[01:19:51] <-- BrainChild has left IRC (Remote host closed the connection)
[01:20:29] --> BrainChild has joined #exult
[01:25:28] <-- BrainChild has left IRC (Ping timeout: 240 seconds)
[01:34:20] --> DominusExult has joined #exult
[01:34:20] <-- DominusExult has left IRC (Changing host)
[01:34:20] --> DominusExult has joined #exult
[01:34:20] --- ChanServ gives channel operator status to DominusExult
[01:35:28] <-- Dominus has left IRC (Ping timeout: 240 seconds)
[01:35:29] --- DominusExult is now known as Dominus
[01:58:27] --> BrainChild has joined #exult
[02:09:41] <-- BrainChild has left IRC (Ping timeout: 250 seconds)
[02:14:30] <-- Rottingbeef-2 has left IRC ()
[02:47:34] --> BrainChild has joined #exult
[03:21:05] <-- BrainChild has left IRC (Ping timeout: 260 seconds)
[07:39:50] --> BrainChild has joined #exult
[07:49:32] <-- BrainChild has left IRC (Ping timeout: 264 seconds)
[07:57:41] <-- Lightkey has left IRC (Ping timeout: 260 seconds)
[08:10:58] --> Lightkey has joined #exult
[08:55:48] --> Marzo has joined #exult
[09:46:37] --> BrainChild has joined #exult
[10:20:01] <-- BrainChild has left IRC (Ping timeout: 256 seconds)
[11:29:35] --> BrainChild has joined #exult
[12:27:02] <Dominus> Marzo: https://www.dropbox.com/s/rj6bifx46bpmnsy/exultrelease.png?dl=0 any opinion on this?
[12:27:29] <Dominus> and Lightkey can you take a look as well? You have proven to have a good eye :)
[12:30:40] --> Rottingbeef-2 has joined #exult
[12:58:49] <-- BrainChild has left IRC (Ping timeout: 250 seconds)
[13:00:40] <Lightkey> Okay, so.. before I look at that, talk me through the versioning after 1.2 please.
[13:01:05] <Lightkey> What happened to 1.3, 1.4, and 1.5?
[13:02:33] <Lightkey> I thought the highlight was for announcing the merge of Exult with ScummVM. :-P
[13:07:27] <Lightkey> So I'm guessing 1.3 and 1.5 are considered development versions, which leaves the question what happened to 1.4.0-1.4.8.
[13:11:03] <Dominus> yes, odd numbers are development, even are release... then we had a 1.4.9rc1 but I don't remember the reasoning anymore. it's been ten years. if you search the logs back then you will likely find something :)
[13:12:19] <Dominus> and no, no plans to merge with ScummVM. It was perfect for Pentagram and xu4, probably for Nuvie, but I don't see Exult in ScummVM (with Exult Studio and everything)
[13:14:32] <Lightkey> Shame, that is the last step to achieve project CABAL. :->
[13:16:45] <Lightkey> And more pragmatically, it would prevent dropping old platforms.
[13:20:09] <Dominus> Not necessarily. The reason for dropping old platforms is that for the share pointers stuff etc. (Marzo knows best), it's better to go with newer c++ standards. Even if we were part of ScummVM this would be a thing to do
[13:20:35] <Dominus> And didn't ScummVM drop support for old OS?
[13:21:56] <Dominus> With Exult Studio maintaining old OS support proved to be its own dll hell. Kirben uses ancient dlls for the snapshots...
[13:22:39] <Dominus> CABAL... just reread that april fools page an hour ago ;)
[13:26:33] <Lightkey> I can't say without knowing which old platforms you are talking about.
[13:28:10] <Lightkey> They recently re-introduced support for Windows 9x, for example, which you don't seem to have.
[13:31:23] <Marzo> The biggest issue with 9x support (and a few other old OSes, like Amiga) is that there is no real testing to see if anything still works there
[13:31:59] <Marzo> Without people actively testing these things, they are most likely broken and not working
[13:32:44] <Marzo> Which is one of the reasons why I actually want to isolate the main code from platform specific code
[13:32:59] <Marzo> And get rid of most #ifdefs everywhere
[13:35:04] <Lightkey> Marzo: Btw, see xu4.sf.net and pentagram.sf.net for what Dominus has been up to, also soon on nuvie.sf.net: https://www.dropbox.com/s/0afbeqh3ronl377/nuvienews.png ;-)
[13:35:11] <Dominus> The release will feature 9x support. One legacy version with SDL1
[13:41:08] <Lightkey> Okay, so in late June 2004, the unstable snapshots were switched from 1.1.x to 1.4.x, now to find out why.
[13:44:13] <Dominus> Ever since the snapshots switched to SDL2 and thus no W9x support, no one complained ;)
[13:46:59] --> BrainChild has joined #exult
[13:48:06] <Dominus> And I think the pointers stuff made Exult require newer c++ standards which also cut off some old OS. But Marzo knows best ;)
[14:00:10] <Marzo> Technically speaking, GCC and clang support Win9x, and there is even a port of GCC for DOS
[14:00:45] <Marzo> What is hampering old OS support is ancient versions of GCC on other OSes, SDL (for Exult), and Gtk+ (for ES)
[14:01:04] <Marzo> GTK+, specifically, due to glib, which dropped support for Win9x
[14:01:51] <Marzo> And for WinXP, but there is a hack I came up that allows it to work on XP
[14:09:04] <Marzo> Huh... DJGPP has GCC 9.2 ftp://ftp.delorie.com/pub/djgpp/current/v2gnu/
[14:09:59] <Marzo> To me, this tells me that any OS/system that does not have as advanced a version of GCC as *DOS* is effectively dead
[14:14:14] <Lightkey> I guess it comes down to mentality then, ScummVM developers obviously disagree. ;-)
[14:15:04] <Lightkey> As I guessed, just a mishap: "I chose 1.4 in a hurry when I realized there were some pretty experimental changes in code that still identified itself as 1.2." -- Jeff Freedman
[14:20:08] <-- BrainChild has left IRC (Ping timeout: 240 seconds)
[14:25:40] <Marzo> I also remember proposing taking a page out of Gnome and keeping odd major numbers for experimental versions and even major numbers for releases
[14:25:54] <Dominus> yes
[14:25:57] <Marzo> So 1.5 is experimental, 1.6 is release, and so on
[14:26:15] <Lightkey> And Freedman agreed but then it stayed that way for some reason.
[14:26:43] <Dominus> I'm normally all for not killing off old OS but with the switch to SDL2, I kind of no longer care. hence this last version with old OS support and then off to new lands
[14:27:28] <Marzo> I would care more about old OSes if we had anyone capable of testing these OSes
[14:28:15] <Dominus> for PPC OS X I'd need to jump through too many hoops (getting a newer gcc to compile etc) and then still not have SDL2
[14:29:44] <Marzo> And then you get into things like Amiga, BeOS, and others
[14:31:33] <Dominus> and overall I have the feeling that to keep compatibility is more a principle and not filling an actual need
[14:33:31] <Dominus> On the Windows side, I barely believe people are still using XP. If they are using Windows 9x they are probably using it in some kind of VM and Exult (or ScummVM) will run natively on that platform
[14:34:43] <Lightkey> Marzo: That is kind of a chicken and egg thing though, ScummVM also has platforms that are broken years on end (stuff that needs to be tested, for building, they have buildbot screaming at developers when they break things on other platforms) but at least they can fix it when someone reports it, which is not as easy if the port does not exist in the first place.
[14:35:25] <Lightkey> Of course this is all hypothetical, as long as there is no answer for Exult Studio, a merge is out of the question.
[14:36:23] <Dominus> that's why all killing off has been tagged so people *can recreate the port if they so wish
[14:37:05] <Marzo> In this day and age, if someone is motivated enough to try Exult on such platforms, they know enough to make a port and submit a PR
[14:40:15] <Marzo> @Dominus: I have been thinking; it may be possible to import the ios fork history into main Exult repository
[14:40:55] <Marzo> Don´t know how important it would be, though
[14:43:06] <Dominus> I think the way we did it, is good enough. As we have it forked in the exult github project, it's all there
[14:44:12] <Dominus> I'd rather have our bugs/patches/feature requests portd over from sf to github
[14:44:23] <Lightkey> Dominus: Of course, many ports have been done because it can be done, not because they make any sense, just look at the Nintendo 64 port. Heck, they added all kinds of useless stuff like filters to make the games look like using Hercules graphics. Point is, if it's merged with ScummVM, the port supports would be done then and there, instead of waiting for porters to come to you.
[14:44:26] <Dominus> ;)
[14:47:51] <Dominus> That is also a valid point. There are many good points like all the stuff like scalers, mt32emu etc all being maintained by the big project.
[14:48:38] <Dominus> But I just don't feel like Exult Studio would work well over there.
[14:50:32] <Dominus> Not to mention that we (DrCode,Marzo, me) might lose our interest with it in ScummVM and thus important intimate knowledge being lost
[14:53:41] <Dominus> The reason why Pentagram thrives atm in ScummVM is that mduggan just showed up when we had merged. He wanted to work on Pentagram and was surprised we were now in ScummVM
[14:54:08] <Dominus> So his work would have been there anyway ;)
[15:11:18] <Dominus> Marzo, I think the time is right now to branch off master as v1.6 and merge devel-merge into master. Could you do this?
[15:12:17] <Marzo> You mean tag master into v1.6?
[15:12:35] <Marzo> Or do you want an actual branch, maybe for bug fixes?
[15:13:05] <Dominus> I think an actual branch would be best, exactly for potential bug fixes
[15:36:54] <Marzo> Its done
[15:37:10] <Marzo> I accidentally pushed a few branches I should not have, but I fixed that too
[15:40:38] <Dominus> thanks! I'll begin to rename stuff to 1.7.0git tonight
[15:41:24] <Marzo> I also updated the coverity branch, so we will see how many new complaints it has :)
[15:41:45] <Dominus> and I asked Kirben for two releases of Exult, SDL2 and SDL1.2x. As well as notified him of the branch changes
[15:43:11] <Dominus> looking forward to the coverty test :)
[15:46:16] <Dominus> if you care about this right now, clang complains about some NULLs in bggame.cc and sigame.cc
[15:46:30] <Dominus> and Gamemenu_gump.cc:57:20: warning: unused variable 'quitmenutext'
[15:46:53] <Dominus> Gump_manager.cc:565:29: warning: zero as null pointer constant
[15:47:38] <Dominus> but that's all the warnings I get now (and some in the ogg/vorbis headers :))
[15:57:36] --> travis-ci has joined #exult
[15:57:36] <travis-ci> exult/exult#288 (clang-tidy - cabfa10 : Marzo Sette Torres Junior): The build has errored.
[15:57:36] <travis-ci> Change view : https://github.com/exult/exult/compare/90c772895419^...cabfa1056cbf
[15:57:36] <travis-ci> Build details : https://travis-ci.org/exult/exult/builds/674094521
[15:57:36] <-- travis-ci has left #exult
[15:58:08] <Lightkey> Dominus: You are right about mduggan, I thought he had come to ScummVM first but in his welcome e-mail he clearly says it was the other way around. I thought he might have been docfu, who showed up in #pentagram a few days before, as they both seem to be from Japan but they have different names.
[15:59:21] <Marzo> Ah, forgot to stop the builds of those branches I accidentally pushed
[16:00:10] <Dominus> I know I am right about mduggan :) we were in email contact
[16:01:57] <Lightkey> Oh, did he contact you per e-mail first?
[16:02:03] <Dominus> yes
[16:03:23] <Lightkey> The Pentagram mailinglist has not been used in the last seven years, so I don't blame him for not trying there. :-)
[16:04:04] <Dominus> yeah, not to mention that SF threw everyone off their lists for dataprotection at one point
[16:07:02] <Dominus> huh... an Exult compile has never been so perfectly clean of warnings...
[16:08:52] <Dominus> later need to see how the optimized images look on iOS
[16:18:08] --> BrainChild has joined #exult
[16:23:04] <Lightkey> So, you wrote on the Exult mailinglist that you had decided on the IRC channel to use 1.4.9 RC1 as the release number but I can't find the corresponding chat in the logs around that time. I see from the ChangeLog that the snapshot versions were increased over time, which would explain why it was up to 1.4.9 at that point. What it does not explain is why choose the development "branch" for the RC. The only reason I
[16:23:04] <Lightkey> could see is to make this retroactively the stable branch and get back to the previous style of odd numbers are development versions.
[16:24:43] <Lightkey> Anyway, in this case, as there never was a stable release coming from that, I would also mention it in the 1.6.0 news post that it is the first stable version since 2004.
[16:24:55] --> travis-ci has joined #exult
[16:24:55] <travis-ci> exult/exult#289 (v1.6.x - aadee2a : DominusExult): The build failed.
[16:24:55] <travis-ci> Change view : https://github.com/exult/exult/compare/v1.6.x
[16:24:55] <travis-ci> Build details : https://travis-ci.org/exult/exult/builds/674094550
[16:24:55] <-- travis-ci has left #exult
[16:32:05] <Dominus> we didn't have a distinction between stable and development branch back then. We still don't... usually :)
[16:33:31] <Dominus> the extensive work on the pointers branches were kind of not usual for us :)
[16:33:58] <Dominus> but yes, I will mention this, especially as the 1.4.9rc1 release was later considered too unstable
[16:37:56] <Lightkey> Regarding typos, first sentence, "the 1.4.9rc1" should be either "version 1.4.9rc1" or "the 1.4.9 release candidate 1". Second sentence, missing comma after "having moved on". Further down "So please grabe". In the changes, Switch to SDL2, the OS at the end should be plural. Fifth line in the changes has a full stop at the end unlike any other line. Whatever a Papederdoll is. Seperate is a misspelling of Separate.
[16:37:56] <Lightkey> Whatever Fire Elemenatals are. Some missing hyphens in the last changes: cached-out, in-depth, and plot-stopping.
[16:39:04] <Dominus> thanks!
[16:47:38] <Dominus> so, yes, after the development versions were set to 1.4x and with previous versions staying on development numbering for beta and rc releases, it did make sense to name it 1.4.9rc1...
[16:48:54] --> travis-ci has joined #exult
[16:48:54] <travis-ci> exult/exult#299 (master - 21bf263 : GH Cao): The build passed.
[16:48:54] <travis-ci> Change view : https://github.com/exult/exult/compare/e5fa9cfdecad...21bf263118c5
[16:48:54] <travis-ci> Build details : https://travis-ci.org/exult/exult/builds/674107295
[16:48:54] <-- travis-ci has left #exult
[16:50:40] <-- BrainChild has left IRC (Ping timeout: 265 seconds)
[16:51:05] <Dominus> huh... marzo, that coverity site sure did change
[16:51:47] <Marzo> Looks the same to me
[16:52:56] <Dominus> ah, just going to coverity.com did forward me to synopsys.com
[16:53:34] <Dominus> scan.coverity.com is the right place :)
[16:56:49] <Lightkey> Indeed, didn't notice that. That is not the way it's usually used though, maybe for beta versions but RCs are always the version that will become the stable version.
[16:57:09] <Marzo> The RC1 was too buggy to become a release
[16:57:40] <Lightkey> I know, we were talking about previous 1.1.x versioning.
[17:00:30] <Lightkey> The release candidates for 1.2 were 1.1.9rc1 and 1.1.10rc2, that's the first I've seen it used that way.
[17:02:14] <Dominus> Different times back then ;)
[17:06:14] <Dominus> And we probably didn't think too much about it
[17:17:05] <Lightkey> Clearly, you've made a fine mess to untangle but it's finally resolved one and a half decades later. ;-)
[18:47:47] --> BrainChild has joined #exult
[19:21:11] <-- BrainChild has left IRC (Ping timeout: 256 seconds)
[19:31:17] --> BrainChild has joined #exult
[19:36:04] <-- BrainChild has left IRC (Client Quit)
[19:41:10] --> BrainChild has joined #exult
[19:58:31] <-- Marzo has left IRC (Ping timeout: 256 seconds)