0000291: Release a more up-to-date source release.

Notes
(0001063) Torr Samaho (administrator) 2011-02-13 18:00	We planned to release the source of 97c3 a while ago, but for some reason we didn't get around doing this yet. I guess we should try to get back on track with this soon. One thing that is probably worth discussing beforehand is the license though: On the one hand, I don't want to force derived works to go open source, but on the other hand, I'd like Skulltag to be able to benefit from works derived form its source. The ZDoom license ensures the former, but doesn't do anything about the latter. > you could also potentially update to the latest GZDoom renderer, which would then let the genie out of the bottle and prevent people after you from closing the source again. If I wanted to ensure that the source I release stays open source, I'd use a proper license designed for this (e.g. CDDL) and not something that was created with the sole purpose of punishing Skulltag for a dumb statement Aabra made on Doomworld that even none of the ST devs agreed with :P. > And for those of you worried about security, read this:'http://www.altdeath.com/fluxbb/viewtopic.php?pid=231#p231 [^]' I didn't follow the link in the linked post, but even without reading this: Can anybody really think that having the source available doesn't make hacking easier? In particular, when completely open source, any kind of hack detection is pretty much impossible since the hacker can just easily check how the hack detection works. Of course you could say, make this kind of stuff closed source, but then you are having the closed source problem again, not to mention that you have to maintain the closed part separately form the rest which adds a considerable maintenance overhead. Actually, I'd like to know how many of the players would be willing to accept more hacks as price for going open source. I personally possibly could be convinced, but it wouldn't make any sense to release the latest source, if this turned Skulltag into a ghost town.

(0001064) AlexMax (developer) 2011-02-13 19:39 edited on: 2011-02-13 19:46	"We planned to release the source of 97c3 a while ago, but for some reason we didn't get around doing this yet. I guess we should try to get back on track with this soon." Would perhaps a more up-to-date version be more palatable? Perhaps 97d5? - - - "One thing that is probably worth discussing beforehand is the license though: On the one hand, I don't want to force derived works to go open source, but on the other hand, I'd like Skulltag to be able to benefit from works derived form its source. The ZDoom license ensures the former, but doesn't do anything about the latter." That's the thing though, I don't know how you can guarantee one without the other. If derived works go closed source and the admins are uncooperative, then I don't really see how you can have one without the other. - - - "If I wanted to ensure that the source I release stays open source, I'd use a proper license designed for this (e.g. CDDL) and not something that was created with the sole purpose of punishing Skulltag for a dumb statement Aabra made on Doomworld that even none of the ST devs agreed with :P." I don't want to speak for Graf, but I suspect that the main reason why he hasn't put more of his contributions under a copyleft license is because he knows that Skulltag, in spite of being closed source, is quite popular, and would rather popular child ports use his contributions instead of being a complete hard-ass about it. If Skulltag goes open source, I wouldn't be surprised to see him put more of his contributions under some similar copyleft license. - - - "Actually, I'd like to know how many of the players would be willing to accept more hacks as price for going open source. I personally possibly could be convinced, but it wouldn't make any sense to release the latest source, if this turned Skulltag into a ghost town." See, that's the crux of it for me. There are already skulltag hacks floating around, and we already have at least one, possibly more people who are already writing hacks for Skulltag that are kept up to date with each new release. It might be eaiser to change a few lines and recompile source than to google "how to inject code into a process" and following directions, but 90% of the hacks out there for other games at least are just script kiddies following some recipe. Besides just because there might be a few more hacks or hacked executables floating around out there doesn't necessarily correlate to more people who suddenly have their morals corrupted. I can't promise you that the number of hackers will be non-zero compared to closed source, but what I suspect will happen...if anything at all...will be a slight bump in players with exploited clients while the kiddies get it out of their system and then things will go back to normal as they get banned and leave. I really think that the fear of hacks is overblown. There are plenty of other open source games out there that get along just fine and aren't overrun with hacks and cheats....I think the only way you get to that point is through admin neglect, if the server admins simply do not care and are not proactive in banning cheaters themselves. What do we gain? The four reasons I push for it are: 1. Prevents the Skulltag community from being trapped by a bad admin/coder. I don't think you are bad admins/coders (or else I wouldn't stick around) but I wouldn't keep nudging you about it unless I thought there was a very real danger of the same thing that happened to ZDaemon happening to Skulltag. I like Skulltag too much to have a replacement set of admins/coders kill it for good. I've been told that the current team would not allow this to happen, but it's hard to read people ahead of time...I know that Nightfang doesn't agree with many of the things that the current ZDaemon team has done. 2. It would be nice to have Skulltag not seen alongside ZDaemon as a second-class citizen in the Doom community anymore. 3. New code contributions. The 97b release gave use Spleens unlagged code and gave birth to ScoreDoomST. 4. Why not. :) Skulltag is built on the shoulders of tons of code that is all open source, it only makes sense to give back.

(0001065) Xenaero (reporter) 2011-02-13 19:54 edited on: 2011-02-13 20:06	"Actually, I'd like to know how many of the players would be willing to accept more hacks as price for going open source. I personally possibly could be convinced, but it wouldn't make any sense to release the latest source, if this turned Skulltag into a ghost town." --- In my long and arduous 'career' as a multiplayer, competitive Doomer, I can say this for absolute certain: There will be a small, barely above negligible yet-still-noticeable increase in exploits and hacks, however the benefits of things that have come from the previous source release will be magnified far beyond the scope of cheats being published. There is more to gain than to lose. AlexMax already touched on the incredible gains profited from the release of the previous codebase. One reason I argue for it is that in the past, ports being closed-source does not in fact inhibit cheating a great deal. Why this is is because they all share the same fundamental base coding. You can make a working inject-wallhack with say, an ancient version of ZDoom, it's almost guaranteed to work in Skulltag, ZDaemon, Odamex and any other port inheriting from the ZDoom codebase. This will not change. Skulltag will not turn into a ghost town at this point because the community thrives on its modding capabilities and active developer support. Any cheats made public will of course be seen by the developing portion of the community and the main developers themselves and likely be patched out. Any cheats that are created and as it stands now, kept out of the public spotlight for use by a select few troublemakers will never reach the point of critical mass where it drives the vast majority of users away. The best thing Skulltag has going for it is that it has a great developer-player relationship, in that if someone hops on a server with an exploit, almost everyone who notices it will have some information on it and awareness will be raised. Since Skulltag is being actively developed this makes a hacking subcommunity exceedingly hard to take root and take over the community. This is the case with a great many multiplayer titles with active developer support, either open-source or closed, old games or new. It won't turn into a ghost town just as it isn't happening now, in my belief. It's the fact that one could just take the old source and make a elementary hack like an aimbot or wallhack just as they have before and do now; yet as before it's been patched out and the users banned. There will always be a cycle of new exploits and patches to shut out these exploits, that's a given. All I want to point out is look at what we've gained from the previous source experiment. I believe the scope of benefits of opening the source heavily outweighs the risks and detriments of keeping the source closed. EDIT: I forgot to say, the old source release isn't really that old, and we didn't see a big spike in hacks as was feared and expected.

(0001066) AlexMax (developer) 2011-02-13 20:11 edited on: 2011-02-13 20:12	Two good examples of open source games with very little protection that aren't ghost towns: Warsow:'http://dpmaster.deathmask.net/?game=warsow [^]' Urban Terror:'http://dpmaster.deathmask.net/?game=urbanterror [^]' Making a hack for one of these games has got to be one of the easiest things in existance, with both games using the Quake engine. Yet look at those player-bases...they're huge! As of this writing, Warosow has 142 players and Urban Terror has almost 4,000! Doesn't look like a ghost town to me. :)

(0001067) mandolore (reporter) 2011-02-13 20:20	Pretty much what has been stated has been stated. Hacking is always going to be there, but in my view the open source has many benefits and less problems than leaving it closed source

(0001068) AlexMax (developer) 2011-02-13 20:28 edited on: 2011-02-13 20:55	One other thing...why CDDL? I thought you were simply throwing that out there as an example licence, but it's not GPL compatible at all, whereas the majority of ZDoom/GZDoom _could_ be if the BUILD code and FMod code could be sorted out, and if Randy and Graf gave the okay (Randy did in fact bless ZDoom 1.22 minus Hexen code as GPL, which is why Odamex is allowed to use it).

(0001069) Torr Samaho (administrator) 2011-02-13 20:47	Ok, for the time being let me just comment on one point, because I think it's necessary that everybody is aware of the following: > Any cheats made public will of course be seen by the developing portion of the community and the main developers themselves and likely be patched out. When we go open source, patching out hacks will be (pretty much) impossible. Currently the hacks are patched out by making some changes so that the hack breaks or can't hook in anymore. When the hacker has the source, he doesn't need to hook in anything, he just goes straight to the line in the code that does the depth check for example, turns if off, recompiles and has a working wallhack. The only way to counter a wallhack in an open source environment would be a denial of information approach, but that severely hampers honest players and I certainly don't want to take this route. Furthermore, denial of information can't do anything to stop an aimbot. <tl;dr> When going open source, hacks can't be patched out anymore. > One other thing...why CDDL? Because CDDL is compatible with a lot of licenses while GPL essentially is only compatible with itself. I could easily slap CDDL on the Skulltag specfic source files without needing to touch the licenses of the non-Skulltag files. In particular, CDDL on the ST files doesn't require the FMOD or build stuff to be removed (at least from what I know, if this is wrong, feel free to correct me). I personally think CDDL is a much better license than GPL. It's much more permissive, yet nicely fits the open source spirit and allows to merge projects using different licenses. Nevertheless, if I should release ST under CDDL (of course only the Skulltag exclusive files would have the license on them), I would allow it to be relicensed to GPL. Note: I wrote the last part before your edit.

(0001070) AlexMax (developer) 2011-02-13 20:57	It's okay, I'll just respond to you proper. :) Blzut3 is straightening me out, apparently "It must allow closed source use without dynamic linking, but also force the release of any changes to the Skulltag source" and I think it's good reasoning, though flawed...what's preventing any potential closed source forks from simply putting in stub methods to closed source bits and releasing that? It would be technically adhering to the license but totally useless to Skulltag. That GPL relicensing bit is really awesome of you though. :)

(0001071) Torr Samaho (administrator) 2011-02-13 21:08	> what's preventing any potential closed source forks from simply putting in stub methods to closed source bits and releasing that? Feasibility? That sounds technically possible, but very cumbersome to maintain. And probably the wrath of the Doom community ;). Flagging the ST files as CDDL would sufficiently express our support for open source and our desire to get access to derived works.

(0001072) Gez (reporter) 2011-02-13 22:08	LGPL is a good alternative, too, because it's completely compatible with GPL but it doesn't require retarded things like only linking to GPL stuff. ZDoom does actually contain some LGPL code.

(0001073) Torr Samaho (administrator) 2011-02-13 22:22	AFAIK one has to keep in mind that the LGPL allows its code to be consumed by a GPL project, i.e. you release some code under LGPL, a GPL code imports and enhances the code and you won't be able to use the enhanced code under the LGPL anymore.

(0001074) Gez (reporter) 2011-02-13 22:27	That would also happen in your proposed "CDDL but with permission to relicense as GPL" scheme. :p

(0001075) Xenaero (reporter) 2011-02-13 22:29	"When going open source, hacks can't be patched out anymore." This might be true to a degree, I doubt anyone can really stop inject-based hacks. Yet like I said with the support of the community behind the developing process of the port, and the stigma of basically being permabanned from everything Skulltag related if you're caught, I doubt hacking would increase tenfold as feared.

(0001077) Gez (reporter) 2011-02-13 23:46	Now people are gonna fake their IP to be Carnevil:'http://www.doomworld.com/vb/thread/54352/ [^]'

(0001086) Gez (reporter) 2011-02-14 19:33	Just checked one thing, Skulltag does feature the OPL emulator, which it shouldn't. It's taken from MAME and under a license that specifies that the full source code should be freely available.

(0001089) Torr Samaho (administrator) 2011-02-15 00:28	Thanks for the info! I wasn't aware of this and will remove it.

(0001093) AlexMax (developer) 2011-02-15 22:13	"AFAIK one has to keep in mind that the LGPL allows its code to be consumed by a GPL project, i.e. you release some code under LGPL, a GPL code imports and enhances the code and you won't be able to use the enhanced code under the LGPL anymore." Okay, now I'm a little confused. What makes "CDDL but with a code donation as GPL exception" palatable but LGPL not?

(0001094) Torr Samaho (administrator) 2011-02-16 01:23	Technically, it's not a big difference, but you don't have the cumbersome dynamic linking requirement when using CDDL code in a closed source project. What's more important to me is that the selection of the license sends a different message. By choosing CDDL, I make clear which kind of license I prefer, and by allowing it to be relicensed to GPL, I show that I intentionally also allow it to be used under a different license even if Skulltag won't be able to benefit from it. If someone chooses LGPL, IMO it's not clear whether or not he was aware that even though he will be guaranteed to see the source of derived works, he possibly will be locked out of using it.

(0001097) AlexMax (developer) 2011-02-17 02:44	I've asked Cyberimp about the source release, he seemed eager to defer the decision to you. 17:35:20 <Cyber> well, in the end overall it's torr's call on whether he releases it or not, i can only influence his decision I posted a community thread here to get the opinions of the community proper on the subject of the open source release: 'http://www.skulltag.com/forum/viewtopic.php?f=3&t=28199&start=0 [^]' Interesting little tidbit from #doom-tech when I was talking to Quasar (lead developer of Eternity Engine) about the potential for a Skulltag source release: 15:58:57 <Quasar```> of course I would personally be a lot more comfortable allowing exceptions to the licensing of code I write when the people receiving the exception have demonstrated respect for the free software idea and for the rest of the community When asked for clarification: 21:38:17 <Quasar```> I am saying if skulltag is open source and usable by other projects then I don't have to throw a hissy when zdoom borrows some of my code again 21:38:48 <Quasar```> cuz it's not that I'm a GPL nazi, it's that I don't like people taking the code I offer openly and closing it So perhaps a source release with a acceptable GPL exemption could be beneficial for ZDoom and GZDoom as well, not just GPL projects...

(0001307) AlexMax (developer) 2011-04-07 16:42	To continue from the login server ticket, what other factors other than the login server make the decision to go open source easier or harder? You replied: > It's better to discus this in 0000291. Just to name one thing: Only relatively few people (less than hundred) voted in the forum poll so far. Considering that there are usually more than hundred players playing online at once, only a fraction of the player base voiced their opinion yet. True, but I think we _did_ get a considerable cross-section of the Skulltag forum community which are among its most active community members. Also keep in mind that on top of that that the hundreds of players we get daily is only some small fraction of the thousands or tens of thousands of skulltag players out there...and I can't really think of a feasible way to get ALL of their opinions. Anything else that would make it easier or harder? Preferably something actionable...something I can actually work on.

This issue is already marked as resolved. If you feel that is not the case, please reopen it and explain why.
Supporters:	No one explicitly supports this issue yet.
Opponents:	No one explicitly opposes this issue yet.

Issue History
Date Modified	Username	Field	Change
2011-02-13 16:34	AlexMax	New Issue
2011-02-13 18:00	Torr Samaho	Note Added: 0001063
2011-02-13 18:00	Torr Samaho	Status	new => feedback
2011-02-13 19:39	AlexMax	Note Added: 0001064
2011-02-13 19:39	AlexMax	Status	feedback => new
2011-02-13 19:40	AlexMax	Note Edited: 0001064	View Revisions
2011-02-13 19:46	AlexMax	Note Edited: 0001064	View Revisions
2011-02-13 19:54	Xenaero	Note Added: 0001065
2011-02-13 20:06	Xenaero	Note Edited: 0001065	View Revisions
2011-02-13 20:11	AlexMax	Note Added: 0001066
2011-02-13 20:12	AlexMax	Note Edited: 0001066	View Revisions
2011-02-13 20:20	mandolore	Note Added: 0001067
2011-02-13 20:28	AlexMax	Note Added: 0001068
2011-02-13 20:32	AlexMax	Note Edited: 0001068	View Revisions
2011-02-13 20:43	AlexMax	Note Edited: 0001068	View Revisions
2011-02-13 20:47	Torr Samaho	Note Added: 0001069
2011-02-13 20:52	Torr Samaho	Status	new => feedback
2011-02-13 20:54	AlexMax	Note Edited: 0001068	View Revisions
2011-02-13 20:55	AlexMax	Note Edited: 0001068	View Revisions
2011-02-13 20:57	AlexMax	Note Added: 0001070
2011-02-13 20:57	AlexMax	Status	feedback => new
2011-02-13 21:08	Torr Samaho	Note Added: 0001071
2011-02-13 22:08	Gez	Note Added: 0001072
2011-02-13 22:22	Torr Samaho	Note Added: 0001073
2011-02-13 22:27	Gez	Note Added: 0001074
2011-02-13 22:29	Xenaero	Note Added: 0001075
2011-02-13 23:46	Gez	Note Added: 0001077
2011-02-14 19:33	Gez	Note Added: 0001086
2011-02-15 00:28	Torr Samaho	Note Added: 0001089
2011-02-15 22:13	AlexMax	Note Added: 0001093
2011-02-16 01:23	Torr Samaho	Note Added: 0001094
2011-02-17 02:44	AlexMax	Note Added: 0001097
2011-04-07 16:42	AlexMax	Note Added: 0001307
2012-02-13 02:00	Blzut3	Status	new => resolved
2012-02-13 02:00	Blzut3	Resolution	open => fixed
2012-02-13 02:00	Blzut3	Assigned To	=> Blzut3
2012-02-13 02:00	Blzut3	Assigned To	Blzut3 =>
2018-09-30 22:49	Blzut3	Status	resolved => closed

Relationships