[rburhum]

The problem of re-designing something as popular Wikipedia (especially something useful that has not changed in 10 years!) is that regardless of the result (i.e. better or worse), you will face _fierce_ opposition. Humans are very sensitive to anything that changes their mental model of how they think something should work.

Think about radically updating some of these to see what I mean:

- Facebook Newsfeed

- HNs homepage

- Vim interface

- Google Search Page

- Craigslist homepage

- Your favorite web mail client

- Your favorite Smartphone main screen

- Reddit homepage

You can come up with the most amazing forward-thinking improvement, and somebody without a doubt will send you so much hate that you'll think it is that person's only goal to wipe out your family line.

This doesn't mean we shouldn't do it, it just means that we have to be aware of that section of the population and come up with a strategy that either does 1) a slow migration for the people with old mental model, 2) provides an alternative "classic" view, or 3) completely ignores that section of the population since they may represent a very small minority. I tend to go for 1,2,3 in that order, but of course this can vary differently depending on the project.

Nice job with the re-design by the way :)

[bichiliad]

Something really interesting (and relevant) is that in cases where a redesign makes an interface markedly easier, it can still perform less favorably simply because people liked knowing how to do something that most people couldn't (i.e. using said interface). Take Ideo's redesign of the Bloomberg example, for example.

http://www.ideo.com/work/bloomberg-terminal-concept

[Jasper_]

What's the point of the virtual post-it note? Why not use a regular post-it note?

[bichiliad]

You know, not too sure. The micro-terminals make sense, but the virtual post-its seem a bit odd.

[3rd3]

I think this is not a general problem, it’s just that it’s very easy to make mistakes when changing a running system. You probably overgeneralize based on the fact that there were so many bad re-design decisions in massively used products lately. Just to name a few:

- YouTube’s failed attempt at improving the comment section

- The new mobile YouTube app entirely covers the video with supersized play/pause button.

- The odd merge of touch and desktop UI elements in Windows 8

I think a huge factor for hate is also whether the re-design was preannounced. I think, ideally the user would have the option to upgrade the page, just as they can decide whether to upgrade their phone or not.

[ronaldx]

We might consider that the existing, originally successful designs are products of competition and natural selection of early alternatives, both within and outside of the company. In other words, it's plausible that good design is evolutionary. The rest of my comment explores this idea.

If so, these existing designs are really the best-of-the-bunch, beating out the competition, and perhaps this is because they are better on criteria that have been overlooked or not fully understood.

Now that Facebook, Google, Wikipedia have "won" the competitions in their original space, we might believe they are no longer under strong selective pressure of competitive evolution.

Facebook and Google might therefore be inclined towards major redesign which benefit them more for commercial reasons. This alienates some people, yes, but since the companies no longer have to fear immediate challenges from competitors, they can afford to provide users with a less optimised and more profitable design.

Wikipedia has no real motivation to change the design away from one which is known to be good: nobody strongly benefits from a major redesign.

[pekk]

What if your "forward-thinking improvement" doesn't capture the reasons the original worked. Then actually you are proposing worse over better, just because it is newer.

But you made it, so you won't say that your own baby is ugly ever no matter how many people provide reasoned arguments

[vezzy-fnord]

You also have to justify that it's necessary, which this page did not.

Rather, it seemed like they were doing a redesign for the hell of it. Out of some belief that if something hasn't changed in X time, it should. It tries to justify itself using itself.

We've been using the Harvard architecture for decades now. Should we change it? It's been some time, after all.