website/sotc/2000/0222

<!-- 22 Feb 2000 -->
   <P>I'd like to say first off that based on the feedback I've gotten from
   my last State of the Code update, I'll probably be doing these more
   often. They'll also be showing up on the webpage soon I hope. I'm
   keeping archives until that happens. This one's a bit long because a
   lot has happened and I should have wrriten this about Friday night.
   Sorry for the delay.
   
<H2>0.1.1 update</H2>

   <P>In general, the update for 0.1.1 is coming along nicely. A lot of the
   changes I have been working on have actually been committed to thre
   tree. If you grab the Release_0_1 tag out of CVS right now, your
   version will still say it is 0.1.0, but it should have most of the
   bugs fixed. All in all, a very good thing. Here's what I've still got
   to do with it:
   <UL>
     <LI>Report of a segfault somewhere. This should be fixed now, but if
       it's not fixed I need to know.
     <LI>I don't like my current fix for noclip in unstable and I don't
       want to backport it. In a few days I hope I will know if there is
       a better solution. I'll fall back on the working solution I have
       in unstable if I can't find a better solution. If I change what
       I'm doing, it'll be tested on unstable first for quality control.
     <LI>Report the backtile isn't getting drawn around the classic
       statusbar in the -3dfx target. It gets drawn for all the others
       and testing this target isn't easy for me to do, so I haven't done
       it yet. Will soon.
   </UL>
       
   <P>That's it! Tentatively, 0.1.1 can be ready to roll as early as this
   weekend. You can still grab the Release_0_1_0 tag for the current
   release, but please grab the current stuff and "test" the hell out of
   it. 0.1.0 was pretty good, but 0.1.1 should fix every known issue I've
   been told about since then. Break it if you can, please!
   
<H2>Unstable updates</H2>

   <P>If you have been living under a rock and don't know it, unstable works
   again! Mostly. Mercury has, in fear for his mortal life, undone many
   of the changes made to unstable which caused so much breakage before.
   The only target not working again is the SDL target because nobody
   else but me seems to be building -sdl and just because I build it
   doesn't mean I understand that target. (Remember, my areas are the QF
   build system and some of the stuff deep in the engine core!) If you
   can fix it, step up and do so.
   
   <P>Win32 support. xC and Endy have Win32 "working" again under VC++, for
   some versions of working. xC is working on the problems, but could use
   a hand if you can help him. Here's what's broken:
   <UL>
     <LI>asm code in software (gas2nasm problems?)
     <LI>software renderer isn't at the moment
     <LI>qw-server seems to be having issues
   </UL>
       
   <P>The protocol breakage between the QW clients and server has been
   fixed. We are still (unfortunately) compatible with Win32 2.3x clients
   with all the problems that invites. We'll probably so remain until
   we're done with the merge because we don't want to change the protocol
   on win32 people until we know we can support them. Currently we just
   don't have enough win32 developers with their hands in the code to
   make sure it remains working.
   
   <P>I've begun working on QER/QSG colored lighting in QF. It's the first
   QER feature I'm working on in the hope that we'll eventually be
   working from a common codebase. That's up to them of course but we all
   know the QF tree is superior to the q1source tree. Still, they may not
   want to have to work around software renderer limitations, so we'll
   see.
   
   <P>The merge continues at impressive speeds. The merge hall of fame has
   a new face today as taniwha has proven his insanity by filling the CVS
   list with commits of merged files.
   
<H2>Project relations</H2>

   <P>We've been talking to QuakeWorld Forever about their secure QuakeWorld
   implementation. I will probably at some point be helping them
   implement public key cryptography. I've explored the best way to do
   this (Netrek uses a similar strategy) and have a few ideas. We'll see
   how they pan out.
   
   <P>As I mentioned above, I'm implementing some QER features and Endy from
   QSG has been picking away at a few of our Win32 problems. I generally
   consider these two groups to be closely related, but we're going to be
   working with them more to make this a truly multi-project effort and
   ensure some standardization and compatibility between our projects.
   
   <P>If you haven't heard, you've lived under a rock for quite some time
   now. QuakeLives is and has been almost since the project's inception,
   in violation of the GNU GPL. They tell us that they did not and do not
   have Copyright infringement problems with that license, the GNU GPL.
   They demanded we remove any statement saying otherwise from our
   website. (Mercury says he left QuakeLives because of GPL
   violations---this is noted on our news page.) Since we can prove
   otherwise, we've refused.
   
   <P>Speaking of QuakeLives, the MegaTF website was distributing copies of
   a release marked 2.54 without source. Slade of QuakeLives told us that
   MegaTF was distributing "warez" and later denied making that
   statement. We've got logs if anybody cares enough to read them,
   naturally. At any rate, Ambush has removed the offending binaries
   until he has some idea where things really stand with QuakeLives. He
   is not retargetting MegaTF in the meantime. Talk to him, not to us,
   etc.