Today I mentioned to Keith that I was thinking about redesigning my site (after all, it’s only been up like this for 3½ years), and he responded that he had already started on his redesign of Asterisk.
Trusting person that he is, he sent me the link to look at an early version of it, and lousy friend that I am, I immediately posted the link on my site for the whole ‘world’ (well, about 50 of you) to see (actually, he said it was okay).
So, here is Asterisk version 3 (I think, right?). First off, don’t look at it in IE as the styles are super broken in that browser at this point.
Overall impressions: Nice! I like the header a lot – the navigation actually looks different to what many sites are doing these days and harks back (IMO) to an older style with the large boxes and use of numbers. I like the toggle tool for the main image – a nice little feature.
The large main headline looks good – although I wonder if there are any plans for some SIFR ‘flashiness’? The Flikr feed is pretty cool; hmmm, I wouldn’t mind adding that to my site, although all the pics would mainly be of my son, so maybe not.
Of course, there’s no mention of the name of the site, just the cryptic logo on the right. That’s a pretty ballsy move; I wonder if it’s going to stay like that? Of course, it’s Keith’s personal site so he can do whatever the hell he wants, and I, for one, like the fact the he could be going out on a limb here.
I immediately noticed that the design was wider than 800px (although not by much). I’m not sure if this is intentional, but I suspect it is. Given Keith’s audience (web designers = higher resolution monitors), I would say go for it and build for the 1024 crowd; I think the design is going to need the extra width in order to add in sufficient white space.
Anyway, great job so far, Keith – I look forward to seeing how things progress.
Since one of the ideas is to split strings not into words, but hopefully into phrases more semantically informative than the words they are made of, doing that better should mean better suggestions, and avoiding what essentially are word n-tuples should make for smaller data and slightly faster querying.