An open letter to Sno

posted 13 May 2005

Sno is the creator of OUTeverywhere, a community website of which I am a heavy user. It has recently undergone a disastrous interface overhaul.

I'm a web developer, and a long-standing member of OUT. I say this so to make it clear that I at least have some idea what I'm talking about, both on the notion of OUT as a community and OUT as a piece of software.

The old OUT had problems, both social and technical. Frames aren't very accessible to the blind. To standards-compliance anal-retentives, it was a nightmare: font tags and frames are both out of fashion on the web these days. Socially, the board were full of whinging fuckwits. I will address all of this issues.

Firstly: FUCK standards-compliance, and as a web developer, I do not say this lightly! Cutting out useful features just so that you can boast that your site is pure XHTML/CSS is exactly the opposite of what standards are, at heart, supposed to achieve: greater usability for all. Instead, you're getting reduced usability for all -- the lowest common denominator. What the old OUT interface had, in spades, was *usability*. That is, after all, why so many people used it.

Frames, much maligned as they are, were *perfect* for OUT's purposes: they kept navigational elements on screen at all time, giving the site a very anchored feel, more like using a specialized "OUT browser" than reading a web page.

Saying that you can't do this and that because browsers won't let you is a cop out. *CSS* won't let you do these things. *Browsers* have let you do this since 1995, which is when you designed the original interface and built all the features we are now crying out for. Just because you can't do it the new, cool way doesn't mean you should stop doing it.

The old interface was "cluttered". But that's another way of saying "information rich". Nobody accuses BBC news front page of being "cluttered", because it has the difficult job of fitting an awful lot of news into a very small space. OUT fits an awful lot of stuff -- boards, polls, profile visits, contacts, incoming messages, and full navigation -- into a small space too. The new interface has hundreds of pixels of white space wasted at the top and sides of every page, making reading and navigating a million times clunkier than the former interface.

A split forum interface that showed topics on one side and threads in the other -- like a newsreader -- was *innovative*. Just because nobody else seems to have done it since doesn't mean it was a bad idea.

A little bar that refreshed status widgets? Genius! Pervasive by not intrusive status notification, years ahead of its time. The rest of the web still hasn't caught up to that. And now it's gone.

Frames are also very bandwidth-efficient. Even on a slow connection, reading messages was speedy and browsing multiple profiles was easy. the new interface now wastes bandwidth in addition to space by re-sending the navigational elements on every single page. Another step backwards.

Socially, the boards feel slighted. I don't that was an intentional move on your part. You are aware, as is everyone else, that while only 1% of the site users are active board members, they are also the same 100% who organize the events, respond to support threads, and in general provide the sense of community that your market research told you the members of OUT valued so much.

Sno, the old OUT interface was a wonderful achievement in software engineering, full of innovation and thoughtful design, the result of years of tweaking. The new interface, even if you spent 6 months designing it, could never match the value of years of evolutionary development. If it was a nightmare behind the scenes, then reengineering was needed. But even if you rebuild the back end from the ground up, you need to KEEP THE INTERFACE, because it WASN'T BROKEN. It need evolution, not revolution.

I love OUT. I use it all the time. A disturbingly high percentage of all my friends are either from OUT in the first place or are on OUT now. I don't want to leave OUT, and I *really* don't want to have to clone it. But if I have to, I will, because this new interface is just wrong.

Please, bring the old interface back and try again.

9 comments