#447 posted by necros on 2010/10/24 20:45:16
wow, there's even more going on than i realized...
thanks for the clarification, i'll just go ahead and start suggesting a 2mb zone or something from now on. :)
Possible Bug?
#448 posted by taniwha on 2010/11/22 12:42:04
I'm in the process of porting protocol 666 into QuakeForge, and I ran into this (in SV_WriteEntitiesToClient):
//johnfitz -- don't send model>255 entities if protocol is 15
if (sv_protocol == PROTOCOL_NETQUAKE && (int)ent->v.modelindex & 0xFF00)
continue;
Shouldn't that be sv.protocol?
#449 posted by gb on 2010/11/22 14:50:08
sv_protocol is the cvar, IIRC.
#450 posted by mh on 2010/11/22 17:25:57
> Shouldn't that be sv.protocol?
I suspect that you're right. sv_protocol is the command whereas sv.protocol is the currently active protocol which only gets updated when a new server is spawned. Using sv_protocol here would mean that server messages would go all wacko if the protocol was changed while a server was running.
Yeah...
#451 posted by metlslime on 2010/11/22 20:27:53
that's a mistake. I'm not sure what happens if you compare the cvar directly rather than cvar.value ... it's a struct which means are you comparing the first element of the struct?
Sv_protocol
#452 posted by O.S. on 2010/11/22 20:58:39
.. is not a cvar, it's a global associated with the sv_protocol console command which sv.protocol is assigned to at server creation. So the comparison in SV_WriteEntitiesToClient() really needs sv.protocol and not sv_protocol.
Ah...
#453 posted by metlslime on 2010/11/22 21:11:12
Yeah it's been a little while since i looked at the code. I guess it works as written but is unsafe. Should be sv.protocol.
Ha ha ;>
Nice to see QuakeForge is getting updated.
#455 posted by taniwha on 2010/11/23 06:49:31
Due to a series of unfortunate events, QF more or less went into hibernation for a few years, though I kept poking at it occasionally (there is no year with 0 commits). I've been fairly busing hacking on things this past year, and lately, the maps on quaddicted have given me some real motivation :). git helps a lot, too (easy branching and stash).
I'm glad to have been of help.
Bug
When running Fitz 0.85 with -quoth, it's impossible to switch to the id1 gamedir with game id1, because the engine thinks that the Quoth dir is the id1 dir, and thus nothing changes.
Not A Bug
#457 posted by negke on 2011/01/23 11:21:24
If it weren't like this, certain releases wouldn't work they way they do. Those that come with an extra pak/mod dir but still require quoth as a base.
But I Know That
I'm just saying that typing 'game id1' in the console should probably revert that behavior.
Yeah...
#459 posted by metlslime on 2011/01/23 21:40:07
that was implemented by request because BJP quake has it. ideally there would be ways to turn that stuff on/off in the console. (maybe multiple gamedirs like darkplaces would do it.) But to get the alternate hipnotic/rogue statusbars, you'd also need special variables for those.
#460 posted by necros on 2011/01/24 00:30:03
i was really happy you added the -quoth thing, but i guess it does make more sense for a more generalized multimod loading order thing.
maybe there'd be a way to detect what kind of hud layout to use based on what the gfx.wad contains? or are they all the same?
#461 posted by mh on 2011/01/24 10:56:31
if (hipnotic || quoth)
????
Ideally
#462 posted by necros on 2011/01/24 19:41:16
it would be awesome if it wasn't determined by folder name of the mod so you can make other mods that use the hud without having to get people to do a -hipnotic -game mymod.
checking quickly, the gfx.wad in hipnotic has all those extra weapon icons and id1 doesn't, so they do use different gfx.wad files. is that possible then, to look at what entries are in the wad file and then use an appropriate hud?
#463 posted by kaffikopp on 2011/01/26 16:09:47
Hey, I have a widescreen monitor (1920x1080) and changed my resolution in FitzQuake accordingly, and it works fine, except the UI (HUD, console, main menu) is terribly tiny. Are there any commands to prevent them from being downsized while maintaining the proper resolution?
Sort Of
#464 posted by rj on 2011/01/26 18:38:31
scr_conscale
scr_menuscale
scr_sbarscale
pretty self-explanatory... '1' is the default value, '2' doubles the size and so on. i find 1.5 a nice value
Cheers
#465 posted by kaffikopp on 2011/01/26 20:10:15
Just what I was looking for, also nice to be able to adjust the speed of the console.
Ahh
#466 posted by rj on 2011/01/26 20:58:17
what's that command? :)
Scr_conspeed
#467 posted by kaffikopp on 2011/01/26 21:31:56
Also, when I started up FitzQuake just now everything suddenly moved twice as fast, as if host_framerate had been altered, even though it hadn't. What gives?
Ah
#468 posted by necros on 2011/01/26 21:45:16
you must have a 64bit system and dual core (or more) cpu.
atm, there isn't a solution that i know of apart from using a different engine. you can use quakespasm for now, which is basically fitzquake, except it uses some sdl stuff for timing which gets rid of the problem.
apparently, the way glquake (and by extension fitzquake) calculates time is with some method that doesn't take into account more than one processor.
#469 posted by kaffikopp on 2011/01/26 21:56:30
Yeah, I do. I normally use DarkPlaces anyway, just thought I'd give FitzQuake a shot. There was nothing wrong with it earlier today though, and I didn't disable any cores or anything in the meantime. Strange...
#470 posted by necros on 2011/01/26 21:57:53
i think i recall some people saying disabling additional cores worked for them, but that didn't work for me.
#471 posted by mh on 2011/01/26 23:54:35
> Also, when I started up FitzQuake just now everything suddenly moved twice as fast, as if host_framerate had been altered, even though it hadn't. What gives?
Quake's default timer is shit and Fitz still uses it. :( 64-bit, multi-core or power saving will wreck it, and it really needs an engine-side fix.
|