Now to fix the bugs they found ...
Other stuff: Been in America for three weeks, went to Siggraph, it was a lot of fun, but nothing of earthshattering coolness got unveiled there. Glitzy though!
What up? Been a while ...
We made a release of Horde-1.3.x yesterday, and a whole bunch of the new modules are going to be snapshotted and cut as tarballs.
Hrm, work is keeping me busy. I'm going to load WINE onto my OpenBSD laptop, get it ported and working, and then get OpenFX working reliably under it. Now that would rule! Also re-writing IMP's MIME system at the moment into a more 'proper' object model, to support the Horde MIME_Viewer framework and some new features.
Off travelling around the States for a month, should be fun - starting in Vegas!
Wow, busy days - been a while! I started my new job with Network Appliance, which totally rules so far :)
Got the final code release from Stuart for OpenFX, and am just going through it and importing into CVS and other administrivia before uploading it to the FTP for a 1.0 release! Here is a nice screenshot to whet your appetite in the meanwhile :-)
Also did a fair bit of Horde stuff, and our CVS-web viewer called Chora is coming along nicely. Put in some branch visualisation stuff, so that you can track the progress of a file across RCS branches more easily. Look at this for an example of what I mean. Should be releasing this stably quite soon now; the MIME stuff needs a massive cleanup, and some other goals, but its almost there. Check out the demo if you are interested.
No database access abstraction, so there's no hope for trying to bolt it up to (for example) PostgreSQL instead without heavy modifications. One site (warning, Spanish ahead) mentioned that they were working on creating an abstraction layer, but it doesn't look like they've released anything yet. <RANT>This has got to be PHP's single biggest failing: lack of a project- independant transparent database API, such as Perl's DBI/DBD scheme, or Python's DB- API.</RANT>
Actually, PHP4 is starting to get a very decent DB abstraction layer in the form of PEAR, which is its form of Perl's CPAN.
If you install the PEAR that comes with the latest PHP4, look in /usr/local/lib/pear/DB.php and check it out. The current development version of IMP is using it now, and everything works rather well.
You should look at OpenFX andrei. It's been under development for years, and was a commercial product, but we are going to release it really really soon now.
Feature list includes a raytracer and scan-line renderer simultaneously (which not even 3DS has I believe), bezier/NURB support, skeletal animation, procedural textures, a variety of conversion utilities, PNG/GIF/JPEG/AVI/FLIC formats, dynamic output plugins, etc. Check out the screenshot gallery I threw together for more ideas. this and this are two of my personal favourite pictures :)
I've had it in a 'private' beta for a while now, but I figure there's no harm in letting the advogato crew have a look at it - we are going to release just as soon as we can get the primary tarball and website ready. This is a Win32 binary ONLY at the moment - if you want source, I'll send you an old snapshot to look at.
Feel free to download it from ftp://ftp.openfx.org/pub/openfx/beta and send me feedback at firstname.lastname@example.org. Incidentally, offers of US mirrors would be greatly appreciated (I run rsyncd).
The catch? It's a Win32 ONLY program at the moment. I'm using the excellent WINE project to compile it natively at the moment, but time pressures on this mean that the dev team eagerly awaits release and other volunteers to help us get the codebase into shape. Luckily, the code only uses pure Win32 calls (no MFC or any of that malarchy) so converting it to ANSI C and GCC will prove no great difficulty using WINE I feel.
In case anyone is interested, I just ran across this referrer analysis I did on my logs of the Mars Polar Lander site on its landing (err, crashing) day.
Kind of interesting, as it shows the relative power of the news-site referrals (they all had it on their front pages at the time, pretty much).
Didnt get much of a chance to hack this weekend, being busy with other things; but I've almost managed to finish off the DocBook ports for OpenBSD, including Norman Walsh's modular stylesheets.
They'll hopefully be ready for testing on Monday, and should be committed soon after that, so OpenBSD-2.8 will finally sate my DocBook needs! I've actually started using the DocBook-XML stuff now, so I guess I better get on and port that as well.
Marc Espie very kindly replied to some of my clueless emails explaining the innards of the OpenBSD ports system - it is very, very, very cool indeed, and the cleanest of all the BSDs that I've seen - hopefully some of his ideas will be used as the basis of the unified BSD ports system !
Hey hey, what's been happening? Had a great weekend here - weather stopped pouring long enough to have a huge game of rollerhockey on Sunday, but been raining ever since!
On the IMP front, I've been hacking trying to replace the 2.3 navigator view with a collapsible DHTML tree. I'm veering towards the idea of cobbling my own tree view together from Dan Steinman's excellent objects, as none of the available ones seem that good. Can anyone point out a good DHTML tree that works on a wide range of browsers?
On the OpenBSD front, I've just been wandering through the tree taking maintainership for poor orphaned ports, and trying to clean up the list of non-FAKEd ports. It will be good to get them all out in time for 2.8, and be left with a really clean tree!
OpenFX has stalled a bit while we try to get WinCVS working; it is such a pain sometimes! Might just do the initial release without anon-cvs access, and add it later if/when we need it...
Has anyone ever used the CPAN Perl module File::Tail ?
If you have, and gotten it to work, and you drop me a mail, I will owe you for life :)
/me spent most of an evening on it trying to get it to work on Linux/OpenBSD/anything!!
New HTML Parser: The long-awaited libxml2 based HTML parser code is live. It needs further work but already handles most markup better than the original parser.
Keep up with the latest Advogato features by reading the Advogato status blog.
If you're a C programmer with some spare time, take a look at the mod_virgule project page and help us with one of the tasks on the ToDo list!