a comment on a Gravatar-enabled blog.

Trackback URL
bugbugbug
The CrunchBoard
  • MediaTemple Logo
  • QuickSprout Logo
  • OpenX Logo
  • Cotendo Logo