ScummVM IRC Log - 08 Dec 2017

[Back to Index]

[00:17] --> Strangerke_ joined #scummvm.
[00:19] <-- Strangerke left irc: Ping timeout: 240 seconds
[00:19] Nick change: Strangerke_ -> Strangerke
[00:23] <-- criezy left irc: Quit: criezy
[00:46] <-- girafe left irc: Read error: Connection reset by peer
[00:50] tomazcuber[m] --> (tomazcuber@gateway/shell/matrix.org/x-utwfxbpqilbtuomh) joined #scummvm.
[01:21] <-- ccawley2011 left irc: Ping timeout: 240 seconds
[01:23] --> flesk_ joined #scummvm.
[01:23] trojan[m] --> (trojanmatr@gateway/shell/matrix.org/x-ebhrgeuytbfzwkkc) joined #scummvm.
[01:23] yashgandhe[m] --> (yashgandhe@gateway/shell/matrix.org/x-vvivhmaxlpaaywox) joined #scummvm.
[01:23] weezywinez[m] --> (weezywinez@gateway/shell/matrix.org/x-afkgkndtuvibvxmt) joined #scummvm.
[01:23] johnwhitlow[m] --> (johnwhitlo@gateway/shell/matrix.org/x-kgazzusqztkcofzt) joined #scummvm.
[01:23] kusushi[m] --> (kusushimat@gateway/shell/matrix.org/x-buqbdfeqtlezhpsi) joined #scummvm.
[01:23] super_fox2[m] --> (superfox2m@gateway/shell/matrix.org/x-aasolqjedpmawgcs) joined #scummvm.
[01:23] Coldwine[m] --> (coldwinema@gateway/shell/matrix.org/x-utsxahnalnhleulr) joined #scummvm.
[01:44] --> ccawley2011 joined #scummvm.
[01:52] <-- ccawley2011 left irc: Quit: Bye
[02:16] <-- SylvainTV left irc: Read error: Connection reset by peer
[02:19] <-- Dominus left irc: Ping timeout: 260 seconds
[02:19] --> Dominus joined #scummvm.
[02:28] <tsoliman> I pushed to master and branch-2-0 in the same push command so from git's POV it is a single "receive" but I don't know if the github bot handles that case. As far as I can see it only sees one branch.
[02:29] Action: tsoliman just noticed the run-on sentence
[02:29] <-- Joefish left irc: Ping timeout: 276 seconds
[02:30] <dreammaster> Heh.. lately the Github bot doesn't seem to handle even single commits from me, most of the time :P
[02:30] --> Joefish joined #scummvm.
[02:30] #scummvm: mode change '+v Joefish' by ChanServ!ChanServ@services.
[03:00] <-- Joefish left irc: Ping timeout: 258 seconds
[03:14] --> GitHub82 joined #scummvm.
[03:14] <GitHub82> [scummvm] dreammaster pushed 1 new commit to master: https://git.io/vbl1E
[03:14] <GitHub82> scummvm/master 4c777a9 Paul Gilbert: XEEN: Separation of scene indoors & outdoors drawing
[03:14] GitHub82 (GitHub82@192.30.252.36) left #scummvm.
[03:21] <-- dreammaster left irc:
[03:29] <-- antlarr left irc: Ping timeout: 240 seconds
[03:30] --> antlarr joined #scummvm.
[03:40] --> Strangerke_ joined #scummvm.
[03:42] <-- Strangerke left irc: Ping timeout: 276 seconds
[03:42] Nick change: Strangerke_ -> Strangerke
[07:08] <-- Simei left irc: Ping timeout: 246 seconds
[07:09] weezywinez[m] <-- (weezywinez@gateway/shell/matrix.org/x-afkgkndtuvibvxmt) left irc: Ping timeout: 246 seconds
[07:14] yashgandhe[m] <-- (yashgandhe@gateway/shell/matrix.org/x-vvivhmaxlpaaywox) left irc: Ping timeout: 240 seconds
[07:14] trojan[m] <-- (trojanmatr@gateway/shell/matrix.org/x-ebhrgeuytbfzwkkc) left irc: Ping timeout: 252 seconds
[07:14] johnwhitlow[m] <-- (johnwhitlo@gateway/shell/matrix.org/x-kgazzusqztkcofzt) left irc: Ping timeout: 252 seconds
[07:14] tomazcuber[m] <-- (tomazcuber@gateway/shell/matrix.org/x-utwfxbpqilbtuomh) left irc: Ping timeout: 240 seconds
[07:15] Coldwine[m] <-- (coldwinema@gateway/shell/matrix.org/x-utsxahnalnhleulr) left irc: Ping timeout: 276 seconds
[07:15] kusushi[m] <-- (kusushimat@gateway/shell/matrix.org/x-buqbdfeqtlezhpsi) left irc: Ping timeout: 264 seconds
[07:15] super_fox2[m] <-- (superfox2m@gateway/shell/matrix.org/x-aasolqjedpmawgcs) left irc: Ping timeout: 258 seconds
[07:15] <-- flesk_ left irc: Ping timeout: 276 seconds
[07:25] <-- heroux left irc: Ping timeout: 250 seconds
[07:38] --> Begasus joined #scummvm.
[07:49] --> Begas_VBox joined #scummvm.
[07:54] <-- kurtwr left irc: Ping timeout: 248 seconds
[07:56] --> GitHub62 joined #scummvm.
[07:56] <GitHub62> [scummvm] sev- closed pull request #1082: SKY: Fix unaligned pointer dereferences (master...sky-unaligned) https://git.io/vbm1H
[07:56] GitHub62 (GitHub62@192.30.252.34) left #scummvm.
[07:56] --> GitHub4 joined #scummvm.
[07:56] <GitHub4> [scummvm] sev- pushed 1 new commit to master: https://git.io/vblNa
[07:56] <GitHub4> scummvm/master 8e82937 Colin Snover: SKY: Fix unaligned pointer dereferences...
[07:56] GitHub4 (GitHub4@192.30.252.40) left #scummvm.
[07:56] --> GitHub66 joined #scummvm.
[07:56] <GitHub66> [scummvm] sev- pushed 1 new commit to master: https://git.io/vblNr
[07:56] <GitHub66> scummvm/master ff8a031 Matan Bareket: Added MD5 for the German version of Imo and the King
[07:56] GitHub66 (GitHub66@192.30.252.34) left #scummvm.
[08:10] --> kurtwr joined #scummvm.
[08:16] --> Ahwhat joined #scummvm.
[08:18] <Begasus> ok, which one should we check for the new release (2.0.0pre or 2.1.0git)?
[08:24] --> Joefish joined #scummvm.
[08:24] #scummvm: mode change '+v Joefish' by ChanServ!ChanServ@services.
[09:00] <-- borosky left irc: Ping timeout: 255 seconds
[09:06] --> Simei joined #scummvm.
[09:14] --> waltervn joined #scummvm.
[09:14] #scummvm: mode change '+o waltervn' by ChanServ!ChanServ@services.
[09:15] <-- Ahwhat left irc:
[09:19] <waltervn> morning
[09:21] tomazcuber[m] --> (tomazcuber@gateway/shell/matrix.org/x-pyxzjcxwhlvfkrou) joined #scummvm.
[09:24] <-- LittleToonCat left irc: Remote host closed the connection
[09:48] <-- Strangerke left irc: Ping timeout: 248 seconds
[09:54] johnwhitlow[m] --> (johnwhitlo@gateway/shell/matrix.org/x-ggqxnkpfetabyghe) joined #scummvm.
[09:54] trojan[m] --> (trojanmatr@gateway/shell/matrix.org/x-lbzcfmunjnungiax) joined #scummvm.
[09:54] yashgandhe[m] --> (yashgandhe@gateway/shell/matrix.org/x-knpnhrbuzgcnhoco) joined #scummvm.
[09:54] --> flesk_ joined #scummvm.
[09:54] weezywinez[m] --> (weezywinez@gateway/shell/matrix.org/x-kecusldqqsgenbfu) joined #scummvm.
[09:54] kusushi[m] --> (kusushimat@gateway/shell/matrix.org/x-fvmomdulcdmyxqyt) joined #scummvm.
[09:54] Coldwine[m] --> (coldwinema@gateway/shell/matrix.org/x-ufrscxwqxtgiasnc) joined #scummvm.
[09:54] super_fox2[m] --> (superfox2m@gateway/shell/matrix.org/x-bniihopexyibuuca) joined #scummvm.
[10:03] --> Strangerke joined #scummvm.
[10:20] --> criezy|Work joined #scummvm.
[10:21] #scummvm: mode change '+o criezy|Work' by ChanServ!ChanServ@services.
[10:22] <criezy|Work> Good morning
[10:23] <criezy|Work> I learnt about MemoServ today :)
[10:23] <criezy|Work> Thanks snover, that seems like a useful feature!
[10:25] --> ignalina joined #scummvm.
[10:25] <wjp> Begasus: 2.0.0pre is what's going to become 2.0.0
[10:29] --> Strangerke_ joined #scummvm.
[10:31] <-- Strangerke left irc: Ping timeout: 240 seconds
[10:31] Nick change: Strangerke_ -> Strangerke
[11:14] <Begasus> ok thank wjp, thought so (and build checked +1) :)
[11:14] --> Strangerke_ joined #scummvm.
[11:16] <-- Strangerke left irc: Ping timeout: 260 seconds
[11:16] Nick change: Strangerke_ -> Strangerke
[11:33] --> Strangerke_ joined #scummvm.
[11:35] <-- Strangerke left irc: Ping timeout: 248 seconds
[11:35] Nick change: Strangerke_ -> Strangerke
[11:46] <-- rootfather left irc:
[11:56] --> TMM joined #scummvm.
[11:56] <-- TMM left irc: Changing host
[11:56] --> TMM joined #scummvm.
[11:56] #scummvm: mode change '+o TMM' by ChanServ!ChanServ@services.
[11:58] --> salty-horse joined #scummvm.
[11:58] <-- salty-horse left irc: Changing host
[11:58] --> salty-horse joined #scummvm.
[11:58] #scummvm: mode change '+o salty-horse' by ChanServ!ChanServ@services.
[12:00] --> jam__ joined #scummvm.
[12:05] --> borosky joined #scummvm.
[12:15] <-- salty-horse left irc: Ping timeout: 240 seconds
[12:30] --> salty-horse joined #scummvm.
[12:30] <-- salty-horse left irc: Changing host
[12:30] --> salty-horse joined #scummvm.
[12:30] #scummvm: mode change '+o salty-horse' by ChanServ!ChanServ@services.
[12:37] <-- ignalina left irc: Quit: ignalina
[12:52] --> ignalina joined #scummvm.
[14:02] <-- ignalina left irc: Quit: ignalina
[14:09] --> Henke37 joined #scummvm.
[14:09] <-- Tomaz^ left irc: Ping timeout: 258 seconds
[14:19] --> Tomaz^ joined #scummvm.
[14:31] --> Littleboy joined #scummvm.
[14:31] #scummvm: mode change '+o Littleboy' by ChanServ!ChanServ@services.
[15:05] --> ignalina joined #scummvm.
[15:13] --> ^^Hawk^^ joined #scummvm.
[15:19] <-- ^^Hawk^^ left irc: Ping timeout: 246 seconds
[15:21] --> ^^Hawk^^ joined #scummvm.
[15:28] <-- ignalina left irc: Quit: ignalina
[15:33] --> Guest13621 joined #scummvm.

[15:35] --> ignalina joined #scummvm.
[15:35] <-- ignalina left irc: Client Quit
[15:36] <-- Tomaz^ left irc: Ping timeout: 255 seconds
[15:51] --> ccawley2011 joined #scummvm.
[15:53] <-- ^^Hawk^^ left irc: Ping timeout: 248 seconds
[15:54] --> ^^Hawk^^ joined #scummvm.
[16:05] <-- Joefish left irc: Ping timeout: 240 seconds
[16:26] --> ignalina joined #scummvm.
[16:38] --> eriktorbjorn1 joined #scummvm.
[16:39] <-- eriktorbjorn left irc: Ping timeout: 255 seconds
[16:43] <-- ^^Hawk^^ left irc: Ping timeout: 248 seconds
[16:44] --> ^^Hawk^^ joined #scummvm.
[16:45] <-- TMM left irc: Quit: Ex-Chat
[16:55] --> ajax16384 joined #scummvm.
[16:55] #scummvm: mode change '+o ajax16384' by ChanServ!ChanServ@services.
[16:58] <snover> criezy|Work: youre welcome. it can be less hassle than a bouncer, for sure :)
[16:58] <criezy|Work> snover: regarding bug 10353, doesn't the latest stable build already use SDL 2 though?
[16:58] <criezy|Work> Or maybe he meant the stable build from buildbot, in which case it is indeed SDL 1 I think.
[16:59] <criezy|Work> (and not the official release 1.9.0 package)
[16:59] <snover> scummvm 1.9 release build is also sdl1 on windows.
[16:59] <criezy|Work> There are issues reported with SDL2 and Windows 10 when using relative mouse: https://bugzilla.libsdl.org/show_bug.cgi?id=3931
[17:00] <criezy|Work> But hopefully that does not impact us.
[17:00] <criezy|Work> Unless the mouse grabing suffers from the same issue.
[17:01] <snover> its a pretty ugly issue.
[17:03] <snover> i guess i am glad that we never use relative mode so it is partially mitigated by accident.
[17:07] Action: wjp adds new GAMEOPTION_RGB_RENDERING to all SCI16 games detection entries
[17:08] <snover> wjp: so many gameoptions! i am looking forward to exploring that PR soon.
[17:08] <wjp> (after first replacing the top three most popular sets of gui/game options in the table by #defines...)
[17:22] <-- jam__ left irc: Ping timeout: 255 seconds
[17:22] --> rootfather joined #scummvm.
[17:22] #scummvm: mode change '+o rootfather' by ChanServ!ChanServ@services.
[17:54] --> Joefish joined #scummvm.
[17:54] #scummvm: mode change '+v Joefish' by ChanServ!ChanServ@services.
[18:13] <criezy|Work> Time to go back home :)
[18:13] <criezy|Work> And it's friday! :)
[18:13] <criezy|Work> I can start looking at what needs to be done for the release tarballs in about one hour.
[18:15] --> LittleToonCat joined #scummvm.
[18:16] <-- criezy|Work left irc: Quit: Page closed
[18:17] --> Strangerke_ joined #scummvm.
[18:18] <-- Strangerke left irc: Ping timeout: 248 seconds
[18:18] Nick change: Strangerke_ -> Strangerke
[18:21] <-- L0ngcat left irc: Read error: Connection reset by peer
[18:22] --> L0ngcat joined #scummvm.
[18:40] Nick change: eriktorbjorn1 -> eriktorbjorn
[18:40] #scummvm: mode change '+o eriktorbjorn' by ChanServ!ChanServ@services.
[18:47] --> marcusramberg joined #scummvm.
[18:50] --> criezy joined #scummvm.
[18:50] #scummvm: mode change '+o criezy' by ChanServ!ChanServ@services.
[18:53] --> heroux joined #scummvm.
[19:02] <-- heroux left irc: Ping timeout: 260 seconds
[19:08] --> heroux joined #scummvm.
[19:11] --> Farmboy0 joined #scummvm.
[19:11] <-- Farmboy0 left irc: Changing host
[19:11] --> Farmboy0 joined #scummvm.
[19:21] <rootfather> snover thanks for your support on the sdl mailing list :D
[19:25] --> GitHub53 joined #scummvm.
[19:25] <GitHub53> [scummvm] criezy pushed 1 new commit to branch-2-0: https://git.io/vb4vl
[19:25] <GitHub53> scummvm/branch-2-0 a963836 Thierry Crozat: I18N: Regenerate translations data file
[19:25] GitHub53 (GitHub53@192.30.252.36) left #scummvm.
[19:58] --> girafe joined #scummvm.
[20:17] <criezy> OK, I think I made almost everything needed for the 2.0.0 branch.
[20:17] <snover> rootfather: one must simply know the magic words :)
[20:17] <criezy> I will add the v2.0.0 tag and generate tarballs after dinner in about one hour.
[20:17] <criezy> And then look at scummvm-tools.
[20:17] <snover> criezy: awesome. thanks!
[20:17] <snover> and thank you for the NEWS updates too
[20:18] <criezy> What sdl mailing list is that? I though they closed it a few months ago and migrated to discourse?
[20:18] <snover> it was the bug tracker, not a mailing list.
[20:22] <criezy> Is that for the small graphical issue on Windows with bilinear filtering?
[20:22] <criezy> I remember a link to the SDL bug tracker was added to the ticket for that on our bug tracker a few weeks ago.
[20:23] <wjp> seems so
[20:23] <criezy> Ha yes, I just found it. And it is apparently fixed now!
[20:23] <criezy> Nice.
[20:30] <snover> yep, that one.
[20:32] <snover> for future reference, the key phrase to goad any maintainer into action is there is a patch that appears to fix the issue and its only <small number> of lines
[20:37] <rootfather> thanks, I'll try to remember this for future use :P
[20:49] <snover> all these branches are confusing me. is it correct that i need to cherry-pick a fix from master into branch-2-0 *and* branch-2-0-0 now?
[20:51] <wjp> wait, what?
[20:52] <wjp> ah, criezy created that branch
[20:52] <wjp> that branch is intended to be the final 2.0.0
[20:52] <wjp> so it shouldn't have to be touched anymore
[20:53] <snover> ok. is this an echo of the days before git where branches and tags were the same thing?

[20:54] <wjp> kind of, yes
[20:55] <wjp> the process is that the version number is updated to 2.0.0 on branch-2-0-0, and branch-2-0 can go on to become 2.0.1pre
[20:56] <wjp> it mirrors what happens for the major branching
[20:57] <snover> ok. so it is not the case currently, but we could just make the tag point to the commit that fixes the version without having a separate branch-2-0-0, and that would serve the same purpose?
[20:58] <snover> (and could potentially retroactively delete the release-version branches too, with no ill effect?)
[20:59] <criezy> I also generated all the MSVC and Code::Blocks project files only in the branch-2-0-0.
[21:00] <wjp> I think in the past there were some concerns about release tags also needing to be on at least one branch, for some reason that I can't quite remember
[21:00] <wjp> at least there were complaints when I deleted old release branches at some point
[21:02] <wjp> but I see that has happened now anyway with everything pre 1.4
[21:03] <wjp> so that memory is presumably entirely out of date
[21:04] <snover> gits default behaviour is to not fetch tags that dont point at objects reachable by a tracked branch head, so maybe that had something to do with it. of course you can fetch them all at any time by passing --tags, or reconfigure git to fetch tags by default from a remote if you want to.
[21:04] <snover> so maybe someone was just confused about this behaviour of not getting the tags when they pulled and didnt know the solution is to use --tags.
[21:04] <wjp> I just did a fresh clone, and it does get all tags
[21:05] <wjp> it seems at some point we closed all pre-1.4 branches by tagging their heads with closed/branch-x-y tags
[21:05] <snover> clone has an opposite behaviour, it pulls all tags unless you use --no-tags
[21:06] <snover> apparently.
[21:06] <wjp> opposite to what exactly?
[21:06] <wjp> fetch?
[21:06] <snover> opposite to fetch/pull
[21:06] <-- ^^Hawk^^ left irc: Quit: ByE ByE :D [Mondololloso Script] v2.0.1
[21:07] <snover> it is all configurable, though, so if someone doesnt like the defaults they can be changed forever in a global config.
[21:07] <rsn8887> I think I asked this before, but 2.0.0 will be built using the old buildbot, correct?
[21:07] <wjp> ah, there's a third option in between --tags and --no-tags
[21:07] <wjp> interesting
[21:07] <wjp> global config? Are you sure?
[21:08] <wjp> last time I looked it was per-remote
[21:08] <snover> rsn8887: buildbot builds no releases, it is porters responsibility to build and send the final releases currently.
[21:09] <rsn8887> snover: Nice, so I could build the release using the latest version sdk and libraries that are not on the old buildbot yet (because manually configuring that thing seems superflous if new buildbot will take over soon-ish).
[21:10] <snover> wjp: ill double-check in this specific case. i havent encountered anything yet that can be passed as a command-line flag which could not also be put into a config.
[21:10] <-- ignalina left irc: Quit: ignalina
[21:10] <wjp> "The default behavior for a remote may be specified with the remote.<name>.tagOpt setting"
[21:11] <wjp> a global setting would also be complicated by the lack of an explicit way of setting the default behaviour
[21:11] <wjp> I think I've noticed before that this tags behaviour feels a bit unpolished
[21:13] <wjp> anyway, it _is_ an option to put tag v2.0.0 directly on branch-2-0 and just re-increment the version to 2.0.1pre after tagging
[21:13] <wjp> we have sometimes had to do some late fixes to the release and re-tag, though
[21:14] <wjp> so incrementing the version would have to wait a while, but that shouldn't really be a problem
[21:16] <snover> it looks like thats how i used to automate releases on some other projects in the last. update versions to release version, commit, tag, update versions to next pre-version, commit.
[21:16] <snover> past*
[21:17] <snover> rsn8887: yes, hoping the new buildbot will be done shortly and then will be able to generate builds automatically moving forward.
[21:17] #scummvm: mode change '+o Strangerke' by ChanServ!ChanServ@services.
[21:18] <snover> if a release was broken in that case it would just end up getting a new patch number, no big deal.
[21:19] <wjp> with automated builds that's a lot easier, yes
[21:19] <rsn8887> snover: Those files like "NEWS" etc. that should be bundled... what is the preferred behaviour: a) those files be present as separate files in the release archive, b) those files actually being installed on the host machine when ScummVM is installed there.
[21:19] <snover> rsn8887: i need to run to an appointment so i will let someone else answer that :)
[21:19] <rsn8887> snover: OK thanks for the help.
[21:19] <wjp> I'm not sure what you see as the difference between those two options
[21:20] <snover> wjp: i ran out of time to try to find some --tags-defaulting, ill look when i get back if it matters, let me know.
[21:20] <wjp> snover: it's more curiosity on my side than that it really matters
[21:21] <criezy> For Windows that has an installer, those files might be displayed by the installer, and might also be installed by the installer (a Windows user might be able to confirm that).
[21:21] <criezy> For macOS, the files are just bundled with ScummVM on a disk image.
[21:21] <criezy> So I guess there is no simple answer ;)
[21:22] <rsn8887> snover: I see I already have a solution for Vita they will be present in the archive as single files and also installed...
[21:23] <wjp> you may have to explain what the word "archive" and "installed" means to you
[21:23] <wjp> in any case, you'll want to do what makes sense and/or is convention on the platform
[21:34] --> ignalina joined #scummvm.
[21:36] <-- rootfather left irc:
[21:37] <rsn8887> wjp: the Vita .vpk file is similar to Android .apk in the sense that it is unpacked on the host into an automatic directory. I would say the logic release format for Vita would be .vpk, similar to the Android .apk releases.
[21:37] <rsn8887> *logical
[21:41] <-- salty-horse left irc: Quit: Leaving
[21:46] <wjp> ok, but then I don't see what "in the archive as single files" means
[21:50] <snover> wjp: im not finding anything, so this could be one of the few cases where theres some non-defaultable option. maybe you could put remote.origin.tagOpt into your global config though.
[21:51] --> rootfather joined #scummvm.
[21:51] #scummvm: mode change '+o rootfather' by ChanServ!ChanServ@services.
[21:53] <snover> or you could set up a shell alias or git alias that includes the --tags flag.
[21:53] <wjp> I'll just live with it how it is :-)
[21:54] <snover> yeah. the default behaviour never really been an issue for me either.
[21:55] <wjp> (except for when it fetches rsn8887's tags :-) )
[21:58] <rsn8887> Haha. Tags on github certainly feel tagged-on like an afterthought.
[21:59] <snover> oh, yeah. rsn8887, if you create release tags like that in the future, you might consider something more like 2.0.0git-rsn8887.1, 2.0.0git-rsn8887.2, etc. where 2.0.0git is the upstream version.
[21:59] <rsn8887> ok
[21:59] <rsn8887> I didn't even want tags, but couldn't release on github without creating them.
[22:00] <-- Begas_VBox left irc: Quit: Vision[0.10.2]: i've been blurred!
[22:01] <-- Begasus left irc: Quit: Ex-Chat
[22:02] <rsn8887> I think there's still one left, because I wanted to give Vita users the option to go back to an earlier release in case the buildbot releases stopped working. Since Vita doesn't have pre-2.0.0 stable there's currently no other fallback. Until next week.
[22:06] --> GitHub29 joined #scummvm.
[22:06] <GitHub29> [scummvm] csnover pushed 2 new commits to master: https://git.io/vb4lw
[22:06] <GitHub29> scummvm/master 97e6ebc Colin Snover: NEWS: Fix portmanteau of macOS (new name) and OS X (old name)
[22:06] <GitHub29> scummvm/master 3938558 Colin Snover: NEWS: Fix Voyeur news not indented consistently with the rest of news
[22:06] GitHub29 (GitHub29@192.30.252.34) left #scummvm.
[22:06] <-- ajax16384 left irc: Quit: Leaving
[22:07] <snover> that would have been more useful to have noticed yesterday. oh well. its fine. :)
[22:07] <criezy> Oups.
[22:07] <criezy> That will give us a reason to make a 2.0.1 release :P
[22:10] <criezy> argh! The tarballs generation script fails because I don't have xz apparently.
[22:10] <criezy> It first create an uncompressed tar archive and then compresses it to gz, bz2 annd xz.
[22:11] <snover> its usually xz-utils
[22:11] <snover> while youre there, could you fix that so that it is only creating xz archives? there is no reason to be generating gz/bz2 ones.
[22:12] <snover> (or at the least there is no reason to generate bz2 ones)
[22:12] <wjp> well
[22:12] <criezy> I can't uncompress xz on the old system I use to generate the PPC binaries
[22:12] <wjp> from the .spec: Source : %{name}-%{version}.tar.bz2
[22:12] <criezy> But yes, I could use gz.
[22:13] <rsn8887> is 2.0.0 frozen now?
[22:13] <wjp> yes
[22:13] <wjp> (but I can live with having to hack around the lack of a .bz2)
[22:14] <snover> i did some research around this a while ago, i think when i was figuring out what the new buildbot should do, it seemed that for most workloads bz2 is slower and has a worse compression ratio than xz/lzma, so usually it seems people use gz for legacy stuff and xz for modern stuff
[22:15] <wjp> sure
[22:15] <snover> hence that recommendation. but, yes, if there are reasons why bz2 must continue to exist, then it may.
[22:15] <snover> i didnt think there were any.
[22:20] <snover> i am prone to occasional bouts of ignorance :)
[22:20] <criezy> OK. Let's run that script again.
[22:20] <criezy> ...
[22:22] <wjp> I'll switch the .spec to .xz. Presumably that should just work
[22:22] <criezy> That xz encoding sure take a long time...
[22:22] <criezy> Success!
[22:23] <criezy> So for info bz2 was a lot faster to compress than xz, but it does indeed have a worse compression ratio.
[22:25] <wjp> decompression speed matters more though
[22:25] <-- waltervn left irc: Quit: Leaving
[22:25] <criezy> Sure. Compression is only done once by one person.
[22:26] --> dreammaster joined #scummvm.
[22:26] #scummvm: mode change '+o dreammaster' by ChanServ!ChanServ@services.
[22:33] <snover> i chose xz 2 on buildbot since some benchmarks indicated it would be fairly close to gzip 9 compression speed and with a better compression ratio.
[22:34] <snover> the compressor definitely ramps up quickly on the slowness.
[22:37] <snover> wjp: i want to email github about the irc bot. i dont have access to admin, can you look at the service hook and see if there are a bunch of failures?
[22:40] <wjp> "Service Timeout"
[22:40] Action: wjp hits the test button
[22:41] --> GitHub76 joined #scummvm.
[22:41] <GitHub76> [scummvm] csnover pushed 2 new commits to master: https://git.io/vb4lw
[22:41] <GitHub76> scummvm/master 97e6ebc Colin Snover: NEWS: Fix portmanteau of macOS (new name) and OS X (old name)
[22:41] <GitHub76> scummvm/master 3938558 Colin Snover: NEWS: Fix Voyeur news not indented consistently with the rest of news
[22:41] GitHub76 (GitHub76@192.30.252.37) left #scummvm.
[22:41] <snover> hey, thats me!
[22:41] <wjp> I switched it over from non-SSL to SSL
[22:42] <wjp> just in case (and it seemed like a good idea anyway)
[22:42] <snover> that sounds good. ill hold off and see if that helps at all.
[22:42] Action: wjp hits the test button again
[22:42] --> GitHub66 joined #scummvm.
[22:42] <GitHub66> [scummvm] csnover pushed 2 new commits to master: https://git.io/vb4lw
[22:42] <GitHub66> scummvm/master 97e6ebc Colin Snover: NEWS: Fix portmanteau of macOS (new name) and OS X (old name)
[22:42] <GitHub66> scummvm/master 3938558 Colin Snover: NEWS: Fix Voyeur news not indented consistently with the rest of news
[22:42] GitHub66 (GitHub66@192.30.252.36) left #scummvm.
[22:43] <wjp> let's see what happens
[22:45] <-- heroux left irc: Ping timeout: 240 seconds
[22:46] --> heroux joined #scummvm.
[22:48] <Dark-Star> dreammaster: I know the Xeen engine is still very early WIP, but I wonder if I'm the only one getting this kind of graphics corruption right at the start
[22:48] <Dark-Star> dreammaster: https://imgur.com/a/Tozc9
[22:49] <criezy> snover: how many times did you push those two commits? :P
[22:49] <snover> as many times as wjp pushes the test button
[22:49] <snover> its like a free commit generator
[22:49] <dreammaster> Yes, I'm aware there's some transition issues in a few places like that. At the moment, I've got the engine to jump directly into the game if you start up with a global debug level set, and it doesn't have that problem when you do
[22:50] <dreammaster> So figuring why that particular problem occurs has been lower on my list of priorities
[22:50] <snover> im going to go out on a limb and say it is happening due to a bug
[22:50] <Dark-Star> yeah, okay, I was just wondering, I already suspected my build (I had a bit of trouble with MSVC lately ;-)
[22:51] <criezy> Wake up bot!
[22:51] <dreammaster> A lot of the windowing code for the current engine is a best guess of how the original worked, since the original had a lot of messy code for background save & restoring that used XMS. So it's possible my current implementation hasn't got the behaviour exactly right yet
[22:53] <snover> the X stands for Xtreme. the M stands for Mightmare. the& ok, ill see myself out.
[22:53] <wjp> criezy: wake up to what exactly?
[22:54] <dreammaster> Yes, lots of nasty things going on like remapping areas of memory, and fiddling around with handles.
[22:54] <criezy> I pushed a commit to scummvm-tools
[22:54] Action: wjp checks that one
[22:54] <criezy> But maybe you only changed the settings for the scummvm repo?
[22:54] <criezy> and not scummvm-tools?
[22:54] <wjp> "maybe" *cough*
[22:54] <-- Lightkey left irc: Ping timeout: 240 seconds
[22:55] <wjp> heh, scummvm-tools still has a _CIA_ hook :-)
[22:55] <Dark-Star> yeah, the inventory windows and character sheets are not completely working too :) but I figured since it's still WIP you don't really need any new bug reports
[22:56] Action: wjp hits a different test button
[22:56] Action: wjp can do this all day
[22:57] <-- Strangerke left irc: Ping timeout: 248 seconds
[22:57] <wjp> maybe it only works on snover's commits
[22:57] <snover> Yay, I'm special!
[22:58] <rsn8887> snover: did you already do something to scummvm-docker vita buildbot to enable those flac flags --without-PIC?
[22:58] <wjp> "Service Timeout"
[22:58] <rsn8887> Or shall I make a PR?
[22:58] --> GitHub126 joined #scummvm.
[22:58] <GitHub126> [scummvm] sev- pushed 4 new commits to master: https://git.io/vb4E3
[22:58] <GitHub126> scummvm/master c728c29 Eugene Sandulenko: FULLPIPE: More debug output to loading
[22:58] <GitHub126> scummvm/master 1408ffc Eugene Sandulenko: FULLPIPE: Better debug output for movement
[22:58] <GitHub126> scummvm/master 7724e51 Eugene Sandulenko: FULLPIPE: Added debug output for GameObject
[22:58] GitHub126 (GitHub126@192.30.252.40) left #scummvm.
[22:59] <snover> rsn8887 still needs fixing, did you learn anything about what section in the elf is causing trouble?
[22:59] --> Strangerke joined #scummvm.
[22:59] <wjp> what's so special about flac?
[23:00] <rsn8887> snover: no I had no time to do some further digging. I am just wondering if I should use new buildbot to build the 2-0-0 release since it includes so many libraries.
[23:00] <rsn8887> I just feel 2-0-0 stable should include as many libraries as possible?
[23:00] <wjp> well, yes
[23:01] <wjp> but why does flac fail while others don't?
[23:01] <wjp> it just uses autoconf/libtool
[23:01] --> GitHub22 joined #scummvm.
[23:01] <GitHub22> [scummvm-tools] criezy tagged v2.0.0 at 15e4a07: https://git.io/vb4Eg
[23:01] GitHub22 (GitHub22@192.30.252.37) left #scummvm.
[23:02] <wjp> (version numbers use dots, by the way, so 2.0.0)
[23:02] <rsn8887> Oh you mean... well on Vita it is crashing unless I configure it with --without-PIC. No clue why. It seems complicated. Can it have something to do with using stripped or non-stripped libraries?
[23:03] <wjp> only if there's a toolchain bug
[23:03] <wjp> but it compiles and links fine, and then crashes?
[23:05] --> GitHub50 joined #scummvm.
[23:05] <GitHub50> [scummvm-tools] criezy pushed 1 new commit to master: https://git.io/vb4Ej
[23:05] <GitHub50> scummvm-tools/master d5c1cd0 Thierry Crozat: TOOLS: Update version to 2.1.0git
[23:05] GitHub50 (GitHub50@192.30.252.40) left #scummvm.
[23:05] <rsn8887> Oh I forgot, it doesn't crash, it fails during .elf generation, "illegal relocation type 25." I read somewhere stripping libraries sometimes can cause crashes. And some Vita libs are stripped.
[23:05] <rsn8887> *crashes->problems
[23:07] <wjp> I've seen broken strip tools a decade ago, but not recently. But who knows.
[23:08] --> Lightkey joined #scummvm.
[23:10] <wjp> snover: speaking of libraries, do we have any way of checking builds have the "right" set of features enabled?
[23:10] <wjp> if we do automated builds it would be good to ensure we don't accidentally disable features
[23:11] <wjp> (by outdated libraries, broken detection, ...)
[23:12] --> macdude22 joined #scummvm.
[23:15] <wjp> I don't immediately see a straightforward way of doing this
[23:17] <snover> Sorry, on phone so bad typing ability
[23:17] <snover> Assuming that testers will report issues and defaults are ok
[23:17] <criezy> We still don't have an automated way for porters to upload their binaries to the server, right?
[23:17] <wjp> something like an accidentally disabled libmpeg2 will be very hard to spot if you're not looking for it
[23:18] <rsn8887> about dialog?
[23:18] <wjp> and in theory it is a very automatable test
[23:18] <wjp> rsn8887: that falls under "looking for it"
[23:19] <-- macdude22 left irc: Quit: Textual IRC Client: www.textualapp.com
[23:19] <snover> definitely addable. Was wondering about similar for libs w optional deps too
[23:24] <-- Mia left irc: Read error: Connection reset by peer
[23:26] --> Mia joined #scummvm.
[23:26] <-- Mia left irc: Changing host
[23:26] --> Mia joined #scummvm.
[23:27] <-- Farmboy0 left irc: Remote host closed the connection
[23:29] --> TMM joined #scummvm.
[23:29] <-- TMM left irc: Changing host
[23:29] --> TMM joined #scummvm.
[23:29] #scummvm: mode change '+o TMM' by ChanServ!ChanServ@services.
[23:32] <rsn8887> is it possible to find out what .h files a library installs on make install without actually running make install?
[23:33] <wjp> which library?
[23:33] <rsn8887> SDL2_net
[23:33] <rsn8887> for now
[23:34] <criezy> And I'm done for tonight. Hopefully I didn't make any mistake.
[23:35] <wjp> rsn8887: if you run 'make DESTDIR=/tmp/somewhere install' it'll do the make install into another directory
[23:35] <wjp> don't know if that's sufficient for your purpose
[23:36] <TMM> rsn8887, there's fakeroot and checkinstall that can kind of do that
[23:37] <rsn8887> I see thanks. Maybe DESTDIR is enough.
[23:37] <rsn8887> Also _nothing_ has uninstall, so how exactly does one do a clean uninstall without hacking around with these kinds of commands for hours?
[23:38] <rsn8887> my hatred for configure/make grows daily.
[23:38] <rsn8887> basically by the time I can cleanly install, package and uninstall I know enough about the library source that I can write my own Makefile and configure script.
[23:39] <TMM> if people use auto* correctly make uninstall should work by default
[23:39] <wjp> uh, SDL2_net has 'make uninstall'
[23:40] <TMM> auto* may be a little cumbersome if you're not used to it but what it's not is *user* unfriendly ;)
[23:40] <TMM> somewhat dev unfriendly if you follow the manuals from the 90s
[23:40] <wjp> but if you're doing something that requires uninstalling packages you may be doing something wrong
[23:41] <rsn8887> and the first commandline on latest SDL_Net file "Makefile.am" says # Makefile.am for the SDL sample image loading library and viewer
[23:42] <rsn8887> How is SDL_net a "sample image loading library and viewer"?
[23:42] <wjp> that would be SDL_image that they no doubt copied this from
[23:42] <rsn8887> Oh my
[23:43] <rsn8887> maybe I should have said "my hatred of SDL" not auto*
[23:43] <wjp> take a deep breath
[23:44] <rsn8887> btw: does the gd cloudsave work on 2.0.0? I couldn't get it to work on Vita, but it is probably some incompatibility with the handheld.
[23:45] <wjp> don't enable cloud features
[23:45] <wjp> they're disabled for the release, and you shouldn't re-enable them for release builds
[23:55] <rsn8887> oh I see. So no need even for SDL_net then?
[23:55] <wjp> right
[23:56] <rsn8887> Ah well probably good that I tested it anyways...
[23:56] <rsn8887> Would be nice especially on handheld...
[23:56] <wjp> yes
[23:57] <wjp> testing and auditing of this code would be very welcome :-) *hint*
[00:00] --- Sat Dec 9 2017