the linked page is blog spam. The content is taken from an earlier Daily Telegraph article without attribution:

http://www.telegraph.co.uk/technology/6125914/How-20-popular...

I've seen earlier screenshots of Amazon.com than that one. http://www.telco2.net/blog/images/amazon%20early%20screen%20...

I was also suspicious of the Amazon homepage. I seem to recall that they didn't even sell music until after they purchased CDNow.

As one commenter pointed out: I didn't know you could buy DVDs in 1995.

Most of these (e.g. Twitter, Flickr) are inaccurate since they're simply using the Wayback Machine which often misses css and images, completely changing the appearance.

Wikipedia strongly reminded me of WikiWikiWeb, http://c2.com/cgi/wiki?WelcomeVisitors. I actually kind of prefer this sparser look, at least for things that don't have a strong visual aspect.

Aside from the inspiration via how quickly knocked together some of these look, it's interesting to see early directions that have since been abandoned, eg: YouTube's dating site style "I am a $gender looking for $gender" drop-down.

I would say that Amazon and Drudge look the most "recognizable" between then and now (with Google a close 3rd).

I wonder if that's simply because they haven't done a major pivot like many of the others.

The design for Twitter that launched in 2006 looks pretty ugly. I am sure there were better designs in 2006. It looks like 1995.

Twitter wasn't even called Twitter when it launched; it was twttr, and it had a pretty hideous logo: http://www.flickr.com/photos/factoryjoe/190576959/

I think the Twitter image shown was preceded by this vowel-less version:

http://www.flickr.com/photos/factoryjoe/190576959/

This reinforces Reid Hoffman's maxim that if you're not embarrassed when you launch version 1, you spent too much time on it.

Amazon hasn't changed much.

craigslist hasn't changed much either, it is still ugly. BBC looked good though.

Ha ha ha - Amazon hasn't changed! They're still stuck in 1995 UI / UX.

They have (http://www.telco2.net/blog/images/amazon%20early%20screen%20...) but once they found something that worked well it makes sense they'd stick with it. A pretty design that causes a 1% drop in sales would be pretty stupid, and I'm sure they've A/B tested such things.

Damnit. I want an eMate 300 now.