The short version - the site was running an incredibly antique version of TikiWiki
which it turns out had security holes you could drive an ocean liner through, so not good. Unfortunately the automated upgrade refused to work correctly, and insisted on giving me a completely broken site instead. Which is more secure, to be sure, but not especially useful. So... site has been down while I manually rebuilt it from scratch with the latest Tiki[0] (4.2).
Hyperfiction, photo galleries, and the forums are gone for now, and may or may not return in the future. All the old blog posts going back to 2001 are still here - looking through it there's some rather embarrassing material in there, in hindsight 
. And I hadn't realised 'till now that I'd started thinking about dreamcast emulation way way back in early 2002. Wow. It has been a while.
[0] I looked at a few other CMS/blog options, but in the end it would still have been more work to switch to something else, and Tiki is really quite nice when it's configured properly.
Nathan's Journal
Last post Tue 09 of Mar., 2010 04:33 UTC
(1 Posts | 106711 Visits | Activity=2.00)
I finally got spiffy fixed up this weekend, mainly to get it out from under my desk and off my todo list (spiffy being our media PC that suffered a total hard disk failure nearly 2 years ago, and has never really recovered properly). The main difficulty has been getting it onto the wireless network - unfortunately linux support for wireless devices (other than orinoco and intel, which seem to be fairly stable) can still be a bit dicey.
Lessons learned:
- There are multiple versions of the Linksys WUSB54G, which have wildly different chipsets, none of which are guaranteed to work out of the box under linux. Frankly, I'd still avoid USB wireless devices as much as possible.
- Firmware updates sometimes actually include new features. In this case, the 1.2.x updates to the Netgear WG602v3 added what they call Client Bridge mode (ie, it bridges an ethernet segment to the primary AP, and doesn't accept wireless clients of its own). Oddly enough, this is largely why I originally bought the second one nearly two years ago. About darn time...
- ATI no longer supports the Radeon 9200 in the fglrx driver. Fortunately the x.org ati driver now actually supports TV output quite well, and we don't need ATI's one any more.
- The XC-Cube is old enough now that things like LIRC actually support its remote control out-of-the-box, without only a moderate amount of messing around with obscure configuration files.
On a completely unrelated point, does anyone want a nearly-new WUSB54Gv4?
And yes, I do still need to post some comments on Japan, sometime soon I promise...
Statistics (based on data from Metacritic, numbers do include downloadable content):
The Xbox 360 has 319 games of which 227 are unique (do not appear on PS3/Wii) (71.1% unique)
The PS3 has 118 games of which 41 are unique (do not appear on XBOX/Wii) (34.7% unique)
For unique games rated 75% or better, XBOX has 89 (39.2% of uniques), PS3 has 18 (43.9 % of uniques)
At an 85% cutoff, XBOX has 19 (7.8%) and PS3 has 5 (12.2%)
Big win for XBOX
Fighting:
PS3: Tekken 5
XBOX: DOA4
We'll call that one a tie, although personally I'm not a big fan of the Tekken series.
Karaoke:
PS3: Singstar
XBOX: ???
One for PS3
Top-rated exclusive:
PS3: Uncharted: Drake's Fortune or Ratchet & Clank (89)
XBOX: Bioshock (96)
Win for XBOX
Price:
PS3 40Gb: $AUD699
XBOX 20Gb: $AUD549
Definite win for XBOX
Runs linux:
PS3: Officially yes, but no GPU access.
XBOX: Officially no. Software hack exists for specific kernel versions.
Win for PS3
Manufacturer:
Sony: Secretly installed rootkits on customers' computers (not noticeably punished for it).
Microsoft: Convicted monopolist (not noticeably punished for it).
Hmm, tied again.
Looks like the xbox is ahead at the moment...
In unrelated news, Super Mario Galaxy is awesomely great . Although not part of this comparison it's interesting to note that the Wii has 94 exclusives out of 135 total titles (69.6%) tracking closely with the xbox, but has fewer top-ranked uniques - 19 (20.2%) at 75% and only 6 (6.38%) at 85% or better. In keeping with tradition, 6 of the top 10 are first-party Nintendo product.
FWIW: On a non-exclusive basis, 15 titles are XBOX/WII, 51 titles are PS3/XBOX, and 26 titles are fully cross-platform. There are no titles at all that are PS3/WII.
This has been your useless post of the day 
I guess this is what I get for accidentally leaving my glasses where Kira could get to them. *sigh*
In unrelated news, elr is now on SourceForge, and will be maintained there for the immediate future. Also if anyone is interested in contributing to it, let me know (in other words, patches gladly accepted )
I've got to stop doing point releases... 
There's a couple of fairly important bug fixes in this one, along with a couple of minor usability features. Full changelog is here as always.
In other news I've finally gotten the btk (binary toolkit) parser up to something approximating "scratch", which means that after a 3 month diversion I'm almost back to where I was 3 months ago when I decided to rewrite it in C++ and with a real parser...
There's a couple of minor tweaks plus some basic C++ support - get it at the Elr page, as always. This will probably be the last release for a little while as I'm working more heavily on the binary toolkit at the moment. Unfortunately it's going very slowly for reasons that have more to do with me being out of practice wrt C++ than anything else. On the bright side it's pretty much coming back to me; on the down side I'm also remembering why I never actually _liked_ C++ all that much
to deafening silence . This version features a couple of minor fixes and a bunch of usability improvements. The biggest change was probably support for some primitive typing across default reductions (ie reductions without an explicit action). Full
changelog is on the Elr page.
I've released a dev version of elr (dubbed 0.5). It builds, installs, seems to work, but isn't very tested and is missing a few rather important features (such as meaningful diagnostics). Still, it's a start. I expect to work on it fairly heavily in the near future, at least up to the point where its stable enough for production usage.
Wow... that's a long time since the last update. Personally I blame Puzzle Quest.
Lately I've been working on resurrecting the old parser generator I wrote for my honours project many years ago . It needed a few updates to bring it up to the current C++ spec, and since then I've been fixing up a few of the many little issues that make the difference between a prototype and a production system.
The main reason I started doing this is that I needed a frontend for the assembly encoder/decoder generator that I was (in turn) writing to support lxdream development. And to be perfectly honest, using yacc/lex is rather like pulling teeth after having had (the persistently inaccurately named) ELR[0] to play with. Plus, the grammar I had in mind pretty much required its automagic parser-state-based lexical-context abilities.
At the moment I'm tidying up a bunch of string and buffering issues in the skel code, and then I'll probably do a point release and get on with the assembly stuff for a while. It would be good to get elr to the point where it's a feasible replacement for yacc/lex for everyone though, so I'm certainly intending to get back to it in the future.
Oh, and in other somewhat out-of-date news we have a spoodle (don't blame me for the jacket...)
[0] It was originally intended to do both minimal-state LR(1), and EBNF - hence the name. I never had time during the original development phase though, but the name stuck nonetheless. Said features may make it in yet though, albeit probably not the first release version.
I've finally gotten around to getting a proper journal setup on lxdream.org, so I'll be using it for all future DC updates, leaving this one for personal / other stuff.
Which probably means this journal will get updated even less than previously
Sidebar
Blogs