Many Twitter fans have become very much used to following big events via Twitter, its immediate nature and 140 character limit provides key points of what exactly is going on.
Unfortunately for Twitter users, Twitter completely failed during Steve Job’s keynote address at the Macworld Expo in San Francisco.
From the floor I thought it may have just been my internet connection, but for the brief time Twitter has been up since the keynote (it’s returning errors as I write this) the Twitterosphere is not very happy, particularly those who stayed up late (it was 4am on the East Coast of Australia when it started, and I know plenty of people who were up for it).
Mike Bogle sums it up well in a post he wrote during the keynote:
In that sense Twitter has really begun to play a central role in significant events these days - not just for information consumers, but for deliverers as well. It provides an opportunity for staff working on major events to follow consumer/public opinion as it’s being formulated in the minds of the public. As such it provides some amazing insight.
However as of right now however, Twitter is unresponsive and no - or minimal - tweets are getting through. This has effectively crippled one of the key real-time modes of information dissemination and discussion and left countless users in the lurch.
The fact that Twitter continues to crash during times like this arises from nothing short of poor planning on the part of the company. Load and capacity planning is a fundamental consideration for providers of enterprise level services and the fact Twitter is not doing that is not only cause for a tremendous amount of frustration from end-users, but also concern about the staying power and future of the company.
For those that did miss the event, you can see my live notes from the event here.
Twitter will probably just shrug their shoulders as they usually do during outages, and ultimately we’ll all keep using it, but if there’s one messge I’d like to get through to Twitter: keep messing with Twitter users and eventually users might flock elsewhere; sure there’s no reasonable alternative now (Pownce is on deadpool watch, Jaiku was killed by Google) but this won’t always be the case.
Arrogance kills. Or starting to.
Twitter should recognize that a lot of people are using their service now far more than what they probably envisioned it to be. I was all set to get some instant feedback from attendees that I know… to no avail. So, what happened Twitter?
still down from what I can tell…and the keynote ended over an hour ago….
I just posted the same thing:
http://www.centernetworks.com/.....bs-keynote
just imagine sxsw!
btw - why is pownce on deadpool watch duncan? do you need to fill a quota this week?
maybe twitter would be more reliable if each tweet didn’t create a new url…seems pretty low tech if you ask me
- Bad publicity for Ruby on Rails. Again.
- Doubt that Pownce is headed to deadpool.
I thought it was just me
it’s free. “i’ll take my non-paying service somewhere else….” doesn’t really make sense in this case.
In the grand scheme of things, does it really matter that you missed your dumb twitter post?
Twitter is a Rails app? Ruby is known to be slower than, say, Java (benchmarks are available), though the upcoming release will be faster than the old one. I wonder if Twitterfolk ever bitch about their language of choice? I’m a BaseCamp user, and it suffers from the same problem and I *pay* for that *business* tool.
But Twitter will work out.
What i find interesting is that no one has figured out how to launch a ‘twitter bomb’ denial of service attack yet. They’re has got to be at least on disaffected 14 year old hacker out there who is pissed because he only has 1 follower (his WoW buddy) and decides to take down the system.
…though…based on twitter’s performance today it looks like that’s unnecessary.
Peter
http://www.twitter.com/corbett3000
I ran my own little experiment, and pretty much everybody I tried (engadget, gizmodo, stevekeynotelive, the apple iphone update server, plus more - you know who you are!) keeled over.
More details: http://www.appistry.com/blogs/.....way-to-go/
Perhaps there’s no alternatives to twitter and jaiku (inasmuch as people are forced to use WinXP/Vista and Google). But that’s only the situation in the U.S.
And with the twitter’s owners/staffs don’t-give-a-shit attitude, good luck (which you’ll have, so, just continue the downtime-as-usual).
MacWorld Keynote killed it?
Imagine what American Idol would do….
Twitter isn’t even Digerati ready, let alone mainstream ready.
It worked for me. I twittered through the whole thing.
Twitters inability to keep their service up is an insult to its users.
-Scott
http://simpable.com/technology.....-to-scale/
I was really frustrated, Instead of enjoying the keynote on twitter, I spent my time ranting http://technozzle.com/?p=40
Rails don’t Scale.
This need updates — http://www.highscalability.com.....ent-faster
Interesting that Twitter held up great during the So Cal fires, but fails during the Macworld keynote.
Pownce is launching publicly next week (no more invites!) with lots of new features. This morning we added an icon to our mobile site for iPhone/iPod Touch.
http://blog.pownce.com/2008/01.....he-iphone/
Pownce headed towards the deadpool? I think not. Headed towards Madrone Lounge for our pre-launch party tomorrow night is more like it. Everyone welcome (even you, Duncan)!
More info about the party:
http://pownce.com/leahculver/notes/1146315/
Speaking of outages….
What do you expect for free?
I’d wager the bottleneck is their architecture, not the language. On a social system like Twitter, it’s far easier to distribute the application layer than the database layer. Their database is probably melting.
-Phil
This was incredibly frustrating for all us Twitterers today. Many of us are starting to get really annoyed by the constant outages — Yes, it’s a free service, but if it’s unreliable, we *are* going to find a better alternative.
I’d still like to see Twitter have a better foundation to accommodate the amount of users it has now, as well as in the future. Heck, give us sidebar banner ads to make it viable. Just please, upgrade your software and systems.
I think I will start a twitter clone calling it fritter, anyone want to join?
I’ve seen talk about using twitter for during times of disaster for communication and support.
Guess this rules that out!
Yeesh - give Twitter a break. It’s a new app and it has some early shortcomings. You cannot tell me that people depend on Twitter for information. They LIKE Twitter for information, but blogs, text messages, email, and standard websites can all convey information in nearly real-time. If the people up at 4 AM in Australia couldn’t get their Tweet fix, I’m sure they just moved over to one of the countless other blogs, etc. that were reporting on Macworld at the same time. And, even better, they got pictures with their info.
Just as with the product manager leaving Google a week or so ago, you guys just like to blow things out of proportion to drive clicks.
Why not explain what happened in your own backyard (*cough*CrunchGear*cough*) before you start trash-talking other sites, mkay? It looks really bad, y’know
That is what happens when you use a weekend framework like Ruby on Rails on ´professional´application. It just can´t handle it.
Those of you choosing to blame this on Ruby or Rails are hilarious. I would assume you fancy yourselves web professionals. I would then assume that you could come to the simple conclusion that, most likely, the language nor the framework are the cause of these outages. I suppose I shouldn’t make assumptions.
As someone mentioned above, it’s much more likely that their array of database servers bit the dust as soon as everyone started trying to tell everyone else what Stevey was wearing today.
To #30, how come Wikipedia, Digg or Facebook, Flickr do not experience outages? Maybe because they’re not using overhyped languages? And of course database clustering is important, but seems doubtful that Twitter doesn’t have a some kind of load solution.
BTW, TechCrunch and Engadget run on PHP and didn’t crush even though the traffic was huge during the keynote.
I don’t think it’s fair bashing Twitter, their choices in technology aside, as the service is free. Though of course, this is bad PR for them if they want to get enterprise contracts.
I’m with #29.
wth is going on here?
Twitter was not the ONLY one who went down during macworld.
I mean’t MilesZS’s post. (#30)
and Dyde, 2 datacenters powering engadget did go down for a short time.
[ ref: http://feeds.engadget.com/~r/w.....217205075/ ]
To 34, true but they quickly recovered from it, unlike Twitter
@25: Here’s your marketing slogan for free. “Bitter with Twitter? Try Fritter!”
Looks like flickr didn’t do so hot either: http://valleywag.com/345381/flickrs-big-failure
I think Twitter rocks and that y’all are being too hard on it. It’s a GREAT service that is in its nascent stages and they are scaling it. Cut them some slack.
Powce? Deadpool.
Duncan, why do you say Jaiku was killed by Google? It’s still there and it works. Assuming Google doesn’t pull a Dodgeball, Jaiku could be a good Twitter replacement.
Just how many individual updates did people need from Macworld anyway?
I followed the keynote perfectly adequately only to find Twitterers repeating everything a thousand fold and then complaining that the site crashed. Jeez!
The fact that the Twitter outage created so much outrage is actually GREAT publicity for Twitter. It shows that even as a free service, its huge user base is heavily relying on it and is disrupted by its absence.
It might not have been on purpose, but it was a well-timed outage - diverting the attention of tons of technology reporters and enthusiasts away from Apple’s news, proving Twitter’s worth and impact to Union Square Ventures.
Notwithstanding, it’s certainly nice to note Twitter’s announcement of the Starling code today, and open sourced –
http://dev.twitter.com/2008/01.....rling.html
at least win back some (deserved) credibilities..
Duncan - This is yawningly late to this thread but I think it’s relevant. We implemented Twitter-backup solution, which hosts twitter posts when Twitter goes down and republishes when Twitter comes back.
http://blogs.messagedance.com/.....our-email/