The Highlight clips from the Gates/Jobs mutual interview is worth watching, even if you’re not a tech. Humorous, but also puts a real historical feel behind these many decades now of the powerhouses behind personal computing. The full interview and a transcript is also available from the page.
Recovered from the Wayback Machine.
Question: How do you turn a Web 1.0 application into a Web 2.0 one?
Answer: Pull the plug.
I’ve been involved in comments over at Robert Scoble’s on the Zooomr crash and burn, and make no mistake: the application is down and out, and there is no plan in place to get it back up other than asking for donations. Donations for what? Well, we heard about server problems, but then we heard about RAID controller problems; discussion about the database server failing was followed by discussion that 10TB is not a big enough server. This was then followed by the problem isn’t CPU: It’s I/O because of the static pictures being served. Kristopher Tate doesn’t seem to be a consistent idea of what the problem is. Other than, We think the controller on our main database server is bad.
This is what I dislike about Web 2.0. Kristopher Tate and Thomas Hawk are darlings of the A List set, so therefore these aren’t two men running an application, neither of whom really has a clue in how to keep an application of this magnitude up and going. No, these are two brilliant, far thinking futurists who only need money and help from someone to keep this going, as the troops rally around with “Way to go, guys!”
This application has been down over a week, after it went down once before with a promised rollout, after missing its initial rollout at its own startup party, following on what sounds like other downtime problems. Do you think system users should be concerned? Perhaps even, dare I say it, critical? Not on your life. Being critical is not the Web 2.0 way.
There isn’t a note on the front page about the server being down for the count. No, they’ve pasted some UStream videos where people can watch Hawk and Tate stare at a computer terminal, drink wine with ice, and talk to people who are asking them questions via some form of chat. However, when the server failed this last time, Thomas Hawk did write about the the little train that could.
(I would pay money, real money, for one of you developers at, oh, Citibank, Boeing, John Hancock insurance, or so on to go into your bosses office next time you have problems with your applications, and tell him or her about the little train that could.)
I have had more than my share of problems with hardware in the past, and normally I would be much more sympathetic. In fact, since the application is database-driven and PHP, I believe, I might even have offered help–I’m better at troubleshooting problems than building apps from scratch. Not for an application and a company, though, that keeps billing itself as the place for photo sharing, with such grandiose pronouncements that one is forced to imagine that Flickr is the po’dunk site, while Zooomr is the tasty noodle waiting to be snapped up.
So now people are being asked to give, Kristopher has bills to pay, and Zooomr needs a new server. Personally, before I started throwing money at the site, I’d ask to see a business AND technical plan upfront, including a detailed estimate and listing of hardware Tate and Hawk need, now and for the next six months to a year. But, as Thomas Hawk has pointed out, I’m not really a part of that whole Flickr/Zooomr scene, so I don’t really understand.
True, I don’t. I don’t understand solving server problems by doubling or more the burden on the server; throwing new features on when the server can’t reliably support existing ones; and not putting an honest, blunt, note in the front of the site telling people exactly what’s happened, and how the site, and their photos and data, will be recovered. However, before other startups think to follow this as an example, all I have to say is: don’t try this at home, kiddies. Not unless you’re good buds with the Big Names.
I’m sure the company will get the support they need. After all, they have all sorts of friends among the movers and shakers. That’s the only thing that really matters in a Web 2.0 universe.
Speaking of movers and shakers, I have an idea for the company: Get Mike Arrington and Techcrunch to fund the equipment they need. After all, Techcrunch has been pushing the company as a Flickr buster since the very beginning. As much as Scoble and Arrington have been touting this site, you’d think they’d both be fighting each other for the honor of investing.
Recovered from the Wayback Machine.
Like the rest of the world, I’ve been exploring the tutorial and examples for the new Google Gears. I was particularly caught by the addition of SQLite for offline storage–I never imagined a day installing a relational database on your client’s machine via the browser on the fly.
It’s remarkably easy to get up and running, and the API is quite simple, too. I copied the database example; now I just need to figure out what to do with it.
I also tried the download static pages example, mainly to check out how the data is stored locally. Interesting storage structure. Could get local disk space intensive if not used wisely.
Should be loads of fun playing in the future.
correction: SQLite is a component of Firefox. I’m assuming it’s installed by Google Gears for IE 6 and up. Support is only for Firefox 1.5 and up, and IE 6 and up. Eventually support will be provided for Safari. No support planned for Opera or other browsers.