Why on earth am I up this early on a Bank Holiday?
I keep doing this - I wake up early when I don't need to, then sleep in when I should be doing something else. :/
Anyway, this starts my last full week of freedom, so I suppose practising waking up early isn't such a bad idea. Next Wednesday is sixth form registration in the morning, and yet another hospital checkup in the afternoon.
...whereas this week is six months since my last Chemo treatment, which I guess passes for a milestone. I seem fine so far, but then I suppose there's nothing like having cancer to turn you into a hypochondriac.
(It's taken me far too long to write a few small paragraphs, so now it's past 9am and hence Not Really Early anymore, but I like the title so I'm keeping it.)
Well, I'm happy to report that the standard back-to-school clothes/stationery shopping procedure is as boring as ever. Hopefully this time it's at least leading to something useful.
Recently I also finished off a stylesheet change to chatzilla.rdmsoft.com (which I would call a "visual refresh" if if wasn't for the fact that I despise that term). There are still a few minor glitches in Opera and it's fairly broken in IE, but then I'm being lazy and assuming most people interested in ChatZilla are using Gecko.
I love UI.
Swiftly skipping over the fact I shouldn't care what a JRE is, I'm still left wondering why I get this dialog when I ask for HTML via File → Export, but not if I use File → Save As. Oh, and [No] and [Cancel] do the same thing.
"Too complex to merge" is also a fun one, which I got playing with tables in Writer. Anyway, it runs out the thing I was trying to test isn't valid HTML anyway, so I gave up.
- The bottom links panel now floats, so you don't have to scroll to the bottom to hit the LXR link. Having anything use
position:fixedseems to come with a scrolling performance loss, but I think it's worth it.
- I added code to read CVS/Entries to get the file version, and show that along with links to CVS Log and Blame on bonsai.mozilla.org.
- Added a script that tracks which line the page is scrolled to and adds a link next to the LXR and Blame links to go to that line in those tools. This works with a handful of non-standard DOM bits and a binary search function that I managed to keep turning into an infinite loop.
- Line numbers that you click on, or stay scrolled to for a while, get added to a list above the other links.
- Hovering over an item in the line link list will scroll the page to that line, then scroll back when the pointer moves away. I'm still a bit unsure about how buggy this bit is, but it seems to work okay most of the time.
It's definitely not prefect yet, but I do like using it more than LXR for ChatZilla stuff. I almost wish it was more scalable - it'd certainly die if I pointed it at anything more than just mozilla/extensions/irc.
- Finished off the ChatZilla patch for NAMESX I'd been sitting on for ages.
- Played a fair bit of Scorched3D. Since I last played there's been visual updates and some pretty interesting mods added.
- Managed to get the latest release of BitlBee to compile under Cygwin. Their news page is currently sporting the old "X before Windows Vista/Duke Nukem Forever/Debian Sarge" line, though since X is MSN file transfer support I think I'll let them off. :)
- Got search fully working on my new RdMise code. I'm now using Yahoo!'s serialized PHP output, as a workaround for XML in PHP sucking so much.
- Tried to reduce my information overload a bit by setting Thunderbird to check for mail less often and unsubscribing from some blogs/podcasts/newsgroups. Limited success so far...
- Dug up a half-finished PHP project I hadn't touched since January. I'll probably just give it up again, but it's worth a try.
Spammers routinely prove the Rules of Spam are valid.
I mean, come on, that's just silly. :|
(It's also a test to see any browser I have does anything useful with the longdesc attribute, though I doubt anyone else cares about that... Future edit: removed longdesc, added better alt.)