From spirolat at gmx.net Tue Mar 6 14:20:57 2007 From: spirolat at gmx.net (uwe koch) Date: Tue, 06 Mar 2007 20:20:57 +0100 Subject: feature: advanced demo navigation Message-ID: <45EDBF19.80907@gmx.net> hi, a nice feature for demo playback would be the possibility to rewind or jump to a specific point in time. especially for moviemakers this makes life a lot easier. other quakeengines already offer this: warsow has CL_DemoJump_f(), q3mme has demoPlaySeek(), so their implementation could be used as inspiration. would be nice cyrri From monk at rq3.com Wed Mar 7 13:57:59 2007 From: monk at rq3.com (monk at rq3.com) Date: Wed, 7 Mar 2007 11:57:59 -0700 (MST) Subject: [quake3] feature: advanced demo navigation In-Reply-To: <45EDBF19.80907@gmx.net> References: <45EDBF19.80907@gmx.net> Message-ID: <56443.63.150.173.150.1173293879.squirrel@mail.rq3.com> I see listed as a feature: AVI video capture of demos. "video [filename]" Start video capture (use with demo command). Outputs to filename. "stopvideo" Stop the video capture. It seems like it would make more sense to dump the entire demo to raw video and then manipulate it in a dedicated video processor like Premier or whatever is used nowadays? Were you wanting to jump to specific parts of a demo to then start a screen capture program to capture the output to a video file? I can only see this being useful if you're low on HDD space and don't want to dump the entire demo, just certain snippets. Maybe a toggleable key that doesn't interrupt demo playback would be useful, as long as you hold it down, it will record to a file, when you let go it stops, when you depress it again, it starts recording again (maybe to another filename instead of appending the original), etc. Anything else with demo manipulation would seem better suited to a mythical "KeyGrip3" which would allow for pausing, flying around wherever you want for camera positioning, etc. rather than shoehorning a subset of those features into the Q3 engine. Although if a Q3 engine project did have a 'studio' mode like KeyGrip it in, that'd be rad too. But it seems like it should be all or nothing. Right now dumping to raw video and then manipulating THAT seems like the best solution? Monk. > hi, > a nice feature for demo playback would be the possibility to rewind or > jump to a specific point in time. especially for moviemakers this makes > life a lot easier. other quakeengines already offer this: warsow has > CL_DemoJump_f(), q3mme has demoPlaySeek(), so their implementation could > be used as inspiration. > would be nice > cyrri > From monk at rq3.com Wed Mar 7 14:04:24 2007 From: monk at rq3.com (monk at rq3.com) Date: Wed, 7 Mar 2007 12:04:24 -0700 (MST) Subject: [quake3] feature: advanced demo navigation In-Reply-To: <45EDBF19.80907@gmx.net> References: <45EDBF19.80907@gmx.net> Message-ID: <56451.63.150.173.150.1173294264.squirrel@mail.rq3.com> Sorry for the extra spam, but when I think about it again, it would be nice to have a dedicated demo playback set of controls. Fast forward, reverse, pause, step by (server?) frame, etc. But I think that'd be good just for viewing and reviewing demos. Good for reviewing demos for league games looking for cheating (I used to administer a Q2 league/ladder for http://www.worldogl.com/), actually. And it's always nice to be able to pop around in demos in other games, too. Like a DVD player control overlay or something. But this is totally separate from machinima considerations. For machinima, I think there's a different set of requirements and I think the dump to AVI solution already in place makes the most sense outside of a dedicated demo editing suite ala KeyGrip. Just my thoughts. Monk. > hi, > a nice feature for demo playback would be the possibility to rewind or > jump to a specific point in time. especially for moviemakers this makes > life a lot easier. other quakeengines already offer this: warsow has > CL_DemoJump_f(), q3mme has demoPlaySeek(), so their implementation could > be used as inspiration. > would be nice > cyrri > From monk at rq3.com Wed Mar 7 14:07:54 2007 From: monk at rq3.com (monk at rq3.com) Date: Wed, 7 Mar 2007 12:07:54 -0700 (MST) Subject: [quake3] feature: advanced demo navigation In-Reply-To: <56451.63.150.173.150.1173294264.squirrel@mail.rq3.com> References: <45EDBF19.80907@gmx.net> <56451.63.150.173.150.1173294264.squirrel@mail.rq3.com> Message-ID: <56492.63.150.173.150.1173294474.squirrel@mail.rq3.com> > And it's always nice to be able to > pop around in demos in other games, too. LIKE in other games. Not have ioq3 play demos from other games. But comparing to, say, the playback features of a racing car game, for example. They tend to have the features I described. Sorry, I re-read that and noticed that line made me look insane. No more spamses I promises. Monk. From spirolat at gmx.net Wed Mar 7 15:14:01 2007 From: spirolat at gmx.net (uwe koch) Date: Wed, 07 Mar 2007 21:14:01 +0100 Subject: [quake3] feature: advanced demo navigation In-Reply-To: <56451.63.150.173.150.1173294264.squirrel@mail.rq3.com> References: <45EDBF19.80907@gmx.net> <56451.63.150.173.150.1173294264.squirrel@mail.rq3.com> Message-ID: <45EF1D09.5050602@gmx.net> monk at rq3.com schrieb: > ... when I think about it again, it would be > nice to have a dedicated demo playback set of controls. Fast forward, > reverse, pause, step by (server?) frame, etc. But I think that'd be good > just for viewing and reviewing demos. Good for reviewing demos for league > games looking for cheating (I used to administer a Q2 league/ladder for > http://www.worldogl.com/), actually. And it's always nice to be able to > pop around in demos in other games, too. Like a DVD player control > overlay or something. thats what i meant. i just used movie makers as example because they spend a lot of time reviewing demos to find scenes worth capturing. whenever they are not sure about a scene and want to see it again, they have to reload the demo and use fast forward. not a big deal unless you want to dig through some 100 scenes... btw: dump to avi is not an option, since it costs to much time and HDD space to dump complete demos (or demo collections). of course anybody else who reviews a lot of demos (like league admins) would benefit as well. cyrri From zakk at timedoctor.org Wed Mar 7 15:18:19 2007 From: zakk at timedoctor.org (Zachary Slater) Date: Wed, 07 Mar 2007 12:18:19 -0800 Subject: [quake3] feature: advanced demo navigation In-Reply-To: <45EF1D09.5050602@gmx.net> References: <45EDBF19.80907@gmx.net> <56451.63.150.173.150.1173294264.squirrel@mail.rq3.com> <45EF1D09.5050602@gmx.net> Message-ID: <45EF1E0B.4040503@timedoctor.org> This is all very good and I'm happy to accept patches on it as long as they do not interfere with 1.32 mod compatibility. -- - Zachary J. Slater zakk at timedoctor.org zacharyslater at gmail.com From jorgepblank at gmail.com Wed Mar 7 16:43:33 2007 From: jorgepblank at gmail.com (Jorge Pena) Date: Wed, 7 Mar 2007 13:43:33 -0800 Subject: [quake3] feature: advanced demo navigation In-Reply-To: <45EF1E0B.4040503@timedoctor.org> References: <45EDBF19.80907@gmx.net> <56451.63.150.173.150.1173294264.squirrel@mail.rq3.com> <45EF1D09.5050602@gmx.net> <45EF1E0B.4040503@timedoctor.org> Message-ID: <28406b400703071343x23dc5abbmc0df7a5a4db3de00@mail.gmail.com> Sweet. I don't know if this is possible, or realistic, but what Monk said about being able to view a demo and move the camera around in the world (Is this what you meant?) would be extremely crazy (As in, cool). Or is this already possible? Heh. I mean like, instead of having to stage the cameras at record time, make it so that the camera could be moved around in view-time, this way I think it'd make for some really neat videos and it'd be really useful. Imagine moving the camera around to see if someone's around the corner, etc., instead of just being stuck to the recorder's perspective. Again, I don't know if this is possible or realistic (Or already implemented, haha), just thought it'd be cool. On 3/7/07, Zachary Slater wrote: > > This is all very good and I'm happy to accept patches on it as long as > they do not interfere with 1.32 mod compatibility. > -- > - Zachary J. Slater > zakk at timedoctor.org > zacharyslater at gmail.com > -------------- next part -------------- An HTML attachment was scrubbed... URL: From ntoronto at cs.byu.edu Wed Mar 7 18:11:34 2007 From: ntoronto at cs.byu.edu (Neil Toronto) Date: Wed, 07 Mar 2007 16:11:34 -0700 Subject: [quake3] feature: advanced demo navigation In-Reply-To: <28406b400703071343x23dc5abbmc0df7a5a4db3de00@mail.gmail.com> References: <45EDBF19.80907@gmx.net> <56451.63.150.173.150.1173294264.squirrel@mail.rq3.com> <45EF1D09.5050602@gmx.net> <45EF1E0B.4040503@timedoctor.org> <28406b400703071343x23dc5abbmc0df7a5a4db3de00@mail.gmail.com> Message-ID: <45EF46A6.6080909@cs.byu.edu> These are called "Multi-View Demos" and (IIRC) OSP mod was the first to do them. (CPMA and Ultra Freeze Tag (mine) did it, along with probably a few more.) It consists of three parts: 1) Send *every* entity to the player taking the MVD, not just those in the PVS (potentially visible set) 2) Mark the demo as an MVD by setting a configstring 3) Enable the cgame keycatcher when playing demos back if playing an MVD, interpret keystrokes This can be as easily done in a mod as in the engine. If the engine did it, it would do basically the same thing. One good argument for doing it in a mod is that the mod can better control who is allowed to make an MVD. Recording an MVD is a great way to maximize the effectiveness of a wall hack. Neil Jorge Pena wrote: > Sweet. I don't know if this is possible, or realistic, but what Monk > said about being able to view a demo and move the camera around in the > world (Is this what you meant?) would be extremely crazy ( As in, > cool). Or is this already possible? Heh. I mean like, instead of > having to stage the cameras at record time, make it so that the camera > could be moved around in view-time, this way I think it'd make for > some really neat videos and it'd be really useful. Imagine moving the > camera around to see if someone's around the corner, etc., instead of > just being stuck to the recorder's perspective. Again, I don't know if > this is possible or realistic ( Or already implemented, haha), just > thought it'd be cool. From jorgepblank at gmail.com Wed Mar 7 18:15:43 2007 From: jorgepblank at gmail.com (Jorge Pena) Date: Wed, 7 Mar 2007 15:15:43 -0800 Subject: [quake3] feature: advanced demo navigation In-Reply-To: <45EF46A6.6080909@cs.byu.edu> References: <45EDBF19.80907@gmx.net> <56451.63.150.173.150.1173294264.squirrel@mail.rq3.com> <45EF1D09.5050602@gmx.net> <45EF1E0B.4040503@timedoctor.org> <28406b400703071343x23dc5abbmc0df7a5a4db3de00@mail.gmail.com> <45EF46A6.6080909@cs.byu.edu> Message-ID: <28406b400703071515n2d07d874re676299135cc73f7@mail.gmail.com> Wow that's pretty cool, it'd be great if you could make some type of patch to the engine, although I appreciate your explanation either way. This is pretty nice! On 3/7/07, Neil Toronto wrote: > > These are called "Multi-View Demos" and (IIRC) OSP mod was the first to > do them. (CPMA and Ultra Freeze Tag (mine) did it, along with probably a > few more.) It consists of three parts: > > 1) Send *every* entity to the player taking the MVD, not just those in > the PVS (potentially visible set) > 2) Mark the demo as an MVD by setting a configstring > 3) Enable the cgame keycatcher when playing demos back if playing an > MVD, interpret keystrokes > > This can be as easily done in a mod as in the engine. If the engine did > it, it would do basically the same thing. One good argument for doing it > in a mod is that the mod can better control who is allowed to make an > MVD. Recording an MVD is a great way to maximize the effectiveness of a > wall hack. > > Neil > > Jorge Pena wrote: > > Sweet. I don't know if this is possible, or realistic, but what Monk > > said about being able to view a demo and move the camera around in the > > world (Is this what you meant?) would be extremely crazy ( As in, > > cool). Or is this already possible? Heh. I mean like, instead of > > having to stage the cameras at record time, make it so that the camera > > could be moved around in view-time, this way I think it'd make for > > some really neat videos and it'd be really useful. Imagine moving the > > camera around to see if someone's around the corner, etc., instead of > > just being stuck to the recorder's perspective. Again, I don't know if > > this is possible or realistic ( Or already implemented, haha), just > > thought it'd be cool. > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From kloppenburg at snt.utwente.nl Wed Mar 7 18:22:43 2007 From: kloppenburg at snt.utwente.nl (Erik Kloppenburg) Date: Thu, 08 Mar 2007 00:22:43 +0100 Subject: [quake3] feature: advanced demo navigation In-Reply-To: <45EF46A6.6080909@cs.byu.edu> References: <45EDBF19.80907@gmx.net> <56451.63.150.173.150.1173294264.squirrel@mail.rq3.com> <45EF1D09.5050602@gmx.net> <45EF1E0B.4040503@timedoctor.org> <28406b400703071343x23dc5abbmc0df7a5a4db3de00@mail.gmail.com> <45EF46A6.6080909@cs.byu.edu> Message-ID: <45EF4943.6030607@snt.utwente.nl> Neil Toronto wrote: > This can be as easily done in a mod as in the engine. If the engine did > it, it would do basically the same thing. One good argument for doing it > in a mod is that the mod can better control who is allowed to make an > MVD. Recordin What if it were serverside only? Server knows already everything that's going on, so there's no extra risk then. From jorgepblank at gmail.com Wed Mar 7 18:27:49 2007 From: jorgepblank at gmail.com (Jorge Pena) Date: Wed, 7 Mar 2007 15:27:49 -0800 Subject: [quake3] feature: advanced demo navigation In-Reply-To: <45EF4943.6030607@snt.utwente.nl> References: <45EDBF19.80907@gmx.net> <56451.63.150.173.150.1173294264.squirrel@mail.rq3.com> <45EF1D09.5050602@gmx.net> <45EF1E0B.4040503@timedoctor.org> <28406b400703071343x23dc5abbmc0df7a5a4db3de00@mail.gmail.com> <45EF46A6.6080909@cs.byu.edu> <45EF4943.6030607@snt.utwente.nl> Message-ID: <28406b400703071527w11caf7d9hc506512ffcc94add@mail.gmail.com> Oooohhh, good idea! So then the server would give the information to the person recording, no? So that'd still pose a risk, no? Either way I'd like to see how this is implemented with a patch if it were possible (If not, it's alright), it just seems very interesting. On 3/7/07, Erik Kloppenburg wrote: > > Neil Toronto wrote: > > This can be as easily done in a mod as in the engine. If the engine did > > it, it would do basically the same thing. One good argument for doing it > > in a mod is that the mod can better control who is allowed to make an > > MVD. Recordin > > What if it were serverside only? Server knows already everything that's > going on, so there's no extra risk then. > -------------- next part -------------- An HTML attachment was scrubbed... URL: From arny at ats.s.bawue.de Wed Mar 7 18:38:16 2007 From: arny at ats.s.bawue.de (Thilo Schulz) Date: Thu, 8 Mar 2007 00:38:16 +0100 Subject: [quake3] feature: advanced demo navigation In-Reply-To: <28406b400703071527w11caf7d9hc506512ffcc94add@mail.gmail.com> References: <45EDBF19.80907@gmx.net> <45EF4943.6030607@snt.utwente.nl> <28406b400703071527w11caf7d9hc506512ffcc94add@mail.gmail.com> Message-ID: <200703080038.20902.arny@ats.s.bawue.de> On Thursday 08 March 2007 00:27, Jorge Pena wrote: > Oooohhh, good idea! So then the server would give the information to the > person recording, no? So that'd still pose a risk, no? Either way I'd like > to see how this is implemented with a patch if it were possible (If not, > it's alright), it just seems very interesting. Don't get your hopes too high. All this daydreaming is nice but up to now there has been noone stepping up saying "I'll code this". As long as there's noone putting his time into it, all this simply won't happen. As Zakk already said - if someone were to make a patch it has the chance of being accepted or at least end up in the patch section for ioquake3. -- Thilo Schulz -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available URL: From kloppenburg at snt.utwente.nl Wed Mar 7 18:45:18 2007 From: kloppenburg at snt.utwente.nl (Erik Kloppenburg) Date: Thu, 08 Mar 2007 00:45:18 +0100 Subject: [quake3] feature: advanced demo navigation In-Reply-To: <28406b400703071527w11caf7d9hc506512ffcc94add@mail.gmail.com> References: <45EDBF19.80907@gmx.net> <56451.63.150.173.150.1173294264.squirrel@mail.rq3.com> <45EF1D09.5050602@gmx.net> <45EF1E0B.4040503@timedoctor.org> <28406b400703071343x23dc5abbmc0df7a5a4db3de00@mail.gmail.com> <45EF46A6.6080909@cs.byu.edu> <45EF4943.6030607@snt.utwente.nl> <28406b400703071527w11caf7d9hc506512ffcc94add@mail.gmail.com> Message-ID: <45EF4E8E.3090304@snt.utwente.nl> Jorge Pena wrote: > Oooohhh, good idea! So then the server would give the information to the > person recording, no? So that'd still pose a risk, no? No, I mean the server would record the demo and not give it out to anyone. Except after the match, it can put the demofile up for download on a site or whatever. From jorgepblank at gmail.com Wed Mar 7 18:54:58 2007 From: jorgepblank at gmail.com (Jorge Pena) Date: Wed, 7 Mar 2007 15:54:58 -0800 Subject: [quake3] feature: advanced demo navigation In-Reply-To: <45EF4E8E.3090304@snt.utwente.nl> References: <45EDBF19.80907@gmx.net> <56451.63.150.173.150.1173294264.squirrel@mail.rq3.com> <45EF1D09.5050602@gmx.net> <45EF1E0B.4040503@timedoctor.org> <28406b400703071343x23dc5abbmc0df7a5a4db3de00@mail.gmail.com> <45EF46A6.6080909@cs.byu.edu> <45EF4943.6030607@snt.utwente.nl> <28406b400703071527w11caf7d9hc506512ffcc94add@mail.gmail.com> <45EF4E8E.3090304@snt.utwente.nl> Message-ID: <28406b400703071554y519956e6h203253f5af85c49b@mail.gmail.com> Ohhh, yeah that seems reasonable. Thilo, yeah, still a cool topic though, I always thought about this but didn't know what it was called or thought anyone else would think it's a nice idea haha. On 3/7/07, Erik Kloppenburg wrote: > > Jorge Pena wrote: > > Oooohhh, good idea! So then the server would give the information to the > > person recording, no? So that'd still pose a risk, no? > > No, I mean the server would record the demo and not give it out to > anyone. Except after the match, it can put the demofile up for download > on a site or whatever. > -------------- next part -------------- An HTML attachment was scrubbed... URL: From ntoronto at cs.byu.edu Wed Mar 7 19:07:46 2007 From: ntoronto at cs.byu.edu (Neil Toronto) Date: Wed, 07 Mar 2007 17:07:46 -0700 Subject: [quake3] feature: advanced demo navigation In-Reply-To: <28406b400703071554y519956e6h203253f5af85c49b@mail.gmail.com> References: <45EDBF19.80907@gmx.net> <56451.63.150.173.150.1173294264.squirrel@mail.rq3.com> <45EF1D09.5050602@gmx.net> <45EF1E0B.4040503@timedoctor.org> <28406b400703071343x23dc5abbmc0df7a5a4db3de00@mail.gmail.com> <45EF46A6.6080909@cs.byu.edu> <45EF4943.6030607@snt.utwente.nl> <28406b400703071527w11caf7d9hc506512ffcc94add@mail.gmail.com> <45EF4E8E.3090304@snt.utwente.nl> <28406b400703071554y519956e6h203253f5af85c49b@mail.gmail.com> Message-ID: <45EF53D2.3050305@cs.byu.edu> Ultra Freeze Tag had server-side demos. It was quite a hack - in a good way. The demo files were only a bit larger than MVDs of the same match. You could even play them back on the server, and have your team connect and watch the demo together. It was just like watching a regular match, but it was recorded. Even though they were so much more useful than MVDs, people didn't tend to use them because they were so hard to distribute. MVDs got more attention just because of that. You don't have to have access to the server to swipe an MVD of your clan match - you just need to be a coach or a spectator, and record it locally. Server-side demos might be easier to do in-engine. We had to jump through quite a few hoops to make it work well. And for the curious: no, I'm not stepping up to implement any of this. I've got grad school and PyQuake3 to do - other than that, I'm just a [quake3] troll. :) Neil Jorge Pena wrote: > Ohhh, yeah that seems reasonable. Thilo, yeah, still a cool topic > though, I always thought about this but didn't know what it was called > or thought anyone else would think it's a nice idea haha. > > On 3/7/07, *Erik Kloppenburg* > wrote: > > Jorge Pena wrote: > > Oooohhh, good idea! So then the server would give the > information to the > > person recording, no? So that'd still pose a risk, no? > > No, I mean the server would record the demo and not give it out to > anyone. Except after the match, it can put the demofile up for > download > on a site or whatever. > > From ntoronto at cs.byu.edu Wed Mar 7 19:19:36 2007 From: ntoronto at cs.byu.edu (Neil Toronto) Date: Wed, 07 Mar 2007 17:19:36 -0700 Subject: [quake3] feature: advanced demo navigation In-Reply-To: <45EF53D2.3050305@cs.byu.edu> References: <45EDBF19.80907@gmx.net> <56451.63.150.173.150.1173294264.squirrel@mail.rq3.com> <45EF1D09.5050602@gmx.net> <45EF1E0B.4040503@timedoctor.org> <28406b400703071343x23dc5abbmc0df7a5a4db3de00@mail.gmail.com> <45EF46A6.6080909@cs.byu.edu> <45EF4943.6030607@snt.utwente.nl> <28406b400703071527w11caf7d9hc506512ffcc94add@mail.gmail.com> <45EF4E8E.3090304@snt.utwente.nl> <28406b400703071554y519956e6h203253f5af85c49b@mail.gmail.com> <45EF53D2.3050305@cs.byu.edu> Message-ID: <45EF5698.10909@cs.byu.edu> Lovely self-reply... Neil Toronto wrote: > *snip* > > Server-side demos might be easier to do in-engine. We had to jump > through quite a few hoops to make it work well. It was here we learned that the QVM has the same endian-ness as the architecture it ran on, which surprised us a bit. We had to write our own hton*() and ntoh*() functions... Anyway, *supposing* someone did this, it might be easiest to record every player's commands and userinfo strings, and play them back as if the players were actually connected. Then the mod wouldn't need support for the feature - it would just think some players were connected. Well, it would need to know when a demo was being played back so it could keep players from joining the game. This assumes, of course, that server events proceed deterministically. Hmm... maybe this is harder than I thought. The engine used to have a "serverrecord" command that recorded server-side demos, but I don't know how it worked or what's become of it, and I don't know how to play them back. Neil From jorgepblank at gmail.com Wed Mar 7 19:24:47 2007 From: jorgepblank at gmail.com (Jorge Pena) Date: Wed, 7 Mar 2007 16:24:47 -0800 Subject: [quake3] feature: advanced demo navigation In-Reply-To: <45EF5698.10909@cs.byu.edu> References: <45EDBF19.80907@gmx.net> <45EF1E0B.4040503@timedoctor.org> <28406b400703071343x23dc5abbmc0df7a5a4db3de00@mail.gmail.com> <45EF46A6.6080909@cs.byu.edu> <45EF4943.6030607@snt.utwente.nl> <28406b400703071527w11caf7d9hc506512ffcc94add@mail.gmail.com> <45EF4E8E.3090304@snt.utwente.nl> <28406b400703071554y519956e6h203253f5af85c49b@mail.gmail.com> <45EF53D2.3050305@cs.byu.edu> <45EF5698.10909@cs.byu.edu> Message-ID: <28406b400703071624o54753d9ch6599181d1b7c7cce@mail.gmail.com> >Anyway, *supposing* someone did this, it might be easiest to record >every player's commands and userinfo strings, and play them back as if >the players were actually connected. Haha, excuse my ignorance, but I've always imagined something like that! In fact, that's how I actually thought it originally worked, the regular recording haha, wow I'm pretty stupid. But yeah that's a good idea. On 3/7/07, Neil Toronto wrote: > > Lovely self-reply... > > Neil Toronto wrote: > > *snip* > > > > Server-side demos might be easier to do in-engine. We had to jump > > through quite a few hoops to make it work well. > > It was here we learned that the QVM has the same endian-ness as the > architecture it ran on, which surprised us a bit. We had to write our > own hton*() and ntoh*() functions... > > Anyway, *supposing* someone did this, it might be easiest to record > every player's commands and userinfo strings, and play them back as if > the players were actually connected. Then the mod wouldn't need support > for the feature - it would just think some players were connected. Well, > it would need to know when a demo was being played back so it could keep > players from joining the game. > > This assumes, of course, that server events proceed deterministically. > Hmm... maybe this is harder than I thought. > > The engine used to have a "serverrecord" command that recorded > server-side demos, but I don't know how it worked or what's become of > it, and I don't know how to play them back. > > Neil > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From monk at rq3.com Wed Mar 7 19:54:08 2007 From: monk at rq3.com (monk at rq3.com) Date: Wed, 7 Mar 2007 17:54:08 -0700 (MST) Subject: [quake3] feature: advanced demo navigation In-Reply-To: <45EF46A6.6080909@cs.byu.edu> References: <45EDBF19.80907@gmx.net> <56451.63.150.173.150.1173294264.squirrel@mail.rq3.com> <45EF1D09.5050602@gmx.net> <45EF1E0B.4040503@timedoctor.org> <28406b400703071343x23dc5abbmc0df7a5a4db3de00@mail.gmail.com> <45EF46A6.6080909@cs.byu.edu> Message-ID: <58471.63.150.173.150.1173315248.squirrel@mail.rq3.com> Sorry, I don't know how to do short emails! Quake 2 had a working command called "serverrecord", if memory serves me correctly. KeyGrip2 was able to then manipulate that. I can't recall if it could use it all directly or if it had to parse out each individual player's demo into a separate file. I was looking into this a few years back for RQ3 and do recall running into MVD. I think the people involved at the time, OSP and CPMA, didn't really return my inquiries. And, as Thilo points out, coding resources to roll our own for the sake of machinima was scarce. I also ran into OpenDemo: http://www.machinima.com/opendemo/ Which seems to have morphed into LMPC or something: http://sourceforge.net/projects/lmpc http://machinima.com/opendemo/blog/ Again, at the time, neither MVD, OpenDemo, or rolling our own solution was attractive due to both lack of resources AND lack of supporting programs. It's all well and good that you have an MVD or OpenDemo, but is there something that can allow you to recam and edit? I suppose LMPC does that, now. One of the main stumbling blocks of Q3 machinima and demo editing was that the network stream (and hence the demos) was encrypted, if my memory serves me. Quake 2 was pretty mature for machinima. For instance, the fabled KeyGrip2 is actually still being used despite not being actively developed for... 6 years? Something like that, I believe: http://jitspoe.planetquake.gamespy.com/tutorials/keygrip2.html (short recamming tutorial) http://z-studios.com/tutorials/keygrip2-tutorials-quake-2/ It's been years since I used it, but I believe KG2 used Quake 2 as its render engine. I recall needing a separate install of Q2 so I could play around with KG2. Are there any specs or anything released for MVD for Q3? All I find are QuakeWorld stuff. From zakk at timedoctor.org Wed Mar 7 20:39:22 2007 From: zakk at timedoctor.org (Zachary Slater) Date: Wed, 07 Mar 2007 17:39:22 -0800 Subject: [quake3] feature: advanced demo navigation In-Reply-To: <58471.63.150.173.150.1173315248.squirrel@mail.rq3.com> References: <45EDBF19.80907@gmx.net> <56451.63.150.173.150.1173294264.squirrel@mail.rq3.com> <45EF1D09.5050602@gmx.net> <45EF1E0B.4040503@timedoctor.org> <28406b400703071343x23dc5abbmc0df7a5a4db3de00@mail.gmail.com> <45EF46A6.6080909@cs.byu.edu> <58471.63.150.173.150.1173315248.squirrel@mail.rq3.com> Message-ID: <45EF694A.4020700@timedoctor.org> monk at rq3.com wrote: > While I'm dreaming, I'd also like a pony! > I would like an 8800 GTX to magically appear inside my computer replacing the 7800 GTX. FYI on the pony side of things I'm installing vista tonight and I'll see how well/poorly ioq3 works with it. Please direct all flames for installing vista to zakk at timedoctor.com -- - Zachary J. Slater zakk at timedoctor.org zacharyslater at gmail.com From ntoronto at cs.byu.edu Thu Mar 8 01:04:06 2007 From: ntoronto at cs.byu.edu (Neil Toronto) Date: Wed, 07 Mar 2007 23:04:06 -0700 Subject: [quake3] feature: advanced demo navigation In-Reply-To: <200703080038.20902.arny@ats.s.bawue.de> References: <45EDBF19.80907@gmx.net> <45EF4943.6030607@snt.utwente.nl> <28406b400703071527w11caf7d9hc506512ffcc94add@mail.gmail.com> <200703080038.20902.arny@ats.s.bawue.de> Message-ID: <45EFA756.7020507@cs.byu.edu> Thilo Schulz wrote: > On Thursday 08 March 2007 00:27, Jorge Pena wrote: > >> Oooohhh, good idea! So then the server would give the information to the >> person recording, no? So that'd still pose a risk, no? Either way I'd like >> to see how this is implemented with a patch if it were possible (If not, >> it's alright), it just seems very interesting. >> > > Don't get your hopes too high. All this daydreaming is nice but up to now > there has been noone stepping up saying "I'll code this". As long as there's > noone putting his time into it, all this simply won't happen. > As Zakk already said - if someone were to make a patch it has the chance of > being accepted or at least end up in the patch section for ioquake3. I'm not volunteering for this - I don't have the time - but I just realized that I *do* have the Ultra Freeze Tag code. (I had lost it in a reformat, but someone sent it to me just recently. That was nice.) It has both MVDs and SSDs. SSDs, IMO, are a bad idea for permanent inclusion, because small changes to the netcode necessitate changes in the SSD code, which is kind of hairy and scary. (Plus, it's very easy to forget.) MVDs, on the other hand, would be perfect. They're very general and frighteningly easy to implement. All you'd need is a restriction that the player taking the demo be a spectator. If anyone wants to do MVDs, I'd be happy to supply a reference implementation. That should cut the work down considerably. Neil From ntoronto at cs.byu.edu Thu Mar 8 01:47:27 2007 From: ntoronto at cs.byu.edu (Neil Toronto) Date: Wed, 07 Mar 2007 23:47:27 -0700 Subject: [quake3] feature: advanced demo navigation In-Reply-To: <45EFA756.7020507@cs.byu.edu> References: <45EDBF19.80907@gmx.net> <45EF4943.6030607@snt.utwente.nl> <28406b400703071527w11caf7d9hc506512ffcc94add@mail.gmail.com> <200703080038.20902.arny@ats.s.bawue.de> <45EFA756.7020507@cs.byu.edu> Message-ID: <45EFB17F.6010907@cs.byu.edu> Neil Toronto wrote: > Thilo Schulz wrote: >> On Thursday 08 March 2007 00:27, Jorge Pena wrote: >> >>> Oooohhh, good idea! So then the server would give the information to >>> the >>> person recording, no? So that'd still pose a risk, no? Either way >>> I'd like >>> to see how this is implemented with a patch if it were possible (If >>> not, >>> it's alright), it just seems very interesting. >>> >> >> Don't get your hopes too high. All this daydreaming is nice but up to >> now there has been noone stepping up saying "I'll code this". As long >> as there's noone putting his time into it, all this simply won't happen. >> As Zakk already said - if someone were to make a patch it has the >> chance of being accepted or at least end up in the patch section for >> ioquake3. > > I'm not volunteering for this - I don't have the time - but I just > realized that I *do* have the Ultra Freeze Tag code. (I had lost it in > a reformat, but someone sent it to me just recently. That was nice.) > It has both MVDs and SSDs. I just discovered that it's an old version without MVD support. I'll try to track down a more recent one. Neil From baztheallmighty at hotmail.com Thu Mar 8 05:00:25 2007 From: baztheallmighty at hotmail.com (Myles Treadwell) Date: Thu, 08 Mar 2007 10:00:25 +0000 Subject: [quake3] feature: advanced demo navigation In-Reply-To: <45EFB17F.6010907@cs.byu.edu> Message-ID: https://svn.sourceforge.net/svnroot/q2pro/q3sdc2 fast forward/slow mo == bind t timescale 4 0.5 etc etc Add rewind and your set. credit where credit is due: [SkulleR] wrote the above code >From: Neil Toronto >Reply-To: quake3 at icculus.org >To: quake3 at icculus.org >Subject: Re: [quake3] feature: advanced demo navigation >Date: Wed, 07 Mar 2007 23:47:27 -0700 >MIME-Version: 1.0 >Received: from icculus.org ([67.106.77.212]) by >bay0-mc5-f10.bay0.hotmail.com with Microsoft SMTPSVC(6.0.3790.2668); Wed, 7 >Mar 2007 22:47:40 -0800 >Received: (qmail 8015 invoked by alias); 8 Mar 2007 01:45:49 -0500 >Received: (qmail 8007 invoked by uid 305); 8 Mar 2007 01:45:49 -0500 >X-Message-Info: LsUYwwHHNt26rouEDe32Z7XJewZk5pCjtDt39Tz/3k0= >Mailing-List: contact quake3-help at icculus.org; run by ezmlm >Precedence: bulk >X-No-Archive: yes >List-Post: >List-Help: >List-Unsubscribe: >List-Subscribe: >Delivered-To: mailing list quake3 at icculus.org >User-Agent: Thunderbird 1.5.0.10 (X11/20070306) >References: <45EDBF19.80907 at gmx.net> <45EF4943.6030607 at snt.utwente.nl> ><28406b400703071527w11caf7d9hc506512ffcc94add at mail.gmail.com> ><200703080038.20902.arny at ats.s.bawue.de> <45EFA756.7020507 at cs.byu.edu> >Return-Path: quake3-return-1650-baztheallmighty=hotmail.com at icculus.org >X-OriginalArrivalTime: 08 Mar 2007 06:47:40.0164 (UTC) >FILETIME=[AA585440:01C7614D] > >Neil Toronto wrote: >>Thilo Schulz wrote: >>>On Thursday 08 March 2007 00:27, Jorge Pena wrote: >>> >>>>Oooohhh, good idea! So then the server would give the information to the >>>>person recording, no? So that'd still pose a risk, no? Either way I'd >>>>like >>>>to see how this is implemented with a patch if it were possible (If not, >>>>it's alright), it just seems very interesting. >>>> >>> >>>Don't get your hopes too high. All this daydreaming is nice but up to now >>>there has been noone stepping up saying "I'll code this". As long as >>>there's noone putting his time into it, all this simply won't happen. >>>As Zakk already said - if someone were to make a patch it has the chance >>>of being accepted or at least end up in the patch section for ioquake3. >> >>I'm not volunteering for this - I don't have the time - but I just >>realized that I *do* have the Ultra Freeze Tag code. (I had lost it in a >>reformat, but someone sent it to me just recently. That was nice.) It has >>both MVDs and SSDs. > >I just discovered that it's an old version without MVD support. I'll try to >track down a more recent one. > >Neil > _________________________________________________________________ Advertisement: Fresh jobs daily. Stop waiting for the newspaper. Search Now! www.seek.com.au http://a.ninemsn.com.au/b.aspx?URL=http%3A%2F%2Fninemsn%2Eseek%2Ecom%2Eau&_t=757263760&_r=Hotmail_EndText_Dec06&_m=EXT From tim at ngus.net Thu Mar 8 05:08:49 2007 From: tim at ngus.net (Tim Angus) Date: Thu, 08 Mar 2007 10:08:49 +0000 Subject: feature: advanced demo navigation In-Reply-To: References: Message-ID: <45EFE0B1.3050100@ngus.net> Myles Treadwell wrote: > credit where credit is due: > [SkulleR] wrote the above code Or did he? He had a reputation for using Q3 source code leaks for various things. From baztheallmighty at hotmail.com Thu Mar 8 05:28:15 2007 From: baztheallmighty at hotmail.com (Myles Treadwell) Date: Thu, 08 Mar 2007 10:28:15 +0000 Subject: [quake3] Re: feature: advanced demo navigation In-Reply-To: <45EFE0B1.3050100@ngus.net> Message-ID: hahaha true. Would be good if some of the mods "CPMA" for instance would release there code. He did say something in www.quakesrc.org/forums about using ogc's code in one of his previous SDC releases. I have no reason to believe that he didn't write it though. Speak up if u have evidence as it should probably be removed from source forge if it contains code from other's projects that are not open source. >From: Tim Angus >Reply-To: quake3 at icculus.org >To: quake3 at icculus.org >Subject: [quake3] Re: feature: advanced demo navigation >Date: Thu, 08 Mar 2007 10:08:49 +0000 >MIME-Version: 1.0 >Received: from icculus.org ([67.106.77.212]) by >bay0-mc5-f10.bay0.hotmail.com with Microsoft SMTPSVC(6.0.3790.2668); Thu, 8 >Mar 2007 02:09:02 -0800 >Received: (qmail 18657 invoked by alias); 8 Mar 2007 05:07:13 -0500 >Received: (qmail 18649 invoked by uid 305); 8 Mar 2007 05:07:13 -0500 >X-Message-Info: LsUYwwHHNt3660MmjhEvYg2f34OAemlK3oXsmRrh6gU= >Mailing-List: contact quake3-help at icculus.org; run by ezmlm >Precedence: bulk >X-No-Archive: yes >List-Post: >List-Help: >List-Unsubscribe: >List-Subscribe: >Delivered-To: mailing list quake3 at icculus.org >User-Agent: Thunderbird 1.5.0.10 (Windows/20070221) >References: >X-Virus-Scanned: by amavisd-new-20030616-p10 (Debian) at relay.picsel.com >Return-Path: quake3-return-1652-baztheallmighty=hotmail.com at icculus.org >X-OriginalArrivalTime: 08 Mar 2007 10:09:02.0338 (UTC) >FILETIME=[CBE1C620:01C76169] > >Myles Treadwell wrote: >>credit where credit is due: >>[SkulleR] wrote the above code > >Or did he? He had a reputation for using Q3 source code leaks for various >things. _________________________________________________________________ Advertisement: Fresh jobs daily. Stop waiting for the newspaper. Search now! www.seek.com.au http://a.ninemsn.com.au/b.aspx?URL=http%3A%2F%2Fninemsn%2Eseek%2Ecom%2Eau&_t=757263760&_r=Hotmail_EndText_Dec06&_m=EXT From zakk at timedoctor.org Tue Mar 13 23:34:50 2007 From: zakk at timedoctor.org (Zachary Slater) Date: Tue, 13 Mar 2007 20:34:50 -0700 Subject: Vista Message-ID: <45F76D5A.6030201@timedoctor.org> So far as I can tell, ioquake3 works fine in Vista. Though it isn't registered in the game browser. I'll need to find out how to implement that. -- - Zachary J. Slater zakk at timedoctor.org zacharyslater at gmail.com From zakk at timedoctor.org Tue Mar 13 23:40:02 2007 From: zakk at timedoctor.org (Zachary Slater) Date: Tue, 13 Mar 2007 20:40:02 -0700 Subject: [quake3] Vista In-Reply-To: <45F76D5A.6030201@timedoctor.org> References: <45F76D5A.6030201@timedoctor.org> Message-ID: <45F76E92.70805@timedoctor.org> Zachary Slater wrote: > So far as I can tell, ioquake3 works fine in Vista. Though it isn't > registered in the game browser. I'll need to find out how to implement > that. http://forums.winamp.com/showthread.php?s=&threadid=258969 apparently. -- - Zachary J. Slater zakk at timedoctor.org zacharyslater at gmail.com From jorgepblank at gmail.com Tue Mar 13 23:48:37 2007 From: jorgepblank at gmail.com (=?ISO-8859-1?Q?Jorge_Pe=F1a?=) Date: Tue, 13 Mar 2007 19:48:37 -0800 Subject: [quake3] Vista In-Reply-To: <45F76E92.70805@timedoctor.org> References: <45F76D5A.6030201@timedoctor.org> <45F76E92.70805@timedoctor.org> Message-ID: <28406b400703132048m4f12f167ja08e078770e0af4e@mail.gmail.com> Sweet. On 3/13/07, Zachary Slater wrote: > > Zachary Slater wrote: > > So far as I can tell, ioquake3 works fine in Vista. Though it isn't > > registered in the game browser. I'll need to find out how to implement > > that. > > http://forums.winamp.com/showthread.php?s=&threadid=258969 > > apparently. > -- > - Zachary J. Slater > zakk at timedoctor.org > zacharyslater at gmail.com > -------------- next part -------------- An HTML attachment was scrubbed... URL: From jorgepblank at gmail.com Tue Mar 13 23:52:24 2007 From: jorgepblank at gmail.com (=?ISO-8859-1?Q?Jorge_Pe=F1a?=) Date: Tue, 13 Mar 2007 19:52:24 -0800 Subject: Freetype Message-ID: <28406b400703132052sd2e3d96kb81e77082cdfd82a@mail.gmail.com> Hey guys, I just spent a while compiling, setting up Freetype inside my source tree, and modifying my visual studio solution settings and the Makefile, now I supposedly have Freetype support, all I did was enable BUILD_FREETYPE. Now, my question is, does this mean I can simply insert ttf's into the game? I think it was inside fonts/, at least that's what Tr3b says for his XreaL engine ( http://www.quakesrc.org/forums/viewtopic.php?t=6040&highlight=freetype ). I know I know, they're different engines, but I opened them both up and to my surprise they seemed to have the same things, at least, from what I could tell. Thanks guys, I'd appreciate any help. I'd test it myself but I don't know how to. I think I basically put them in there and then do something with the .menu files? What if I want to use it throughout the game, such as inside the console and what not? It's okay, I don't need ultra specific instructions if you guys don't have time, I'd appreciate at least a general answer though, if not, it's alright. Thanks again! -------------- next part -------------- An HTML attachment was scrubbed... URL: From arny at ats.s.bawue.de Wed Mar 14 09:26:45 2007 From: arny at ats.s.bawue.de (Thilo Schulz) Date: Wed, 14 Mar 2007 14:26:45 +0100 Subject: [quake3] Freetype In-Reply-To: <28406b400703132052sd2e3d96kb81e77082cdfd82a@mail.gmail.com> References: <28406b400703132052sd2e3d96kb81e77082cdfd82a@mail.gmail.com> Message-ID: <200703141426.48222.arny@ats.s.bawue.de> On Wednesday 14 March 2007 04:52, Jorge Pe?a wrote: > I'd test it myself but I don't know how to. I think I basically put them in > there and then do something with the .menu files? What if I want to use it > throughout the game, such as inside the console and what not? It's okay, I > don't need ultra specific instructions if you guys don't have time, I'd > appreciate at least a general answer though, if not, it's alright. I am not really well versed in the usage of these fonts, so all I can give you is a few hints. The first thing is that it does not simply suffice to enable the define BUILD_FREETYPE, so you'll need the library on your system and available to your build environment... but you probably know that already. The freetype stuff pretty much seems to be done in renderer/tr_fonts.c, so you should concentrate on that file if you want to know how freetype integration works. Use a debugger to step through the functions and see what everything does and you'll quickly grasp what the code already there is doing :) -- Thilo Schulz -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available URL: From jorgepblank at gmail.com Wed Mar 14 09:57:03 2007 From: jorgepblank at gmail.com (=?ISO-8859-1?Q?Jorge_Pe=F1a?=) Date: Wed, 14 Mar 2007 05:57:03 -0800 Subject: [quake3] Freetype In-Reply-To: <200703141426.48222.arny@ats.s.bawue.de> References: <28406b400703132052sd2e3d96kb81e77082cdfd82a@mail.gmail.com> <200703141426.48222.arny@ats.s.bawue.de> Message-ID: <28406b400703140657s129d4eday200f45048b26dbc2@mail.gmail.com> Okay cool, thanks. Yeah, I set everything up (The library, environment, since I'm on windows at the moment, etc. etc.), and defined BUILD_FREETYPE and it seems to work. I was just wondering if anyone knew whether this allowed me to simply drop in ttf files into the fonts directory, but it's okay I'll try to figure that out, thanks Thilo! On 3/14/07, Thilo Schulz wrote: > > On Wednesday 14 March 2007 04:52, Jorge Pe?a wrote: > > I'd test it myself but I don't know how to. I think I basically put them > in > > there and then do something with the .menu files? What if I want to use > it > > throughout the game, such as inside the console and what not? It's okay, > I > > don't need ultra specific instructions if you guys don't have time, I'd > > appreciate at least a general answer though, if not, it's alright. > > I am not really well versed in the usage of these fonts, so all I can give > you > is a few hints. The first thing is that it does not simply suffice to > enable > the define BUILD_FREETYPE, so you'll need the library on your system and > available to your build environment... but you probably know that already. > The freetype stuff pretty much seems to be done in renderer/tr_fonts.c, so > you > should concentrate on that file if you want to know how freetype > integration > works. Use a debugger to step through the functions and see what > everything > does and you'll quickly grasp what the code already there is doing :) > > -- > Thilo Schulz > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From vsteiss at web.de Thu Mar 29 10:21:13 2007 From: vsteiss at web.de (vsteiss at web.de) Date: Thu, 29 Mar 2007 16:21:13 +0200 Subject: advanced admin and user authentification Message-ID: <910917552@web.de> Hello, I have a question regarding this comment in bug#3019. https://bugzilla.icculus.org/show_bug.cgi?id=3019: >>>> (In reply to comment #13) > Configurable prefix used to compute the MD5 serving as a GUID will still be a > security hole. The only difference is that the attacked will have first to join > the server it wants to hack to record the prefix used there before configuring > his own server with the same prefix value. > > As it is now, it is still possible to hack the system ( although highly > unlikely ) if you can use a tool like tcpdump to capture the packets between > your victim and the server. Although greatly limited over the internet, such > hack method can still be used in more constrained environments like a big LAN > sharing a single internet connection. > > A good solution which wouldn't involve changing too much the client and the > server consists in using the Diffie-Hellman key agreement system. Such system > is considered secure against eavesdroppers as long as the keys are carefully > chosen. And that way, server operators can still share a single server key > between many servers and change the server IP. > > http://en.wikipedia.org/wiki/Diffie-Hellman_key_exchange > > Although the page says to discard a and b in the end. We want to keep them here > since they are the new qkeys. > > And for the implement you say? What about libcrypto, part of OpenSSL? > http://www.openssl.org/docs/crypto/dh.html I don't see how we can add any type of key exchange protocol without breaking the game protocol. Breaking the game protocol conflicts with ioq3 goals I think. [...] <<<< As it was closed with >>I'm closing this bug, the keying stuff belongs in a new "enhancement" ticket I guess.<< my question is: Is there any work in progress to make an advanced secure authentification? Regards Beselius _______________________________________________________________ SMS schreiben mit WEB.DE FreeMail - einfach, schnell und kostenguenstig. Jetzt gleich testen! http://f.web.de/?mc=021192 From arny at ats.s.bawue.de Thu Mar 29 11:28:14 2007 From: arny at ats.s.bawue.de (Thilo Schulz) Date: Thu, 29 Mar 2007 17:28:14 +0200 Subject: [quake3] advanced admin and user authentification In-Reply-To: <910917552@web.de> References: <910917552@web.de> Message-ID: <200703291728.19719.arny@ats.s.bawue.de> On Thursday 29 March 2007 16:21, vsteiss at web.de wrote: > As it was closed with >>I'm closing this bug, the keying stuff > belongs in a new "enhancement" ticket I guess.<< my question is: > Is there any work in progress to make an advanced secure authentification? no. -- Thilo Schulz -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available URL: From jorgepblank at gmail.com Fri Mar 30 00:16:03 2007 From: jorgepblank at gmail.com (=?ISO-8859-1?Q?Jorge_Pe=F1a?=) Date: Thu, 29 Mar 2007 20:16:03 -0800 Subject: Building On Mac Message-ID: <28406b400703292116j6708d8edn789a1e1d825d1d8d@mail.gmail.com> Hey guys, I already asked on IRC so, sorry for bothering. But I was able to SSH into my friend's mac for the first time in a long time (Probably like, 3 months) and a lot in my game has changed since then. I enabled freetype support with BUILD_FREETYPE, and I also enabled OGG Vorbis support as well as curl support. All of this works perfectly on Linux and Windows, however, when I tried it on Mac OS X, it didn't work. Note that before, the game built perfectly. I realized I needed the libraries, so I got them and compiled and installed but it still didn't find the libraries, so I checked in the /lib and /include directories but there was nothing there. So I basically did a ./configure --prefix=/tmp/whateverlibitwas and then did the rest as normal, and at the very end moved the things within /tmp/whateverlibitwas/lib to ioquake3's libs/macosx folder, since I had already seen libSDL in there I figured it'd pick those up too. Well, I started getting 'No such file or directory' errors in files where I'd do #include , and so I had to change them to be relevant inclusions instead of as they were (What I mean is, for example, I'd change it to #include "../blah/blah.h"), although inconvenient, I was desperate to build, so I tried. I noticed the libraries weren't being picked up after all, so after some ignorant skimming through the Makefile (Ignorant cause I'm not Makefile writer, I mean, I can hack things up so they work usually but I can't write them out of thin air), and I noticed that libSDL was being linked to explicitly, so I looked for a section in the Makefile that had to do with Mac, and to the LDFLAGS section I noticed an -L line, I man gcc'd this and then /-L'd, and I found out that this had to do with the inclusion of the directories it should look in for libraries. Logically, I added -Lcode/libs/macosx, but this didn't work either. I'm sorry for going through this a roundabout way instead of just getting to the point but I want to make sure I'm understood (Not because I think you guys are stupid, but because, since I'm no expert at this, I'm afraid I might be wording my question wrong). I now have every dylib and a file (Dynamic and static?) libraries in libs/macosx, but I'm still having problems with the header inclusions and library linking. At first I noticed it was asking for -lvorbisfile, for example, and the library was actually called libvorbisfile.a, so I renamed it to vorbisfile (Same with the rest), to no avail. Seeing as I personally don't have a mac or have ever used one for that matter (Other than through SSH, which I only know thanks to Linux), I'm rather confused as to the whole development structure of it. Where are headers and libraries held, is there a central place? What do you guys suggest I do? Tweaking everything over again just for it to work on Mac seems inconvenient, I'm sure I'm doing something wrong. Thanks guys, sorry for the long message, but I'd really appreciate it. And yes, I AM using the mac build script (Always have been). Thanks again, I really appreciate it. -------------- next part -------------- An HTML attachment was scrubbed... URL: From arny at ats.s.bawue.de Fri Mar 30 00:36:07 2007 From: arny at ats.s.bawue.de (Thilo Schulz) Date: Fri, 30 Mar 2007 06:36:07 +0200 Subject: [quake3] Building On Mac In-Reply-To: <28406b400703292116j6708d8edn789a1e1d825d1d8d@mail.gmail.com> References: <28406b400703292116j6708d8edn789a1e1d825d1d8d@mail.gmail.com> Message-ID: <200703300636.09913.arny@ats.s.bawue.de> On Friday 30 March 2007 06:16, Jorge Pe?a wrote: > Thanks guys, sorry for the long message, but I'd really appreciate it. And > yes, I AM using the mac build script (Always have been). Thanks again, I > really appreciate it. No problem mate, I think you have made very clear that you have gone to some trouble already. You don't strike me as the casual dork wanting to get free tech support without doing work and learning on his own, so I am glad to help :) First of all, your approach was already a step in the right direction. I think curl is present on MacOSX already so I don't think you'll need to build it yourself. Now on the topic of ogg I can advise you on the steps that I took to get a working MacOSX build. Get the source to the libraries libogg, libvorbis and libvorbisfile (libvorbisfile is built with libvorbis). run configure using --enable-static --disable-shared and build all three of them this way .. if you redistribute these builds you really want static ogg libraries as users most likely won't have them installed. If you want to make a universal binary, you must cross compile these three libraries at least for the other platform you want to support. After the build process is finished, you'll find the .a files for the static libraries somewhere in the lib/.libs/ directory. I suggest you copy them out of there and put them into some collective directory as you indicated you already did. Then you have to edit the Makefile to indicate the include directories + where to find these libraries in the build process. This is what the relevant sections of my Makefile look like in the macosx section: for ppc: BASE_CFLAGS += -arch ppc -DSMP \ -I/Users/thilo/libogg-1.1.3/include \ -I/Users/thilo/libvorbis-1.1.2/include \ -DMAC_OS_X_VERSION_MIN_REQUIRED=1020 -nostdinc \ -F/Developer/SDKs/MacOSX10.2.8.sdk/System/Library/Frameworks \ ... LDFLAGS += -arch ppc \ -L/Users/thilo/libmerge/ppc \ -L/Developer/SDKs/MacOSX10.2.8.sdk/usr/lib/gcc/darwin/3.3 \ -F/Developer/SDKs/MacOSX10.2.8.sdk/System/Library/Frameworks \ as you can see, I added a few lines. libmerge/ppc happens to be the place where I put the ppc ogg and vorbis lib stuff into. Similarly, for i386: BASE_CFLAGS += -arch i386 -DSMP \ -mmacosx-version-min=10.4 \ -I/Users/thilo/libogg-1.1.3/include \ -I/Users/thilo/libvorbis-1.1.2/include \ -DMAC_OS_X_VERSION_MIN_REQUIRED=1040 -nostdinc \ -F/Developer/SDKs/MacOSX10.4u.sdk/System/Library/Frameworks \ -I/Developer/SDKs/MacOSX10.4u.sdk/usr/lib/gcc/i686-apple-darwin8/4.0.1/include -isystem /Developer/SDKs/MacOSX10.4u.sdk/usr/include LDFLAGS = -arch i386 -mmacosx-version-min=10.4 \ -L/Users/thilo/libmerge/i386 \ -L/Developer/SDKs/MacOSX10.4u.sdk/usr/lib/gcc/i686-apple-darwin8/4.0.1 \ -F/Developer/SDKs/MacOSX10.4u.sdk/System/Library/Frameworks \ -Wl,-syslibroot,/Developer/SDKs/MacOSX10.4u.sdk,-m These are only meant as examples, you will have to give the directories depending on your setup of course. Notice, that I added a ",-m" in the very last block compared to the original, where no -m can be found after: MacOSX10.4u.sdk This may be required if building fails and the linker complains about some double defined OpenAL symbols if I remember correctly. You will also want to defuse this line: # Always include debug symbols...you can strip the binary later... BASE_CFLAGS += -gfull by commenting it. You don't want debugging if you're not a developer or trying to hunt a bug. It will only bloat your binary and hurt speed. I hope these instructions were sufficient. Good luck! -- Thilo Schulz -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available URL: From jorgepblank at gmail.com Fri Mar 30 00:57:13 2007 From: jorgepblank at gmail.com (=?ISO-8859-1?Q?Jorge_Pe=F1a?=) Date: Thu, 29 Mar 2007 20:57:13 -0800 Subject: [quake3] Building On Mac In-Reply-To: <200703300636.09913.arny@ats.s.bawue.de> References: <28406b400703292116j6708d8edn789a1e1d825d1d8d@mail.gmail.com> <200703300636.09913.arny@ats.s.bawue.de> Message-ID: <28406b400703292157x5cd5bb0fx281f0b823d4c8e15@mail.gmail.com> Thanks I really really appreciate it! I definitely worked hard on it haha, I was at it for about two hours, all trial and error as I personally have no exposure/experience at all to Macs. One question though, in the -L and -I switches, you put absolute paths, is this necessary? Sorry, I didn't find any information on this in the manual page, and I was wondering because, seeing how it was already hard enough for me, having other people build it might be difficult, so I was thinking about including these things in my subversion repository, for which I'd need relevant paths, no? Or maybe there's a variable such as: -I/Users/$(USER)/libogg-1.1.3/include Just double checked, looks like there is. No wait, better yet: -I/$(PWD)/code/whatever Seems right. I think this will definitely help, next time my friend is available I'll SSH and keep trying, but it all makes sense now, thanks I really, honestly appreciate your help! On 3/29/07, Thilo Schulz wrote: > > On Friday 30 March 2007 06:16, Jorge Pe?a wrote: > > Thanks guys, sorry for the long message, but I'd really appreciate it. > And > > yes, I AM using the mac build script (Always have been). Thanks again, I > > really appreciate it. > > No problem mate, I think you have made very clear that you have gone to > some > trouble already. You don't strike me as the casual dork wanting to get > free > tech support without doing work and learning on his own, so I am glad to > help :) > > First of all, your approach was already a step in the right direction. > I think curl is present on MacOSX already so I don't think you'll need to > build it yourself. > Now on the topic of ogg I can advise you on the steps that I took to get a > working MacOSX build. > > Get the source to the libraries libogg, libvorbis and libvorbisfile > (libvorbisfile is built with libvorbis). run configure > using --enable-static --disable-shared and build all three of them this > way .. if you redistribute these builds you really want static ogg > libraries > as users most likely won't have them installed. If you want to make a > universal binary, you must cross compile these three libraries at least > for > the other platform you want to support. > After the build process is finished, you'll find the .a files for the > static > libraries somewhere in the lib/.libs/ directory. I suggest you copy them > out > of there and put them into some collective directory as you indicated you > already did. > Then you have to edit the Makefile to indicate the include directories + > where > to find these libraries in the build process. > > This is what the relevant sections of my Makefile look like in the macosx > section: > for ppc: > BASE_CFLAGS += -arch ppc -DSMP \ > -I/Users/thilo/libogg-1.1.3/include \ > -I/Users/thilo/libvorbis-1.1.2/include \ > -DMAC_OS_X_VERSION_MIN_REQUIRED=1020 -nostdinc \ > -F/Developer/SDKs/MacOSX10.2.8.sdk/System/Library/Frameworks \ > ... > > LDFLAGS += -arch ppc \ > -L/Users/thilo/libmerge/ppc \ > -L/Developer/SDKs/MacOSX10.2.8.sdk/usr/lib/gcc/darwin/3.3 \ > -F/Developer/SDKs/MacOSX10.2.8.sdk/System/Library/Frameworks \ > > as you can see, I added a few lines. libmerge/ppc happens to be the place > where I put the ppc ogg and vorbis lib stuff into. > > Similarly, for i386: > BASE_CFLAGS += -arch i386 -DSMP \ > -mmacosx-version-min=10.4 \ > -I/Users/thilo/libogg-1.1.3/include \ > -I/Users/thilo/libvorbis-1.1.2/include \ > -DMAC_OS_X_VERSION_MIN_REQUIRED=1040 -nostdinc \ > -F/Developer/SDKs/MacOSX10.4u.sdk/System/Library/Frameworks \ > > -I/Developer/SDKs/MacOSX10.4u.sdk/usr/lib/gcc/i686-apple-darwin8/4.0.1/include > -isystem /Developer/SDKs/MacOSX10.4u.sdk/usr/include > LDFLAGS = -arch i386 -mmacosx-version-min=10.4 \ > -L/Users/thilo/libmerge/i386 \ > -L/Developer/SDKs/MacOSX10.4u.sdk/usr/lib/gcc/i686-apple-darwin8/4.0.1 > \ > -F/Developer/SDKs/MacOSX10.4u.sdk/System/Library/Frameworks \ > -Wl,-syslibroot,/Developer/SDKs/MacOSX10.4u.sdk,-m > > These are only meant as examples, you will have to give the directories > depending on your setup of course. Notice, that I added a ",-m" in the > very > last block compared to the original, where no -m can be found after: > MacOSX10.4u.sdk > This may be required if building fails and the linker complains about some > double defined OpenAL symbols if I remember correctly. > > You will also want to defuse this line: > # Always include debug symbols...you can strip the binary later... > BASE_CFLAGS += -gfull > > by commenting it. You don't want debugging if you're not a developer or > trying > to hunt a bug. It will only bloat your binary and hurt speed. > > I hope these instructions were sufficient. Good luck! > > -- > Thilo Schulz > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From ludwig.nussel at suse.de Fri Mar 30 03:04:05 2007 From: ludwig.nussel at suse.de (Ludwig Nussel) Date: Fri, 30 Mar 2007 09:04:05 +0200 Subject: [quake3] advanced admin and user authentification In-Reply-To: <200703291728.19719.arny@ats.s.bawue.de> References: <910917552@web.de> <200703291728.19719.arny@ats.s.bawue.de> Message-ID: <200703300904.05993.ludwig.nussel@suse.de> Thilo Schulz wrote: > On Thursday 29 March 2007 16:21, vsteiss at web.de wrote: > > As it was closed with >>I'm closing this bug, the keying stuff > > belongs in a new "enhancement" ticket I guess.<< my question is: > > Is there any work in progress to make an advanced secure authentification? > > no. Would be waste of time and effort anyways. ssh (1) - OpenSSH SSH client (remote login program) screen (1) - screen manager with VT100/ANSI terminal emulation cu Ludwig -- (o_ Ludwig Nussel //\ SUSE Labs V_/_ http://www.suse.de/ SUSE LINUX Products GmbH, GF: Markus Rex, HRB 16746 (AG Nuernberg) From bnoordhuis at gmail.com Sat Mar 31 04:18:21 2007 From: bnoordhuis at gmail.com (Ben Noordhuis) Date: Sat, 31 Mar 2007 10:18:21 +0200 Subject: Map cycle breaks on empty or bot only servers Message-ID: <54885e060703310118l4a78fae4l652de8915e1aa760@mail.gmail.com> Dear all, This is the e-mail accompanying https://bugzilla.icculus.org/show_bug.cgi?id=3076 - I hereby submit to your approval a patch to remedy the behavior as described both below and in the bug report: *** The summary says it all, really: due to a logic error in code/g_main.c's CheckIntermissionExit(), the ExitLevel() function is never called if the server is empty or populated by bots only and the server thus remains in intermission mode indefinitely. Reproducibility: always *** For completeness sake, I have attached the patch to this message also. Sincerely, Ben Noordhuis a.k.a. mis -------------- next part -------------- Index: code/game/g_main.c =================================================================== --- code/game/g_main.c (revision 1053) +++ code/game/g_main.c (working copy) @@ -1189,7 +1189,7 @@ ================= */ void CheckIntermissionExit( void ) { - int ready, notReady; + int ready, notReady, playerCount; int i; gclient_t *cl; int readyMask; @@ -1202,8 +1202,9 @@ ready = 0; notReady = 0; readyMask = 0; + playerCount = 0; for (i=0 ; i< g_maxclients.integer ; i++) { - cl = level.clients + i; + cl = level.clients + i; if ( cl->pers.connected != CON_CONNECTED ) { continue; } @@ -1211,6 +1212,7 @@ continue; } + playerCount++; if ( cl->readyToExit ) { ready++; if ( i < 16 ) { @@ -1236,16 +1238,21 @@ return; } - // if nobody wants to go, clear timer - if ( !ready ) { - level.readyToExit = qfalse; - return; - } + // 070331 mis - empty or bot only servers make the server + // hang indefinitely in the intermission, so execute the + // logic below only if there are any *real* players here. + if ( playerCount > 0 ) { + // if nobody wants to go, clear timer + if ( !ready ) { + level.readyToExit = qfalse; + return; + } - // if everyone wants to go, go now - if ( !notReady ) { - ExitLevel(); - return; + // if everyone wants to go, go now + if ( !notReady ) { + ExitLevel(); + return; + } } // the first person to ready starts the ten second timeout From q3urt.undead at gmail.com Sat Mar 31 14:28:47 2007 From: q3urt.undead at gmail.com (un dead) Date: Sat, 31 Mar 2007 14:28:47 -0400 Subject: cl_consoleHistory stores passwords in q3config.cfg Message-ID: Hi all, Sorry I couldn't check bugzilla to see if this is a known issue. I get an internal error when I try to connect to bugs or bugzilla. cl_consoleHistory saves all commands you type including /password and /rcon password. A new user who doesn't view the config may not even be aware of this issue. Here's how they could be exposed: 1) Join a server and use /password or /rcon password. 2) Exit the server and send someone your config A lot of users aren't aware their q3config.cfg has the passwords. To them it's a file generated by quake that they don't worry about. I don't think you should scrub for 'password' because then that lets typos pass through. Could you keep the console history in a separate file like what id does with q3key? A lot of people have multiple configs so perhaps changing/adding a cvar that points to a filename would be best. As long as you set the cvar to a reasonable default, the novice user doesn't have to worry about it. From tim at ngus.net Sat Mar 31 14:53:23 2007 From: tim at ngus.net (Tim Angus) Date: Sat, 31 Mar 2007 19:53:23 +0100 Subject: cl_consoleHistory stores passwords in q3config.cfg In-Reply-To: References: Message-ID: <20070331195323.a629e879.tim@ngus.net> On Sat, 31 Mar 2007 14:28:47 -0400 un wrote: > A lot of users aren't aware their q3config.cfg has the passwords. To > them it's a file generated by quake that they don't worry about. A system specific generated file should not be being sent to anyone else. It's better to write config files separately for things like scripts or specific sets of key bindings. Regardless, the password systems in Q3 are hardly a pantheon of security. Passwords are input and broadcast in cleartext, they really don't hold very much value. Therefore it's really pretty questionable whether implementing special cases is sensible where it is only to protect information that is already insecure. From kloppenburg at snt.utwente.nl Sat Mar 31 15:26:19 2007 From: kloppenburg at snt.utwente.nl (Erik Kloppenburg) Date: Sat, 31 Mar 2007 21:26:19 +0200 Subject: [quake3] Re: cl_consoleHistory stores passwords in q3config.cfg In-Reply-To: <20070331195323.a629e879.tim@ngus.net> References: <20070331195323.a629e879.tim@ngus.net> Message-ID: <460EB5DB.7070200@snt.utwente.nl> Tim Angus wrote: > A system specific generated file should not be being sent to anyone > else. It's better to write config files separately for things like > scripts or specific sets of key bindings. In the entire Quake 3 community it is very common practice to share configs with each other. 'Normal' users don't think about what consequences this can have and usually it has none. The risk of something happening is much higher thanks to this console history. Even if you think people shouldn't do it, it's still the reality that they do and it would be wise to take this into account. > Regardless, the password systems in Q3 are hardly a pantheon of > security. Passwords are input and broadcast in cleartext, they really > don't hold very much value. Therefore it's really pretty questionable > whether implementing special cases is sensible where it is only to > protect information that is already insecure. As I said: config sharing is a very common happening. Sniffing someone's packets is an entirely different area. From q3urt.undead at gmail.com Sat Mar 31 15:32:49 2007 From: q3urt.undead at gmail.com (un dead) Date: Sat, 31 Mar 2007 15:32:49 -0400 Subject: [quake3] Re: cl_consoleHistory stores passwords in q3config.cfg In-Reply-To: <20070331195323.a629e879.tim@ngus.net> References: <20070331195323.a629e879.tim@ngus.net> Message-ID: On 3/31/07, Tim Angus wrote: > On Sat, 31 Mar 2007 14:28:47 -0400 un wrote: > > A lot of users aren't aware their q3config.cfg has the passwords. To > > them it's a file generated by quake that they don't worry about. > > A system specific generated file should not be being sent to anyone > else. It's better to write config files separately for things like > scripts or specific sets of key bindings. But are new people going to do that? I know a lot of people who aren't very knowledgeable about configs. They aren't newbies but they would happily pass along configs to someone else. I see people on forums regularly post their configs or send configs to other people. There's no intuition for most people that it could lead to exposing your passwords. > Regardless, the password systems in Q3 are hardly a pantheon of > security. Passwords are input and broadcast in cleartext, they really > don't hold very much value. Therefore it's really pretty questionable > whether implementing special cases is sensible where it is only to > protect information that is already insecure. Hmm I don't think this is the same class of insecurity. True the passwords are broadcast in plain text but you would have to sniff their traffic in order to discover it. In the cl_consoleHistory case, all you need to do is send your q3config. It seems innocent enough to many users. Anyway, I just wanted to bring it up and now I'll drop it. :) I enjoy the work you guys have done on ioquake3. It's great. From jorgepblank at gmail.com Sat Mar 31 15:35:27 2007 From: jorgepblank at gmail.com (=?ISO-8859-1?Q?Jorge_Pe=F1a?=) Date: Sat, 31 Mar 2007 11:35:27 -0800 Subject: [quake3] Building On Mac In-Reply-To: <28406b400703292157x5cd5bb0fx281f0b823d4c8e15@mail.gmail.com> References: <28406b400703292116j6708d8edn789a1e1d825d1d8d@mail.gmail.com> <200703300636.09913.arny@ats.s.bawue.de> <28406b400703292157x5cd5bb0fx281f0b823d4c8e15@mail.gmail.com> Message-ID: <28406b400703311235w64df186flfdfa8b7cdef7344d@mail.gmail.com> Hey sorry to keep bothering, but here are the things I added (The relevant stuff at least): For PPC: BASE_CFLAGS += -arch ppc -DSMP \ -I$(PWD)/code \ -I$(PWD)/code/vorbis/include \ -I$(PWD)/code/freetype \ -DMAC_OS_X_VERSION_MIN_REQUIRED=1020 -nostdinc \ -F/Developer/SDKs/MacOSX10.2.8.sdk/System/Library/Frameworks \ -I/Developer/SDKs/MacOSX10.2.8.sdk/usr/include/gcc/darwin/3.3 \ -isystem /Developer/SDKs/MacOSX10.2.8.sdk/usr/include LDFLAGS += -arch ppc \ -L$(PWD)/code/libs/macosx \ -L/Developer/SDKs/MacOSX10.2.8.sdk/usr/lib/gcc/darwin/3.3 \ -F/Developer/SDKs/MacOSX10.2.8.sdk/System/Library/Frameworks \ -Wl,-syslibroot,/Developer/SDKs/MacOSX10.2.8.sdk,-m ARCH=ppc I did the same for i386, I didn't show it though because I'm actually building on PPC at the moment. Anyways, I also commented out the line: # Always include debug symbols...you can strip the binary later... #BASE_CFLAGS += -gfull It gets a lot further into the build process now, and then it spits out the following: ld: Undefined symbols: _fprintf$LDBLStub _sprintf$LDBLStub _inflate _inflateEnd _inflateInit2_ _inflateReset make[1]: *** [build/release-darwin-ppc/tip.ppc] Error 1 make: *** [build_release] Error 2 I google'd this and I found out it's a linker error (ld, duh). I google'd for the symbols, and I can't find what library they correspond to, I thinkit's zlib. Does anything make use of zlib from the following: libogg, libvorbis, libfreetype, libcurl? I'm just wondering, because before this, the engine built perfectly. But come to think of it, it probably IS a zlib dependency, I'll go and get the library. Actually, they said the library should already be in there, it wasn't there so I downloaded the tarball and built it myself. I then put the resulting static library (libz.a) in the macosx libs directory which has already been -L'd by the makefile, I ran the bash make script and it still gave me the error. So I edited the Makefile and added a -lz to a random part (To a part where it does -lvorbisfile as well), since I'm building with OGG I figured It'd pick it up, but it's still giving me the error. Anyways, I'm just curious because I didn't find any reference to zlib in the Makefile and I only found a reference in the README, but it only includes header files and source files of it, in other platforms like Windows, I don't have to go and download it and build it along with the game. I'm just wondering if you guys know anything about this, I've been investigating since yesterday I was up till 12 haha, I think I'm getting closer to finally being able to build though, I think it's just this that's left. On 3/29/07, Jorge Pe?a wrote: > > Thanks I really really appreciate it! I definitely worked hard on it haha, > I was at it for about two hours, all trial and error as I personally have no > exposure/experience at all to Macs. One question though, in the -L and -I > switches, you put absolute paths, is this necessary? Sorry, I didn't find > any information on this in the manual page, and I was wondering because, > seeing how it was already hard enough for me, having other people build it > might be difficult, so I was thinking about including these things in my > subversion repository, for which I'd need relevant paths, no? Or maybe > there's a variable such as: > > -I/Users/$(USER)/libogg-1.1.3/include > > Just double checked, looks like there is. No wait, better yet: > > -I/$(PWD)/code/whatever > > Seems right. I think this will definitely help, next time my friend is > available I'll SSH and keep trying, but it all makes sense now, thanks I > really, honestly appreciate your help! > > On 3/29/07, Thilo Schulz wrote: > > > > On Friday 30 March 2007 06:16, Jorge Pe?a wrote: > > > Thanks guys, sorry for the long message, but I'd really appreciate it. > > And > > > yes, I AM using the mac build script (Always have been). Thanks again, > > I > > > really appreciate it. > > > > No problem mate, I think you have made very clear that you have gone to > > some > > trouble already. You don't strike me as the casual dork wanting to get > > free > > tech support without doing work and learning on his own, so I am glad to > > > > help :) > > > > First of all, your approach was already a step in the right direction. > > I think curl is present on MacOSX already so I don't think you'll need > > to > > build it yourself. > > Now on the topic of ogg I can advise you on the steps that I took to get > > a > > working MacOSX build. > > > > Get the source to the libraries libogg, libvorbis and libvorbisfile > > (libvorbisfile is built with libvorbis). run configure > > using --enable-static --disable-shared and build all three of them this > > > > way .. if you redistribute these builds you really want static ogg > > libraries > > as users most likely won't have them installed. If you want to make a > > universal binary, you must cross compile these three libraries at least > > for > > the other platform you want to support. > > After the build process is finished, you'll find the .a files for the > > static > > libraries somewhere in the lib/.libs/ directory. I suggest you copy them > > out > > of there and put them into some collective directory as you indicated > > you > > already did. > > Then you have to edit the Makefile to indicate the include directories + > > where > > to find these libraries in the build process. > > > > This is what the relevant sections of my Makefile look like in the > > macosx > > section: > > for ppc: > > BASE_CFLAGS += -arch ppc -DSMP \ > > -I/Users/thilo/libogg-1.1.3/include \ > > -I/Users/thilo/libvorbis-1.1.2/include \ > > -DMAC_OS_X_VERSION_MIN_REQUIRED=1020 -nostdinc \ > > -F/Developer/SDKs/MacOSX10.2.8.sdk/System/Library/Frameworks \ > > ... > > > > LDFLAGS += -arch ppc \ > > -L/Users/thilo/libmerge/ppc \ > > -L/Developer/SDKs/MacOSX10.2.8.sdk/usr/lib/gcc/darwin/3.3 \ > > -F/Developer/SDKs/MacOSX10.2.8.sdk/System/Library/Frameworks \ > > > > as you can see, I added a few lines. libmerge/ppc happens to be the > > place > > where I put the ppc ogg and vorbis lib stuff into. > > > > Similarly, for i386: > > BASE_CFLAGS += -arch i386 -DSMP \ > > -mmacosx-version-min=10.4 \ > > -I/Users/thilo/libogg-1.1.3/include \ > > -I/Users/thilo/libvorbis-1.1.2/include \ > > -DMAC_OS_X_VERSION_MIN_REQUIRED=1040 -nostdinc \ > > -F/Developer/SDKs/MacOSX10.4u.sdk/System/Library/Frameworks \ > > > > -I/Developer/SDKs/MacOSX10.4u.sdk/usr/lib/gcc/i686-apple-darwin8/4.0.1/include > > -isystem /Developer/SDKs/MacOSX10.4u.sdk/usr/include > > LDFLAGS = -arch i386 -mmacosx-version-min=10.4 \ > > -L/Users/thilo/libmerge/i386 \ > > -L/Developer/SDKs/MacOSX10.4u.sdk/usr/lib/gcc/i686-apple-darwin8/4.0.1 > > \ > > -F/Developer/SDKs/MacOSX10.4u.sdk/System/Library/Frameworks \ > > -Wl,-syslibroot,/Developer/SDKs/MacOSX10.4u.sdk,-m > > > > These are only meant as examples, you will have to give the directories > > depending on your setup of course. Notice, that I added a ",-m" in the > > very > > last block compared to the original, where no -m can be found after: > > MacOSX10.4u.sdk > > This may be required if building fails and the linker complains about > > some > > double defined OpenAL symbols if I remember correctly. > > > > You will also want to defuse this line: > > # Always include debug symbols...you can strip the binary later... > > BASE_CFLAGS += -gfull > > > > by commenting it. You don't want debugging if you're not a developer or > > trying > > to hunt a bug. It will only bloat your binary and hurt speed. > > > > I hope these instructions were sufficient. Good luck! > > > > -- > > Thilo Schulz > > > > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From q3urt.undead at gmail.com Sat Mar 31 16:18:19 2007 From: q3urt.undead at gmail.com (un dead) Date: Sat, 31 Mar 2007 16:18:19 -0400 Subject: [quake3] Building On Mac In-Reply-To: <28406b400703311235w64df186flfdfa8b7cdef7344d@mail.gmail.com> References: <28406b400703292116j6708d8edn789a1e1d825d1d8d@mail.gmail.com> <200703300636.09913.arny@ats.s.bawue.de> <28406b400703292157x5cd5bb0fx281f0b823d4c8e15@mail.gmail.com> <28406b400703311235w64df186flfdfa8b7cdef7344d@mail.gmail.com> Message-ID: On 3/31/07, Jorge Pe?a wrote: > It gets a lot further into the build process now, and then it spits out the > following: > > ld: Undefined symbols: > _fprintf$LDBLStub > _sprintf$LDBLStub > _inflate > _inflateEnd > _inflateInit2_ > _inflateReset > make[1]: *** [build/release-darwin-ppc/tip.ppc] Error 1 > make: *** [build_release] Error 2 > > I google'd this and I found out it's a linker error (ld, duh). I google'd > for the symbols, and I can't find what library they correspond to, I think > it's zlib. Yeah that's from libz (zlib). Do you guys have 'nm' in MacOSX? If so, you can run: $ nm /usr/lib/libz.a | egrep "inflate(|End|Init2_|Reset)$" U inflate U inflateEnd U inflateInit2_ U inflateReset U inflate U inflateEnd 00000000000005c0 T inflate 0000000000000100 T inflateEnd 00000000000021c0 T inflateInit2_ 0000000000000000 T inflateReset From jorgepblank at gmail.com Sat Mar 31 16:28:59 2007 From: jorgepblank at gmail.com (=?ISO-8859-1?Q?Jorge_Pe=F1a?=) Date: Sat, 31 Mar 2007 12:28:59 -0800 Subject: [quake3] Building On Mac In-Reply-To: <28406b400703311235w64df186flfdfa8b7cdef7344d@mail.gmail.com> References: <28406b400703292116j6708d8edn789a1e1d825d1d8d@mail.gmail.com> <200703300636.09913.arny@ats.s.bawue.de> <28406b400703292157x5cd5bb0fx281f0b823d4c8e15@mail.gmail.com> <28406b400703311235w64df186flfdfa8b7cdef7344d@mail.gmail.com> Message-ID: <28406b400703311328x51695c7o4257e9d843b0effc@mail.gmail.com> I found this ( http://www.ipodlinux.org/forums/viewtopic.php?t=216&view=previous&sid=bde0a3071e484f13eea61ffa33c122e8) which looks like I need zlib for freetype, but I built freetype and it seemed to have compiled perfectly (freetype, I mean). By the way, I didn't do the "run configure using --enable-static --disable-shared", because I found the .a files in the lib folder anyways, is this okay? If not then I'll rebuild the things. I also tried with USE_CURL_DLOPEN and USE_OPENAL_DLOPEN set to 0, no go. Okay, I made it so it looks in /usr/local/lib and /usr/local/include now instead of my /libs/macosx directory. I configured all the paths and they should be working now, but now I'm getting mismatching architecture reports, having to do with the dylib, so I'm recompiling the libraries with --enable-shared=no and --enable-static=yes, hopefully that does it. Thanks un dead, but now I know that it's from zlib :D On 3/31/07, Jorge Pe?a wrote: > > Hey sorry to keep bothering, but here are the things I added (The relevant > stuff at least): > > For PPC: > > BASE_CFLAGS += -arch ppc -DSMP \ > -I$(PWD)/code \ > -I$(PWD)/code/vorbis/include \ > -I$(PWD)/code/freetype \ > -DMAC_OS_X_VERSION_MIN_REQUIRED=1020 -nostdinc \ > -F/Developer/SDKs/MacOSX10.2.8.sdk/System/Library/Frameworks \ > -I/Developer/SDKs/MacOSX10.2.8.sdk/usr/include/gcc/darwin/3.3 \ > -isystem /Developer/SDKs/MacOSX10.2.8.sdk/usr/include > > LDFLAGS += -arch ppc \ > -L$(PWD)/code/libs/macosx \ > -L/Developer/SDKs/MacOSX10.2.8.sdk/usr/lib/gcc/darwin/3.3 \ > -F/Developer/SDKs/MacOSX10.2.8.sdk/System/Library/Frameworks \ > -Wl,-syslibroot,/Developer/SDKs/MacOSX10.2.8.sdk,-m > ARCH=ppc > > I did the same for i386, I didn't show it though because I'm actually > building on PPC at the moment. Anyways, I also commented out the line: > > # Always include debug symbols...you can strip the binary later... > #BASE_CFLAGS += -gfull > > It gets a lot further into the build process now, and then it spits out > the following: > > ld: Undefined symbols: > _fprintf$LDBLStub > _sprintf$LDBLStub > _inflate > _inflateEnd > _inflateInit2_ > _inflateReset > make[1]: *** [build/release-darwin-ppc/tip.ppc] Error 1 > make: *** [build_release] Error 2 > > I google'd this and I found out it's a linker error (ld, duh). I google'd > for the symbols, and I can't find what library they correspond to, I thinkit's zlib. Does anything make use of zlib from the following: libogg, > libvorbis, libfreetype, libcurl? I'm just wondering, because before this, > the engine built perfectly. But come to think of it, it probably IS a zlib > dependency, I'll go and get the library. Actually, they said the library > should already be in there, it wasn't there so I downloaded the tarball and > built it myself. I then put the resulting static library ( libz.a) in the > macosx libs directory which has already been -L'd by the makefile, I ran the > bash make script and it still gave me the error. So I edited the Makefile > and added a -lz to a random part (To a part where it does -lvorbisfile as > well), since I'm building with OGG I figured It'd pick it up, but it's still > giving me the error. > > Anyways, I'm just curious because I didn't find any reference to zlib in > the Makefile and I only found a reference in the README, but it only > includes header files and source files of it, in other platforms like > Windows, I don't have to go and download it and build it along with the > game. > > I'm just wondering if you guys know anything about this, I've been > investigating since yesterday I was up till 12 haha, I think I'm getting > closer to finally being able to build though, I think it's just this that's > left. > > On 3/29/07, Jorge Pe?a wrote: > > > > Thanks I really really appreciate it! I definitely worked hard on it > > haha, I was at it for about two hours, all trial and error as I personally > > have no exposure/experience at all to Macs. One question though, in the -L > > and -I switches, you put absolute paths, is this necessary? Sorry, I didn't > > find any information on this in the manual page, and I was wondering > > because, seeing how it was already hard enough for me, having other people > > build it might be difficult, so I was thinking about including these things > > in my subversion repository, for which I'd need relevant paths, no? Or maybe > > there's a variable such as: > > > > -I/Users/$(USER)/libogg-1.1.3/include > > > > Just double checked, looks like there is. No wait, better yet: > > > > -I/$(PWD)/code/whatever > > > > Seems right. I think this will definitely help, next time my friend is > > available I'll SSH and keep trying, but it all makes sense now, thanks I > > really, honestly appreciate your help! > > > > On 3/29/07, Thilo Schulz < arny at ats.s.bawue.de> wrote: > > > > > > On Friday 30 March 2007 06:16, Jorge Pe?a wrote: > > > > Thanks guys, sorry for the long message, but I'd really appreciate > > > it. And > > > > yes, I AM using the mac build script (Always have been). Thanks > > > again, I > > > > really appreciate it. > > > > > > No problem mate, I think you have made very clear that you have gone > > > to some > > > trouble already. You don't strike me as the casual dork wanting to get > > > free > > > tech support without doing work and learning on his own, so I am glad > > > to > > > help :) > > > > > > First of all, your approach was already a step in the right direction. > > > I think curl is present on MacOSX already so I don't think you'll need > > > to > > > build it yourself. > > > Now on the topic of ogg I can advise you on the steps that I took to > > > get a > > > working MacOSX build. > > > > > > Get the source to the libraries libogg, libvorbis and libvorbisfile > > > (libvorbisfile is built with libvorbis). run configure > > > using --enable-static --disable-shared and build all three of them > > > this > > > way .. if you redistribute these builds you really want static ogg > > > libraries > > > as users most likely won't have them installed. If you want to make a > > > universal binary, you must cross compile these three libraries at > > > least for > > > the other platform you want to support. > > > After the build process is finished, you'll find the .a files for the > > > static > > > libraries somewhere in the lib/.libs/ directory. I suggest you copy > > > them out > > > of there and put them into some collective directory as you indicated > > > you > > > already did. > > > Then you have to edit the Makefile to indicate the include directories > > > + where > > > to find these libraries in the build process. > > > > > > This is what the relevant sections of my Makefile look like in the > > > macosx > > > section: > > > for ppc: > > > BASE_CFLAGS += -arch ppc -DSMP \ > > > -I/Users/thilo/libogg-1.1.3/include \ > > > -I/Users/thilo/libvorbis-1.1.2/include \ > > > -DMAC_OS_X_VERSION_MIN_REQUIRED=1020 -nostdinc \ > > > -F/Developer/SDKs/MacOSX10.2.8.sdk/System/Library/Frameworks \ > > > ... > > > > > > LDFLAGS += -arch ppc \ > > > -L/Users/thilo/libmerge/ppc \ > > > -L/Developer/SDKs/MacOSX10.2.8.sdk/usr/lib/gcc/darwin/3.3 \ > > > -F/Developer/SDKs/MacOSX10.2.8.sdk/System/Library/Frameworks \ > > > > > > as you can see, I added a few lines. libmerge/ppc happens to be the > > > place > > > where I put the ppc ogg and vorbis lib stuff into. > > > > > > Similarly, for i386: > > > BASE_CFLAGS += -arch i386 -DSMP \ > > > -mmacosx-version-min=10.4 \ > > > -I/Users/thilo/libogg-1.1.3/include \ > > > -I/Users/thilo/libvorbis-1.1.2/include \ > > > -DMAC_OS_X_VERSION_MIN_REQUIRED=1040 -nostdinc \ > > > -F/Developer/SDKs/MacOSX10.4u.sdk/System/Library/Frameworks \ > > > > > > -I/Developer/SDKs/MacOSX10.4u.sdk/usr/lib/gcc/i686-apple-darwin8/4.0.1/include > > > -isystem /Developer/SDKs/MacOSX10.4u.sdk/usr/include > > > LDFLAGS = -arch i386 -mmacosx-version-min=10.4 \ > > > -L/Users/thilo/libmerge/i386 \ > > > -L/Developer/SDKs/MacOSX10.4u.sdk/usr/lib/gcc/i686-apple-darwin8/4.0.1 > > > \ > > > -F/Developer/SDKs/MacOSX10.4u.sdk/System/Library/Frameworks \ > > > -Wl,-syslibroot,/Developer/SDKs/MacOSX10.4u.sdk,-m > > > > > > These are only meant as examples, you will have to give the > > > directories > > > depending on your setup of course. Notice, that I added a ",-m" in the > > > very > > > last block compared to the original, where no -m can be found after: > > > MacOSX10.4u.sdk > > > This may be required if building fails and the linker complains about > > > some > > > double defined OpenAL symbols if I remember correctly. > > > > > > You will also want to defuse this line: > > > # Always include debug symbols...you can strip the binary later... > > > BASE_CFLAGS += -gfull > > > > > > by commenting it. You don't want debugging if you're not a developer > > > or trying > > > to hunt a bug. It will only bloat your binary and hurt speed. > > > > > > I hope these instructions were sufficient. Good luck! > > > > > > -- > > > Thilo Schulz > > > > > > > > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From jorgepblank at gmail.com Sat Mar 31 16:31:55 2007 From: jorgepblank at gmail.com (=?ISO-8859-1?Q?Jorge_Pe=F1a?=) Date: Sat, 31 Mar 2007 12:31:55 -0800 Subject: [quake3] Building On Mac In-Reply-To: <28406b400703311328x51695c7o4257e9d843b0effc@mail.gmail.com> References: <28406b400703292116j6708d8edn789a1e1d825d1d8d@mail.gmail.com> <200703300636.09913.arny@ats.s.bawue.de> <28406b400703292157x5cd5bb0fx281f0b823d4c8e15@mail.gmail.com> <28406b400703311235w64df186flfdfa8b7cdef7344d@mail.gmail.com> <28406b400703311328x51695c7o4257e9d843b0effc@mail.gmail.com> Message-ID: <28406b400703311331u6fe247f4w5c3b987b9a912c55@mail.gmail.com> Yeah, sorry for the long following output, but it's what I'm getting: /usr/bin/ld: warning /usr/local/lib/libcurl.dylib cputype (18, architecture ppc) does not match cputype (7) for specified -arch flag: i386 (file not loaded) /usr/bin/ld: warning /usr/local/lib/libvorbisfile.dylib cputype (18, architecture ppc) does not match cputype (7) for specified -arch flag: i386 (file not loaded) /usr/bin/ld: warning /usr/local/lib/libvorbis.dylib cputype (18, architecture ppc) does not match cputype (7) for specified -arch flag: i386 (file not loaded) /usr/bin/ld: warning /usr/local/lib/libogg.dylib cputype (18, architecture ppc) does not match cputype (7) for specified -arch flag: i386 (file not loaded) /usr/bin/ld: warning /usr/local/lib/libfreetype.dylib cputype (18, architecture ppc) does not match cputype (7) for specified -arch flag: i386 (file not loaded) /usr/bin/ld: Undefined symbols: _ov_clear _ov_info _ov_open_callbacks _ov_pcm_total _ov_read _ov_seekable _ov_streams _curl_easy_cleanup _curl_easy_getinfo _curl_easy_init _curl_easy_setopt _curl_easy_strerror _curl_multi_add_handle _curl_multi_cleanup _curl_multi_info_read _curl_multi_init _curl_multi_perform _curl_multi_remove_handle _curl_version _FT_Done_FreeType _FT_Get_Char_Index _FT_Init_FreeType _FT_Load_Glyph _FT_New_Memory_Face _FT_Outline_Get_Bitmap _FT_Outline_Translate _FT_Set_Char_Size collect2: ld returned 1 exit status make[1]: *** [build/release-darwin-i386/tip.i386] Error 1 make: *** [build_release] Error 2 It apparently has to do with the i386 things, and I built for ppc, you guys know how I can build the libraries (libcurl and all that) for i386? If not, I'll try and disable building for i386 at the moment until I get it solved. On 3/31/07, Jorge Pe?a wrote: > > I found this ( http://www.ipodlinux.org/forums/viewtopic.php?t=216&view=previous&sid=bde0a3071e484f13eea61ffa33c122e8 > ) which looks like I need zlib for freetype, but I built freetype and it > seemed to have compiled perfectly (freetype, I mean). By the way, I didn't > do the "run configure using --enable-static --disable-shared", because I > found the .a files in the lib folder anyways, is this okay? If not then I'll > rebuild the things. I also tried with USE_CURL_DLOPEN and USE_OPENAL_DLOPEN > set to 0, no go. > > Okay, I made it so it looks in /usr/local/lib and /usr/local/include now > instead of my /libs/macosx directory. I configured all the paths and they > should be working now, but now I'm getting mismatching architecture reports, > having to do with the dylib, so I'm recompiling the libraries with > --enable-shared=no and --enable-static=yes, hopefully that does it. > > Thanks un dead, but now I know that it's from zlib :D > > On 3/31/07, Jorge Pe?a wrote: > > > > Hey sorry to keep bothering, but here are the things I added (The > > relevant stuff at least): > > > > For PPC: > > > > BASE_CFLAGS += -arch ppc -DSMP \ > > -I$(PWD)/code \ > > -I$(PWD)/code/vorbis/include \ > > -I$(PWD)/code/freetype \ > > -DMAC_OS_X_VERSION_MIN_REQUIRED=1020 -nostdinc \ > > -F/Developer/SDKs/MacOSX10.2.8.sdk/System/Library/Frameworks \ > > -I/Developer/SDKs/MacOSX10.2.8.sdk/usr/include/gcc/darwin/3.3 \ > > -isystem /Developer/SDKs/MacOSX10.2.8.sdk/usr/include > > > > LDFLAGS += -arch ppc \ > > -L$(PWD)/code/libs/macosx \ > > -L/Developer/SDKs/MacOSX10.2.8.sdk/usr/lib/gcc/darwin/3.3 \ > > -F/Developer/SDKs/MacOSX10.2.8.sdk/System/Library/Frameworks \ > > -Wl,-syslibroot,/Developer/SDKs/MacOSX10.2.8.sdk,-m > > ARCH=ppc > > > > I did the same for i386, I didn't show it though because I'm actually > > building on PPC at the moment. Anyways, I also commented out the line: > > > > # Always include debug symbols...you can strip the binary later... > > #BASE_CFLAGS += -gfull > > > > It gets a lot further into the build process now, and then it spits out > > the following: > > > > ld: Undefined symbols: > > _fprintf$LDBLStub > > _sprintf$LDBLStub > > _inflate > > _inflateEnd > > _inflateInit2_ > > _inflateReset > > make[1]: *** [build/release-darwin-ppc/tip.ppc] Error 1 > > make: *** [build_release] Error 2 > > > > I google'd this and I found out it's a linker error (ld, duh). I > > google'd for the symbols, and I can't find what library they correspond to, > > I think it's zlib. Does anything make use of zlib from the following: > > libogg, libvorbis, libfreetype, libcurl? I'm just wondering, because before > > this, the engine built perfectly. But come to think of it, it probably IS a > > zlib dependency, I'll go and get the library. Actually, they said the > > library should already be in there, it wasn't there so I downloaded the > > tarball and built it myself. I then put the resulting static library ( > > libz.a) in the macosx libs directory which has already been -L'd by the > > makefile, I ran the bash make script and it still gave me the error. So I > > edited the Makefile and added a -lz to a random part (To a part where it > > does -lvorbisfile as well), since I'm building with OGG I figured It'd pick > > it up, but it's still giving me the error. > > > > Anyways, I'm just curious because I didn't find any reference to zlib in > > the Makefile and I only found a reference in the README, but it only > > includes header files and source files of it, in other platforms like > > Windows, I don't have to go and download it and build it along with the > > game. > > > > I'm just wondering if you guys know anything about this, I've been > > investigating since yesterday I was up till 12 haha, I think I'm getting > > closer to finally being able to build though, I think it's just this that's > > left. > > > > On 3/29/07, Jorge Pe?a < jorgepblank at gmail.com> wrote: > > > > > > Thanks I really really appreciate it! I definitely worked hard on it > > > haha, I was at it for about two hours, all trial and error as I personally > > > have no exposure/experience at all to Macs. One question though, in the -L > > > and -I switches, you put absolute paths, is this necessary? Sorry, I didn't > > > find any information on this in the manual page, and I was wondering > > > because, seeing how it was already hard enough for me, having other people > > > build it might be difficult, so I was thinking about including these things > > > in my subversion repository, for which I'd need relevant paths, no? Or maybe > > > there's a variable such as: > > > > > > -I/Users/$(USER)/libogg-1.1.3/include > > > > > > Just double checked, looks like there is. No wait, better yet: > > > > > > -I/$(PWD)/code/whatever > > > > > > Seems right. I think this will definitely help, next time my friend is > > > available I'll SSH and keep trying, but it all makes sense now, thanks I > > > really, honestly appreciate your help! > > > > > > On 3/29/07, Thilo Schulz < arny at ats.s.bawue.de> wrote: > > > > > > > > On Friday 30 March 2007 06:16, Jorge Pe?a wrote: > > > > > Thanks guys, sorry for the long message, but I'd really appreciate > > > > it. And > > > > > yes, I AM using the mac build script (Always have been). Thanks > > > > again, I > > > > > really appreciate it. > > > > > > > > No problem mate, I think you have made very clear that you have gone > > > > to some > > > > trouble already. You don't strike me as the casual dork wanting to > > > > get free > > > > tech support without doing work and learning on his own, so I am > > > > glad to > > > > help :) > > > > > > > > First of all, your approach was already a step in the right > > > > direction. > > > > I think curl is present on MacOSX already so I don't think you'll > > > > need to > > > > build it yourself. > > > > Now on the topic of ogg I can advise you on the steps that I took to > > > > get a > > > > working MacOSX build. > > > > > > > > Get the source to the libraries libogg, libvorbis and libvorbisfile > > > > (libvorbisfile is built with libvorbis). run configure > > > > using --enable-static --disable-shared and build all three of them > > > > this > > > > way .. if you redistribute these builds you really want static ogg > > > > libraries > > > > as users most likely won't have them installed. If you want to make > > > > a > > > > universal binary, you must cross compile these three libraries at > > > > least for > > > > the other platform you want to support. > > > > After the build process is finished, you'll find the .a files for > > > > the static > > > > libraries somewhere in the lib/.libs/ directory. I suggest you copy > > > > them out > > > > of there and put them into some collective directory as you > > > > indicated you > > > > already did. > > > > Then you have to edit the Makefile to indicate the include > > > > directories + where > > > > to find these libraries in the build process. > > > > > > > > This is what the relevant sections of my Makefile look like in the > > > > macosx > > > > section: > > > > for ppc: > > > > BASE_CFLAGS += -arch ppc -DSMP \ > > > > -I/Users/thilo/libogg-1.1.3/include \ > > > > -I/Users/thilo/libvorbis-1.1.2/include \ > > > > -DMAC_OS_X_VERSION_MIN_REQUIRED=1020 -nostdinc \ > > > > -F/Developer/SDKs/MacOSX10.2.8.sdk/System/Library/Frameworks \ > > > > ... > > > > > > > > LDFLAGS += -arch ppc \ > > > > -L/Users/thilo/libmerge/ppc \ > > > > -L/Developer/SDKs/MacOSX10.2.8.sdk/usr/lib/gcc/darwin/3.3 \ > > > > -F/Developer/SDKs/MacOSX10.2.8.sdk/System/Library/Frameworks \ > > > > > > > > as you can see, I added a few lines. libmerge/ppc happens to be the > > > > place > > > > where I put the ppc ogg and vorbis lib stuff into. > > > > > > > > Similarly, for i386: > > > > BASE_CFLAGS += -arch i386 -DSMP \ > > > > -mmacosx-version-min=10.4 \ > > > > -I/Users/thilo/libogg-1.1.3/include \ > > > > -I/Users/thilo/libvorbis-1.1.2/include \ > > > > -DMAC_OS_X_VERSION_MIN_REQUIRED=1040 -nostdinc \ > > > > -F/Developer/SDKs/MacOSX10.4u.sdk/System/Library/Frameworks \ > > > > > > > > -I/Developer/SDKs/MacOSX10.4u.sdk/usr/lib/gcc/i686-apple-darwin8/4.0.1/include > > > > -isystem /Developer/SDKs/MacOSX10.4u.sdk/usr/include > > > > LDFLAGS = -arch i386 -mmacosx-version-min=10.4 \ > > > > -L/Users/thilo/libmerge/i386 \ > > > > -L/Developer/SDKs/MacOSX10.4u.sdk/usr/lib/gcc/i686-apple-darwin8/4.0.1 > > > > \ > > > > -F/Developer/SDKs/MacOSX10.4u.sdk/System/Library/Frameworks \ > > > > -Wl,-syslibroot,/Developer/SDKs/MacOSX10.4u.sdk,-m > > > > > > > > These are only meant as examples, you will have to give the > > > > directories > > > > depending on your setup of course. Notice, that I added a ",-m" in > > > > the very > > > > last block compared to the original, where no -m can be found after: > > > > MacOSX10.4u.sdk > > > > This may be required if building fails and the linker complains > > > > about some > > > > double defined OpenAL symbols if I remember correctly. > > > > > > > > You will also want to defuse this line: > > > > # Always include debug symbols...you can strip the binary later... > > > > BASE_CFLAGS += -gfull > > > > > > > > by commenting it. You don't want debugging if you're not a developer > > > > or trying > > > > to hunt a bug. It will only bloat your binary and hurt speed. > > > > > > > > I hope these instructions were sufficient. Good luck! > > > > > > > > -- > > > > Thilo Schulz > > > > > > > > > > > > > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From jorgepblank at gmail.com Sat Mar 31 16:41:12 2007 From: jorgepblank at gmail.com (=?ISO-8859-1?Q?Jorge_Pe=F1a?=) Date: Sat, 31 Mar 2007 12:41:12 -0800 Subject: [quake3] Building On Mac In-Reply-To: <28406b400703311331u6fe247f4w5c3b987b9a912c55@mail.gmail.com> References: <28406b400703292116j6708d8edn789a1e1d825d1d8d@mail.gmail.com> <200703300636.09913.arny@ats.s.bawue.de> <28406b400703292157x5cd5bb0fx281f0b823d4c8e15@mail.gmail.com> <28406b400703311235w64df186flfdfa8b7cdef7344d@mail.gmail.com> <28406b400703311328x51695c7o4257e9d843b0effc@mail.gmail.com> <28406b400703311331u6fe247f4w5c3b987b9a912c55@mail.gmail.com> Message-ID: <28406b400703311341k66e64ef5h64f4ecd4a15aeed7@mail.gmail.com> Perfect, it builds now. I had to remove the i386 make instruction from the mac make bash script, then I removed all references to i386 to avoid unnecessary problems and it works perfectly now. So I can't build i386 things on mac? I have a feeling I can, I just need the i386 versions of the libraries right (Like libcurl, libfreetype, etc.)? Anyone know how I can go about doing this? If it's not possible then I'm fine, if demand for i386 comes someone can always just volunteer like someone did for this build. Thanks guys! Sorry for bothering! On 3/31/07, Jorge Pe?a wrote: > > Yeah, sorry for the long following output, but it's what I'm getting: > > /usr/bin/ld: warning /usr/local/lib/libcurl.dylib cputype (18, > architecture ppc) does not match cputype (7) for specified -arch flag: i386 > (file not loaded) > /usr/bin/ld: warning /usr/local/lib/libvorbisfile.dylib cputype (18, > architecture ppc) does not match cputype (7) for specified -arch flag: i386 > (file not loaded) > /usr/bin/ld: warning /usr/local/lib/libvorbis.dylib cputype (18, > architecture ppc) does not match cputype (7) for specified -arch flag: i386 > (file not loaded) > /usr/bin/ld: warning /usr/local/lib/libogg.dylib cputype (18, architecture > ppc) does not match cputype (7) for specified -arch flag: i386 (file not > loaded) > /usr/bin/ld: warning /usr/local/lib/libfreetype.dylib cputype (18, > architecture ppc) does not match cputype (7) for specified -arch flag: i386 > (file not loaded) > /usr/bin/ld: Undefined symbols: > _ov_clear > _ov_info > _ov_open_callbacks > _ov_pcm_total > _ov_read > _ov_seekable > _ov_streams > _curl_easy_cleanup > _curl_easy_getinfo > _curl_easy_init > _curl_easy_setopt > _curl_easy_strerror > _curl_multi_add_handle > _curl_multi_cleanup > _curl_multi_info_read > _curl_multi_init > _curl_multi_perform > _curl_multi_remove_handle > _curl_version > _FT_Done_FreeType > _FT_Get_Char_Index > _FT_Init_FreeType > _FT_Load_Glyph > _FT_New_Memory_Face > _FT_Outline_Get_Bitmap > _FT_Outline_Translate > _FT_Set_Char_Size > collect2: ld returned 1 exit status > make[1]: *** [build/release-darwin-i386/tip.i386] Error 1 > make: *** [build_release] Error 2 > > It apparently has to do with the i386 things, and I built for ppc, you > guys know how I can build the libraries (libcurl and all that) for i386? If > not, I'll try and disable building for i386 at the moment until I get it > solved. > > On 3/31/07, Jorge Pe?a wrote: > > > > I found this ( http://www.ipodlinux.org/forums/viewtopic.php?t=216&view=previous&sid=bde0a3071e484f13eea61ffa33c122e8 > > ) which looks like I need zlib for freetype, but I built freetype and it > > seemed to have compiled perfectly (freetype, I mean). By the way, I didn't > > do the "run configure using --enable-static --disable-shared", because I > > found the .a files in the lib folder anyways, is this okay? If not then I'll > > rebuild the things. I also tried with USE_CURL_DLOPEN and USE_OPENAL_DLOPEN > > set to 0, no go. > > > > Okay, I made it so it looks in /usr/local/lib and /usr/local/include now > > instead of my /libs/macosx directory. I configured all the paths and they > > should be working now, but now I'm getting mismatching architecture reports, > > having to do with the dylib, so I'm recompiling the libraries with > > --enable-shared=no and --enable-static=yes, hopefully that does it. > > > > Thanks un dead, but now I know that it's from zlib :D > > > > On 3/31/07, Jorge Pe?a < jorgepblank at gmail.com > wrote: > > > > > > Hey sorry to keep bothering, but here are the things I added (The > > > relevant stuff at least): > > > > > > For PPC: > > > > > > BASE_CFLAGS += -arch ppc -DSMP \ > > > -I$(PWD)/code \ > > > -I$(PWD)/code/vorbis/include \ > > > -I$(PWD)/code/freetype \ > > > -DMAC_OS_X_VERSION_MIN_REQUIRED=1020 -nostdinc \ > > > -F/Developer/SDKs/MacOSX10.2.8.sdk/System/Library/Frameworks \ > > > -I/Developer/SDKs/MacOSX10.2.8.sdk/usr/include/gcc/darwin/3.3 \ > > > -isystem /Developer/SDKs/MacOSX10.2.8.sdk/usr/include > > > > > > LDFLAGS += -arch ppc \ > > > -L$(PWD)/code/libs/macosx \ > > > -L/Developer/SDKs/MacOSX10.2.8.sdk/usr/lib/gcc/darwin/3.3 \ > > > -F/Developer/SDKs/MacOSX10.2.8.sdk/System/Library/Frameworks \ > > > -Wl,-syslibroot,/Developer/SDKs/MacOSX10.2.8.sdk,-m > > > ARCH=ppc > > > > > > I did the same for i386, I didn't show it though because I'm actually > > > building on PPC at the moment. Anyways, I also commented out the line: > > > > > > # Always include debug symbols...you can strip the binary later... > > > #BASE_CFLAGS += -gfull > > > > > > It gets a lot further into the build process now, and then it spits > > > out the following: > > > > > > ld: Undefined symbols: > > > _fprintf$LDBLStub > > > _sprintf$LDBLStub > > > _inflate > > > _inflateEnd > > > _inflateInit2_ > > > _inflateReset > > > make[1]: *** [build/release-darwin-ppc/tip.ppc] Error 1 > > > make: *** [build_release] Error 2 > > > > > > I google'd this and I found out it's a linker error (ld, duh). I > > > google'd for the symbols, and I can't find what library they correspond to, > > > I think it's zlib. Does anything make use of zlib from the following: > > > libogg, libvorbis, libfreetype, libcurl? I'm just wondering, because before > > > this, the engine built perfectly. But come to think of it, it probably IS a > > > zlib dependency, I'll go and get the library. Actually, they said the > > > library should already be in there, it wasn't there so I downloaded the > > > tarball and built it myself. I then put the resulting static library ( > > > libz.a) in the macosx libs directory which has already been -L'd by > > > the makefile, I ran the bash make script and it still gave me the error. So > > > I edited the Makefile and added a -lz to a random part (To a part where it > > > does -lvorbisfile as well), since I'm building with OGG I figured It'd pick > > > it up, but it's still giving me the error. > > > > > > Anyways, I'm just curious because I didn't find any reference to zlib > > > in the Makefile and I only found a reference in the README, but it only > > > includes header files and source files of it, in other platforms like > > > Windows, I don't have to go and download it and build it along with the > > > game. > > > > > > I'm just wondering if you guys know anything about this, I've been > > > investigating since yesterday I was up till 12 haha, I think I'm getting > > > closer to finally being able to build though, I think it's just this that's > > > left. > > > > > > On 3/29/07, Jorge Pe?a < jorgepblank at gmail.com> wrote: > > > > > > > > Thanks I really really appreciate it! I definitely worked hard on it > > > > haha, I was at it for about two hours, all trial and error as I personally > > > > have no exposure/experience at all to Macs. One question though, in the -L > > > > and -I switches, you put absolute paths, is this necessary? Sorry, I didn't > > > > find any information on this in the manual page, and I was wondering > > > > because, seeing how it was already hard enough for me, having other people > > > > build it might be difficult, so I was thinking about including these things > > > > in my subversion repository, for which I'd need relevant paths, no? Or maybe > > > > there's a variable such as: > > > > > > > > -I/Users/$(USER)/libogg-1.1.3/include > > > > > > > > Just double checked, looks like there is. No wait, better yet: > > > > > > > > -I/$(PWD)/code/whatever > > > > > > > > Seems right. I think this will definitely help, next time my friend > > > > is available I'll SSH and keep trying, but it all makes sense now, thanks I > > > > really, honestly appreciate your help! > > > > > > > > On 3/29/07, Thilo Schulz < arny at ats.s.bawue.de> wrote: > > > > > > > > > > On Friday 30 March 2007 06:16, Jorge Pe?a wrote: > > > > > > Thanks guys, sorry for the long message, but I'd really > > > > > appreciate it. And > > > > > > yes, I AM using the mac build script (Always have been). Thanks > > > > > again, I > > > > > > really appreciate it. > > > > > > > > > > No problem mate, I think you have made very clear that you have > > > > > gone to some > > > > > trouble already. You don't strike me as the casual dork wanting to > > > > > get free > > > > > tech support without doing work and learning on his own, so I am > > > > > glad to > > > > > help :) > > > > > > > > > > First of all, your approach was already a step in the right > > > > > direction. > > > > > I think curl is present on MacOSX already so I don't think you'll > > > > > need to > > > > > build it yourself. > > > > > Now on the topic of ogg I can advise you on the steps that I took > > > > > to get a > > > > > working MacOSX build. > > > > > > > > > > Get the source to the libraries libogg, libvorbis and > > > > > libvorbisfile > > > > > (libvorbisfile is built with libvorbis). run configure > > > > > using --enable-static --disable-shared and build all three of > > > > > them this > > > > > way .. if you redistribute these builds you really want static ogg > > > > > libraries > > > > > as users most likely won't have them installed. If you want to > > > > > make a > > > > > universal binary, you must cross compile these three libraries at > > > > > least for > > > > > the other platform you want to support. > > > > > After the build process is finished, you'll find the .a files for > > > > > the static > > > > > libraries somewhere in the lib/.libs/ directory. I suggest you > > > > > copy them out > > > > > of there and put them into some collective directory as you > > > > > indicated you > > > > > already did. > > > > > Then you have to edit the Makefile to indicate the include > > > > > directories + where > > > > > to find these libraries in the build process. > > > > > > > > > > This is what the relevant sections of my Makefile look like in the > > > > > macosx > > > > > section: > > > > > for ppc: > > > > > BASE_CFLAGS += -arch ppc -DSMP \ > > > > > -I/Users/thilo/libogg-1.1.3/include \ > > > > > -I/Users/thilo/libvorbis-1.1.2/include \ > > > > > -DMAC_OS_X_VERSION_MIN_REQUIRED=1020 -nostdinc \ > > > > > -F/Developer/SDKs/MacOSX10.2.8.sdk/System/Library/Frameworks > > > > > \ > > > > > ... > > > > > > > > > > LDFLAGS += -arch ppc \ > > > > > -L/Users/thilo/libmerge/ppc \ > > > > > -L/Developer/SDKs/MacOSX10.2.8.sdk/usr/lib/gcc/darwin/3.3 \ > > > > > -F/Developer/SDKs/MacOSX10.2.8.sdk/System/Library/Frameworks > > > > > \ > > > > > > > > > > as you can see, I added a few lines. libmerge/ppc happens to be > > > > > the place > > > > > where I put the ppc ogg and vorbis lib stuff into. > > > > > > > > > > Similarly, for i386: > > > > > BASE_CFLAGS += -arch i386 -DSMP \ > > > > > -mmacosx-version-min=10.4 \ > > > > > -I/Users/thilo/libogg-1.1.3/include \ > > > > > -I/Users/thilo/libvorbis-1.1.2/include \ > > > > > -DMAC_OS_X_VERSION_MIN_REQUIRED=1040 -nostdinc \ > > > > > -F/Developer/SDKs/MacOSX10.4u.sdk/System/Library/Frameworks > > > > > \ > > > > > > > > > > -I/Developer/SDKs/MacOSX10.4u.sdk/usr/lib/gcc/i686-apple-darwin8/4.0.1/include > > > > > -isystem /Developer/SDKs/MacOSX10.4u.sdk/usr/include > > > > > LDFLAGS = -arch i386 -mmacosx-version-min=10.4 \ > > > > > -L/Users/thilo/libmerge/i386 \ > > > > > -L/Developer/SDKs/MacOSX10.4u.sdk/usr/lib/gcc/i686-apple-darwin8/4.0.1 > > > > > \ > > > > > -F/Developer/SDKs/MacOSX10.4u.sdk/System/Library/Frameworks > > > > > \ > > > > > -Wl,-syslibroot,/Developer/SDKs/MacOSX10.4u.sdk,-m > > > > > > > > > > These are only meant as examples, you will have to give the > > > > > directories > > > > > depending on your setup of course. Notice, that I added a ",-m" in > > > > > the very > > > > > last block compared to the original, where no -m can be found > > > > > after: > > > > > MacOSX10.4u.sdk > > > > > This may be required if building fails and the linker complains > > > > > about some > > > > > double defined OpenAL symbols if I remember correctly. > > > > > > > > > > You will also want to defuse this line: > > > > > # Always include debug symbols...you can strip the binary later... > > > > > BASE_CFLAGS += -gfull > > > > > > > > > > by commenting it. You don't want debugging if you're not a > > > > > developer or trying > > > > > to hunt a bug. It will only bloat your binary and hurt speed. > > > > > > > > > > I hope these instructions were sufficient. Good luck! > > > > > > > > > > -- > > > > > Thilo Schulz > > > > > > > > > > > > > > > > > > > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From arny at ats.s.bawue.de Sat Mar 31 17:01:39 2007 From: arny at ats.s.bawue.de (Thilo Schulz) Date: Sat, 31 Mar 2007 23:01:39 +0200 Subject: [quake3] Re: cl_consoleHistory stores passwords in q3config.cfg In-Reply-To: References: <20070331195323.a629e879.tim@ngus.net> Message-ID: <200703312301.42599.arny@ats.s.bawue.de> On Saturday 31 March 2007 21:32, un dead wrote: > I see people on forums regularly post their configs or send configs to > other people. There's no intuition for most people that it could lead > to exposing your passwords. I very much agree.. your concern is valid. This must be resolved one way or another. -- Thilo Schulz -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available URL: