[01:34] anyone know why my pages are overlaping the Header? [01:34] I'm using PageStack [01:38] the reddit page is private? https://www.reddit.com/r/UbuntuPhone/ [01:45] https://www.youtube.com/watch?v=92voAhlZeHY === chihchun_afk is now known as chihchun === chihchun is now known as chihchun_afk [06:47] good morning [07:00] hello all [07:17] Good morning o/ [07:17] o/ [07:18] anyone know who the ubuntu-file-manager app developer is? I know popey worked on it. I have a patch that adds thumbnails for jpegs, png's, and gifs, but autopilot is failing 10 tests [07:18] however, autopilot is failing 10 tests even without my changes === chriadam is now known as chriadam|away [08:08] good morning all [08:09] Is there anyway from QML/C++ (without compositor) to define on wich display (monitor) we can display a Window ? [08:12] lets say i have monitor1 and monitor2 and i want programatically to show a Windows on one of them [08:30] zubozrout: hi [08:30] Ok, I am here mcphail [08:30] mcphail: Hi :0 [08:30] * :) [08:30] :) - how much experience do you have making .click packages? [08:31] mcphail: practically zero, I only made some stuff using the SDK [08:32] zubozrout: we don't have a working sdl template for the SDK yet, but that doesn't matter. Building and creating a .click is easy. SturmFlut is working on the template [08:32] mcphail: Well, "working" is a bit much to say ;) [08:33] Well, it would be nice to have it documented at least. [08:33] SturmFlut: :) [08:33] zubozrout: the first thing is to set up a build environment. Give me a sec and I'll find a link === dbarth_ is now known as dbarth [08:34] zubozrout: this is essential reading: https://developer.ubuntu.com/en/blog/2015/04/13/inner-workings-sdk/ [08:35] zubozrout: getting a chroot set up takes time, but it is very automated and foolproof [08:35] mcphail: ok, thanks :) [08:36] zubozrout: the next thing you have to do is get a version of SDL which works with Mir. See: https://raw.githubusercontent.com/mcphail/ubuntu-touch-sdl-template/master/lib/src/how_to_build_sdl_for_ubuntu_phone.txt [08:36] zubozrout: you can take the SDL libs from that github repo. They "wrok" [08:36] *work [08:37] mcphail: All right, that is fine to know. So it is not just as simple as apt-get install somesdlstuff and then running it. [08:37] zubozrout: not quite, but when you have the framework set up is is simple [08:38] zubozrout: the difficulty is getting the first steps done [08:38] zubozrout: you can then compile or download the ARM binary for your SDL app and roll it in a .click package [08:38] mcphail: well, the app has to be compiled with the above SDL libs, yes. [08:39] zubozrout: you can compile _against_ any SDL2 libs, but you have to roll that specific version into your .click [08:40] zubozrout: that means you can take any precompiled binary from the Ubuntu repos, but roll it with my SDL2 [08:40] mcphail: Ah :), ok, thanks :) [08:40] zubozrout: of course, the biggest problem is that SDL2 still doesn't quite work... [08:41] zubozrout: there are issues with coulours, transparency and rotation which are still getting smoothed out [08:41] bschaefer is on the case [08:42] popey what happened to the ubuntu update podcast? [08:43] mcphail: Yes, I know of some issues. And what will happen once click is replaced with snappy packages on Ubuntu? [08:43] zubozrout: not much difference [08:43] zubozrout: for these purposes, click and snappy are almost identical [08:44] (i think) [08:44] mcphail: ok :) [08:44] zubozrout: the hassle with building a .click is you have to bundle all the needed libraries into the click itself. Snappy will be the same [08:45] mcphail: Well, thank you very much for the info. I'll read it and try it - and then ask here again if I bump into any problem [08:45] zubozrout: enjoy! - it is easier than it looks [08:45] mcphail: Thanks :), you were very helpful. [09:54] mzanetti: ping :) [09:58] An Ubuntu phone will launch in India! [09:59] mzanetti: (or anyone who knows) what I understand in OTA-5 there will be full shell rotation and if I want to force for example portrait orientation for my app, I need to specify that in .desktop file. (thanks popey) Will there be an option to specify app orientation when the app is already running? === dholbach_ is now known as dholbach [10:10] SturmFlut: The meizu? [10:10] http://i.imgur.com/TZPiI8B.png << :) [10:12] akiva-thinkpad: Most likely, don't think it would make much sense to assume it's bq. We already know that Meizu wants to launch the MX4 with Flyme in India. [10:12] SturmFlut: lucky :/ [10:12] I still don't have one. === greyback__ is now known as greyback [10:12] nothing in North America~ [10:13] Hopefully by the end of the year! [10:21] akiva-thinkpad: Don't really think so, there hasn't been any committment from a manufacturer for the US market yet. And porting to a new device takes time. [10:22] SturmFlut: Well it will come here eventually. No reason to delay it~ [10:23] SturmFlut: And what about MX5 with Ubuntu? I suppose no one wants to spread the word now, but I would soo much love to see it with Ubuntu. I wanted to purchase MX4 but MX5 is better :), more bleeding-edge. [10:24] is it that much better :o [10:24] ? [10:27] akiva-thinkpad: Not to mention 64bit processor - I wanted to have a phone with that for a long time and if MX5 was available with Ubuntu I wouldn't have to wait another round :). [10:28] akiva-thinkpad: so I think it is great, in my eyes [10:28] :O But does it work in Canada?! [10:28] akiva-thinkpad: Why wouldn't it? [10:32] zubozrout: north american has different bands than the rest of the world [10:32] mx4 does not have our band. [10:32] best we can get is 2g [10:34] akiva-thinkpad: Well, I know there is a problem with LTE and MX4 + MX5 in the Czech Republic (where I live), so with the supported frequencies I can only have LTE in the capital (Prague) but almost nowhere else although I have LTE with my Nexus 7 deb tablet all over the country. [10:35] lucky europeans :P [10:37] akiva-thinkpad: Well, maybe lucky europeans except for the Czech Republic. I think Canada has better (nicer) mobile carriers than are here. [10:37] We have wind, thats it === chihchun_afk is now known as chihchun [10:37] Rogers... oh my blood pressure... [10:38] "What do you mean I owe you $500 ? You cancelled my phone 5 months ago; I couldnt even use it!" Reply: [10:38] akiva-thinkpad: hmm, and no Vodafone or T-mobile? I know they are in the USA. [10:39] "oh no, we just cancelled your service, you still need to pay for your plan" [10:39] No, and the best network in the US is Ting [10:39] which ironically is from Canada, but can't operate in Canada [10:39] akiva-thinkpad: lol, well, that's sad. Yep, I heard about Ting, it has pretty nice reviews. [10:39] mivoligo, not in the beginning, but there are plans for such an api [10:40] mzanetti: thanks, is that the rest I said correct? [10:41] zubozrout, akiva-thinkpad: At least at the moment there isn't any information regarding the MX5, and Ubuntu also doesn't officially run on 64-bit ARM yet. I've asked about 64-bit ARM, because it would be nice if the Convergence device by bq came with it, but no info. [10:41] mivoligo, yep [10:41] SturmFlut: 64bit would just add a bit more speed, and +2gb of ram? or is it different than x86? [10:42] mzanetti: ok, so "automaticOrientation: false" is no longer supported? [10:42] exactly [10:42] SturmFlut: Hmm, well, it sounds like I shouldn't hold my breath if there is no 64bit ARM support yet. So even if MX5 is going to be available with Ubuntu it won't be that soon as I want it to be, right? [10:43] akiva-thinkpad: The AArch64 mode of ARMv8 is completely different than ARMv7. It's not an extension, it's a new design. [10:43] mzanetti: btw what about the lock screen, does it rotate as well? [10:43] mivoligo: haha, no [10:44] SturmFlut: sounds complex~ [10:44] mivoligo: https://bugs.launchpad.net/bugs/1471609 [10:44] Launchpad bug 1471609 in unity8 (Ubuntu) "Cannot unlock phone in landscape" [Critical,Confirmed] [10:44] popey: :D [10:45] ^^ [10:47] mzanetti: are you planning to update MvsM any time soon? [10:48] zbenjamin: yeah, fun, SDK works just fine as long as I don't touch the laptop's internal touchpad :S [10:51] mivoligo, probably yes [10:52] mzanetti: I'll try to update that level 23 and push it in the evening then [10:52] ok [10:54] akiva-thinkpad: Okay, I was wrong I hear, the bq Convergence device should be 64-bit :) [10:54] SturmFlut: So the Aquarius that is released is 64bit? [10:54] SturmFlut, who told you that ? [10:54] (then it cant be arm ) [10:56] ogra_: akiva-thinkpad ment 64bit arm [10:57] Well its all greek to me [10:57] havn't been following hte arm hardware scene. [10:57] the* [10:57] well, thats more a software issue :) [10:57] only about 60% of the phone packages even build on arm64 yet [10:58] ogra_: So it is in works to support arm64? [10:59] the ubuntu archive has arm64 support since ages already ... [10:59] meh, all I care about is autopilot... and why trunk fails for me, but not for jenkins. [10:59] but the packages havent been fixed to build on it ... [10:59] nobody bothered [10:59] partialkly they even have arm64 completely disabled [11:00] getting that fixed and getting arm64 images at all will be quite some developer effort that will draw a lot of manpower [11:02] ogra_: Well, this answers my question about MX5 releasing with Ubuntu :). Sad, I wanted to wait for it even if it meant few months but this sounds like it wont be ready for another year or so. [11:02] zubozrout, well, its not impossible [11:02] ah I think I figured out why... [11:03] technically everything is there... but i dont see us having the spare manpower to get all the missing pieces in place in time [11:04] popey: import Ubuntu.Components 1.2 << that is the current state right? [11:04] the current version? [11:04] because jenkins is complaining: file:///usr/share/filemanager/qml/components/FolderListDelegate.qml:19:1: module "Ubuntu.Components" version 1.2 is not installed [11:04] import Ubuntu.Components 1.2 [11:05] akiva-thinkpad: jenkins is running on utopic :( [11:05] balloons: when you wake, any more poking we can do to move to vivid? [11:05] mzanetti: one more question: what about forcing screen saver not to show up, is it implemented in OTA-5? [11:05] yes [11:05] popey: so... just give it a few days? [11:06] mzanetti: \o/ [11:06] going to drop bigmovingtext from openstore [11:06] mivoligo, you should change your code to not always do "screensaverEnabled: false" [11:07] mzanetti: I'm working right now on the update :) [11:07] but rather: screensaverEnabled: textAnimation.running && Qt.application.active [11:07] something like that would be best imo [11:07] ogra_: I understand, ... but maybe it would be best for me to buy something cheaper than MX4, so BQ ... stick with it for a while and then switch to something better once it is available. And I must agree with SturmFlut that amr64 would be gorgeous for the converfence device. [11:07] * arm64 [11:08] mzanetti: thanks, is it from: "import QtSystemInfo 5.0"? [11:09] oh btw popey, I just discovered your podcast; I love it. Anyways gotta run~ See you all [11:09] o/ === chihchun is now known as chihchun_afk [11:54] Has anyone had any success setting up an emulated environment for comiling ARM binaries on AMD64 (rather than cross-compiling)? Have been trying with debootstrap and qemu-user-static but I am hitting roadblocks. How does the PPA build system achieve this? === _salem is now known as salem_ [12:26] which apt mirror hosts armhf binaries? [12:27] ports.ubuntu.com/ubuntu-ports [12:27] ogra_: cheers [12:29] mcphail: I'm tempted to buy a cheap arm chromebook for doing stuff like this [12:29] popey: I'm 95.989% of the way to solving it [12:29] \o/ [12:29] can you make an image and put it online somewhere? :) [12:29] popey: yep [12:29] \o/ [12:29] popey: if it works... [12:29] we have faith in you [12:30] is it working yet? [12:30] etc [12:30] the "royal" we? [12:30] ya [12:41] renatu: when you get a moment - https://bugs.launchpad.net/sync-monitor/+bug/1472973 - one of the insiders reported issues syncing [12:41] Launchpad bug 1472973 in sync-monitor "calendar fails to sync google calendar" [Undecided,New] [12:54] popey, thanks, I added a comment on the bug report [12:54] popey: my base image is 693M. Don't think I'm going to be able to host that anywhere useful. Would be faster for anyone to reproduce the build... [12:55] mcphail: if it's reproducible I can do that and host it [12:55] ping me the instructions and I'll take a look [12:56] popey: in channel or in /msg? [12:56] whatever :) [12:56] a paste somewhere or something is fine [12:56] popey: hokay [12:57] ta [13:10] popey: think this covers it - http://paste.ubuntu.com/11848280/ . That should give a base armhf vivid install. Would need to add the development packages etc [13:11] kk, thanks [13:11] popey: gzipped base image is about 350MB [13:12] qemu-user-static is a great little invention [13:14] ya [14:20] Is the Ubuntu-SDK PPA built for ARM? [14:23] mcphail: the ppa has that enabled, yes [14:23] good [14:26] If I don't install the SDK, is the "click" package in the standard repos sufficient to build phone packages? [14:38] mcphail: I don't always build click packages, but when I do, I use the ppa. [14:38] ok, I'll see how I get on [14:39] jhodapp: question - we now have the ability for apps to intelligently prevent screen off (e.g. in a game) - how about in mediaplayer - when watchin a video in a browser, I don't want the screen to go off. [14:39] is that on your roadmap? [14:43] popey, the oxide team is actually adding that right now [14:43] sweet! === charles_ is now known as charles [15:09] nemo: ping [15:22] mcphail: \o/ [15:22] got it working?? [15:22] nemo: need to remind me of the tweaks you want [15:23] mcphail: set MOBILE flag in build, and delete line 43 of hedgewars/options.inc [15:23] nemo: great. I'm on it [15:24] cool [15:24] * nemo crosses fingers [15:24] mcphail: just to doublecheck. line 43 is the HWENGINE one ☺ [15:25] since the desktop frontend expects a separate process right now [15:25] er [15:25] HWLIBRARY [15:25] CMake Warning: Manually-specified variables were not used by the project: MOBILE - is that a problem? [15:27] (btw, I'm using the source tarball in this chroot rather than the git repo...) [15:27] popey, here's the MR in fact: https://code.launchpad.net/~justinmcp/oxide/1328494/+merge/263998 [15:31] mcphail: ack [15:31] srs? [15:31] hm [15:31] wierd... [15:32] mcphail: checking cmake setup for engine [15:32] also pinging koda [15:33] mcphail: since we shifted engine to use cmake build it has had more weird issues and a bit more opaque a build process [15:33] (like failing to rebuild on code changes ☹ ) [15:35] nemo: that's a big problem for a build system... [15:35] no kidding [15:36] spent like an hour once trying to figure out why the hell my change wasn't fixing a bug [15:36] before the obv occurred to me [15:37] well this is frustrating. I don't see any toggle in CMakeLists.txt for this. might have to manually add it to that file ☹ [15:37] but let me poke koda about this [15:37] mcphail: that is, I don't see a add_definitions(-dMOBILE) anywhere [15:37] Bah - it has linked against SDL1.2 :( [15:37] mcphail: only desktop version [15:37] mcphail: mobile has always built against SDL2 [15:38] mcphail: the CMakeLists.txt already has: [15:38] if(${SDL_VERSION} VERSION_GREATER 1.3) add_definitions(-dSDL2) [15:39] endif() [15:39] but SDL2 is enabled if you can set MOBILE [15:39] hopefully linking is same. seems to be [15:39] nemo: I just used "-DMOBILE=1" as a build flag. Was that the correct way? [15:39] mcphail: that's for a cmake flag, but I don't see it being passed on to the engine there [15:40] mcphail: I *thought* our CMakeLists.txt would have relayed it. that's what's annoying me [15:40] I'm thinking the lack of mobile devs in past few years has triggered bitrot as koda messed w/ CMake [15:41] mcphail: you could just add add_definitions(-dMOBILE) to hedgewars/CMakeLists.txt for now [15:41] pending consulting w/ him on this [15:41] maybe just above add_definitions(-dDEBUGFILE) [15:41] on line 209 [15:43] nemo: do I need to delete the Makefile after changing CMakeLists.txt? [15:43] mcphail: should detect the change [15:43] if you rerun cmake [15:43] hell. if you run make [15:43] ok :) [15:45] mcphail: what worries me though, is because this is required, it makes me think the CMakeLists.txt currently has no mobile support. which is bad for linking [15:45] mcphail: I do hope since we ocassionally screw around with SDL2 that that should work at least ☹ [15:45] nemo: I'm going to do a clean build because this is throwing multiple errors [15:45] find_package(SDL1or2) [15:45] hopefully it finds 2 ☺ [16:24] nemo: getting stuck with this one: http://paste.ubuntu.com/11849555/ - any ideas? At least (I think) this suggests it is using SDL2... [16:24] oh good [16:25] mcphail: that means it is hitting the MOBILE block in the engine which is adding the gles11 unit [16:26] mcphail: now, in the old android build, a gles11.pp was provieded in project_files/Android-build but ideally we'd get this from ubuntu's package management [16:26] mcphail: I'm checking to see if you guys do in fact package this anywhere ☺ [16:26] mcphail: otherwise we could toss it into the dir [16:26] ok [16:26] nemo: as I said. I don't have any particular knowledge on the issue. but figured can't hurt to hang out here :P [16:26] hi sheepluva [16:26] trying to coax koda here but apparently hectic day at work for him [16:28] hum... not finding it anywhere ☹ [16:29] 18:26 <@koda> needs gles11 unit [16:29] 18:26 <@koda> should be checked under fpc unit folder [16:29] sheepluva: the only gles11 unit I have is that android custom one [16:30] it binds against External_library='GLESv1_CM'; {Setup as you need} [16:30] Loadgles11('libGLESv1_CM.so'); [16:30] sheepluva: so those would need changing at a minimum [16:30] mcphail: hm.. does ubuntu touch need gles11? ☺ [16:31] sheepluva: you know, this kinda surprises me 'cause I thought we were using GLES11 across the board now [16:31] nemo: I honestly don'y know :) [16:32] {$IFDEF MOBILE} {$DEFINE HWLIBRARY} [16:32] {$DEFINE GLunit:=gles11} {$DEFINE USE_LANDSCAPE_AMMOMENU} {$DEFINE USE_TOUCH_INTERFACE} [16:32] {$ELSE} [16:32] sheepluva: boooo [16:32] sheepluva: 1) we got that far already, 2) pastebin 😝 [16:32] nemo: off to do some domestic chores for a while. If you get a solution, I'll catch it in the /lastlog [16:33] does the build destination? the ubuntu touch thing? doesn't have regular gl? [16:33] sheepluva: well ubuntu touch targets mobile devices, so I'm gonna guess GLES is standard on the device. but I had no idea if maybe it had an abstraction layer or something [16:33] sheepluva: also, I know that MOBILE sets gles11 but I wasn't sure if that was redundant or not [16:35] hi [16:35] sorry [16:38] well. eyeballing the flag, seems gles11 only impacts our code in extension loading [16:38] so. assuming opengl apps normally build on an ubuntu touch, and opengl is available.. [16:39] gonna guess we could maybe remove GLunit:=gles11 and, possibly if it causes a problem on startup, kill the extension loading [16:40] we don't use it for anything important anyway [16:40] video recording and 3d rendering [16:40] mcphail: ok. you mind deleting that line too? ☺ [16:43] mcphail: you mean that you are using full fledged gl unit in fpc ? :o [16:43] hi [16:44] how can i create a button in qml for ubuntu touch? [16:45] koda: well... my problem was that the gles11 stuff seemed custom to android (and maybe iOS) [16:45] koda: gles11 isn't provided by the system [16:45] so I was hoping we could get away w/ full gl unit [16:45] no it's custom to *mobile* opengl [16:45] there are missing functions between gles11 and gl [16:45] koda: I mean gles11.pp is under Android-build [16:46] koda: uh. why isn't that reflected in the code? [16:46] only thing that gles11 seems to flag is extension loading [16:46] I thought we targeted GLES11 across the board [16:46] and that it was a subset of full gl [16:46] because the code uses gles11 which matches gl1.0 [16:46] so it works fine on desktop and amobile [16:47] koda: right? thought that's what I was saying [16:49] koda: but my point is... if system has gl unit in fpc, then it should just work if gles11 unit is deleted [16:49] koda: if it doesn't, because ubuntu touch has phone issues on that front, then will need to adapt the android build unit right? [16:50] since I couldn't find an ubuntu package explicitly for it, like, say, fp-unit-gles [16:50] i don't think so, gles11 comes bundled in the arm fpc package afair [16:50] oh realllly [16:50] then he should not have gotten a build error eh [16:54] it depends how the fpc pkg is distributed on ubuntu mobile :p [16:56] * nemo pokes mcphail [16:56] you mind checking to see if gles11 is under the fpc tree on your build machine? [16:56] or, for that matter, any gl unit at all? [16:57] that is, under /usr/lib/fpc/*/units probably [17:02] koda: the thing is, I *think* this package already exists for ubuntu touch [17:02] so that suggests someone is building the entire ecosystem [17:02] well. maybe I misunderstood. [17:02] nemo: no gles under /usr/lib/fpc/ anywhere [17:02] perhaps it is just packaged for arm, and if it works under ubuntu touch is just luck [17:02] mcphail: how about opengl [17:03] nemo: yes - but can't use that on phone [17:03] bleh [17:03] mcphail: ok... so. the android build used a custom gles11.pp that added bindings against a gles so [17:03] it would be *nice* if ubuntu could package this in fashion koda mentioned ☺ [17:03] but failing that, could do it manually I guess [17:04] nemo: :) [17:04] mcphail: I was kinda hoping that possibly ubuntu was providing an opengl stub against gles11 - for apps that might possibly work w/ it. [17:04] mcphail: hm... [17:04] koda: hey. question. [17:05] koda: if the build was against the gl fpc, but we only used gles11 routines... [17:05] er. wait. no... would require the gl lib to target the phone gles. nm [17:06] guess that unit isn't even built for touch [17:06] probably would be missing stuff [17:06] nemo: I don't know anything about pascal so can't offer any suggestions [17:07] mcphail: well... if I knew the name of the gles11 library on the phone could try adapting https://code.google.com/p/hedgewars/source/browse/project_files/Android-build/gles11.pp I guess ☹ [17:07] mcphail: could file a bug against ubuntu touch asking for that mobile targetted arm package, but I guess that's lol [17:07] nemo: just the name of the package giving the C libs for gles? [17:08] mcphail: well... reading that gles11.pp says it was automatically generated [17:08] but, most of it seems pretty generic [17:08] so I was hoping I could simply update Loadgles11('libGLESv1_CM.so'); to be your libgles name [17:09] then toss gles11.pp into hedgewars/ [17:09] or well. have you do both of those 😝 [17:10] libGLESv2.so.2.0.0 and libGLESv1_CM.so.1.1.0 [17:10] ooooh [17:10] srs! [17:11] cool. so maybe don't have to do anything [17:11] I bet libGLESv1_CM.so.1.1.0 is probably symlinked to libGLESv1_CM.so [17:11] mcphail: ok. cp project_files/Android-build/gles11.pp hedgewars/ [17:11] aaaand cross fingers I guess [17:12] nemo: well, we get a _different_ error :) [17:12] yay [17:12] 😝 [17:13] nebtw the gles11.pp is just a copy of the fpc version, so there is no problem in bundling for ubuntu touch imho [17:13] nemo: http://paste.ubuntu.com/11849907/ [17:13] hmmm [17:14] ♥ hg grep [17:14] is teh awesome [17:14] specifically hg grep --all [17:15] thought so. sheepluva added that last june [17:15] symptom of bitrot in our mobile stuff [17:15] good news tho [17:15] mcphail: he just failed to ifdef enough I think [17:15] {$IFDEF USE_S3D_RENDERING OR USE_VIDEO_RECORDING} [17:15] aah. [17:16] is around the body of the createframebuffer [17:16] but not the header [17:16] nemo: paste me a patch or new file and I'll switch it around [17:16] mcphail: hm... [17:16] mcphail: well. I'm not sure what you are using. the branch? [17:17] I just want that ifdef above around those 2 calls [17:17] on your lines 81 and 82 of uRender [17:17] mcphail: 0.9.21 + ubuntu patches? [17:18] nemo: just the 0.9.21 tarball from the site [17:19] 'k [17:19] hang on then [17:20] can I just comment those lines out? [17:20] you could, but I was testing locally and hit another one [17:20] USE_S3D_RENDERING is not enabled for mobile, and RenderClear has same issue [17:20] procedure RenderClear (mode: TRenderMode); [17:20] will need that too [17:21] ugh. DrawWorld too. maaaan [17:30] mcphail: ok. made some minor changes to dev for those flags [17:30] mcphail: going to transplant them into the .21 branch and pack 'em up for you as a diff [17:30] just eyeballing it *seems* like the game should still work just w/ the ifdef tweaks ☺ [17:31] transplant 'cause I think the branch is too old to merge anymore [17:41] Months ago we couldn't even get SDL working, and now you guys are working on the third SDL game port I know of. Oh wow. [17:43] mcphail: ok. builds w/ s3d and video recording turned off now. transplanted it onto the branch. but not worth you getting it I think [17:43] mcphail: simply disabling those 2 lines is enough on the branch. all the rest is in .22-dev [17:44] well. "the rest" one more function and a compiler error in uWorld [17:44] https://code.google.com/p/hedgewars/source/detail?r=a0b9b678b46eccd0c66273c4f6bb016b9586d2be&name=0.9.21 fwiw [17:45] but nice to get mobile bitrot under control again \o/ [17:45] mcphail: we need to properly support MOBILE as a cmake flag, and maybe provide gles11 if the system isn't providing it [17:46] don't want to add gles11.pp outright ofc, to avoid overriding system version [17:52] nemo: was afk. Give me a minute and I'll check that out [17:54] SturmFlut: TBH I thought this was just gonna be flip a flag and go ☺ [17:54] SturmFlut: or I would have probably stayed in my channel 😝 [17:54] but nice to see all that mobile work from iOS and Android days so many years ago actually get some use [17:55] mcphail: hm... gotta go in a couple of minutes [17:55] or I'l get stuck in traffic for sure and miss kid's class [17:56] nemo: np - will have to go soon anyway. Got a bit further then - http://paste.ubuntu.com/11850158/ [17:56] huh... [17:57] koda: ↑ [17:58] don't see where that should be coming from [17:58] but gotta go [17:58] nemo: 365: fingers[index].id := nilFingerId; [17:59] some leftover probably [17:59] mcphail: should be enough to remove it [18:02] ok - it compiles but I'm now getting a raft of SDL linker errors. Might have to give up on this for the evening [18:04]