I have to say that after being an app developer on Facebook... | Hacker News Reader

I have to say that after being an app developer on Facebook...

mattmaroon.com

32 pointstwampss10 comments

Is there a way they could test improvements without pushing them on 150 million users? I have to imagine they already have some internal testing process.

What's a good practice?

Also, Matt, love the new site theme. I think I've been convinced to finally give up white on black (Except for my terminal. It'll be a cold day in hell before those design nazis convince me to change the colors on my terminal).

For partial roll-outs, A/B testing for features and more, check out http://startuplessonslearned.blogspot.com/

Well, the problem isn't the scope. The test would have just told them that a lot of people would complain they didn't like it, which they probably already knew. A lot of people will do that with a good change or a bad one. What they need to find out is if it will make their site into something bigger in the long run, the way the mini feed and such did. I don't really see how you can test that, because you're looking for something that is at best hard to quantify and at worst impossible.

Leave it to Matt to always go against the grain of conventional wisdom.

I definitely agree that Facebook has balls pulling out this redesign but I do not agree that the design is for the better.

My thoughts on the new facebook homepage: http://blog.clutterme.com/2009/03/thoughts-on-new-facebook-h...

    * It's going to stink for apps.
    * Facebook is killing its own page views

On the converse side, the redesign is a definite upper for brands and the visibility of their pages:

Mashable talks about some of the effects that the redesign has had on its Facebook Page, http://mashable.com/2009/03/19/new-facebook-brands/

    * It's going to stink for apps.

Apps stunk for the user experience anyway. They were an experiment gone horribly wrong.

Would you consider a pageview decrease a downside? Even though they may lose some ad revenue (I don't know the details), their bandwidth costs may dip a bit, and ads aren't their real path to monetization anyway. I'm not sure it really matters as much to their long-term vision as apps.

Would it be impossible for webapps to just have different interface versions always available? Like if I really liked something about the very first release, I should be able to roll back to that version and use that. We should all be able to use the interface that we like the most and upgrade as we see fit.

It's possible but you have to jump through quite a few hoops to do it. The web isn't built for this unfortunately...

I'm sorry, but does that page say anything ? All I saw was solid black with some dark gray smudges ...

It's about design, you say ? Curious ...

(Thanks for changing the theme, Matt. I'll send you the bill for the points I got docked here for the snark ;-)

Well, now that I read it, I might as well comment: fine, keeping still kills you. But the Google-leaving Douglas Bowman hit the nail: you can either hope a designer does an Apple-class job, or you take the stolid measure-twice, cut-once engineering approach. And it's a matter of company DNA.

Plus, what if the designer lands you with dark-grey-on-black ?

Reply on news.ycombinator.com