**** BEGIN LOGGING AT Tue Oct 08 02:59:57 2019 Oct 08 05:19:40 good morning Oct 08 05:25:40 i'm having segfaults. so, disagreed. Oct 08 05:27:30 rostam: isn't that just a flag in the fstab? Oct 08 05:46:48 mornings Oct 08 05:47:14 agreed, i'm struggeling with kernel module semaphores Oct 08 05:47:34 but at least i've got coffee! Oct 08 07:14:19 LetoThe2nd: If I may suggest, I think it could be a good idea to put the main topics of the live coding session in the title of the "Live Coding with Yocto Project" Youtube videos Oct 08 07:14:56 ndec: ^^^^^^ Oct 08 07:15:22 qschulz: i don't have access there, but nico can probably take care of this. Oct 08 07:15:35 I think I put that in the description of each video, no? Oct 08 07:16:36 ndec: it's not visible in the excerpt so i have to start the video, click more in the description to know what's talked about Oct 08 07:16:37 right. i usually copy the content of the announcement/email in the video description. are you saying, this is not enough? Oct 08 07:18:08 ndec: "Live Coding with Yocto Project. Bring your questions, see Yocto Project developers step through live tutorials and share your ideas for next sessions. We talk, explore devtools and write code along..." which is identical for all videos Oct 08 07:18:27 (except video 1 :p) Oct 08 07:18:29 i can see that now. maybe I should remove that sentence indeed. Oct 08 07:18:52 just a suggestion, not a must-do :) Oct 08 07:19:08 Or just re-order the sentence so the topics are in the first line? Oct 08 07:19:48 though, having it in the title could help if the video gets recommended on the right side of the screen where only the title is displayed Oct 08 07:20:03 or on the main page of youtube Oct 08 07:20:39 good morning! Oct 08 07:21:32 ndec: actually my main problem with YT is that i don't get notified when there are comment which i am kind-of supposed to answer. and i really don't want to open each video like one a week, polling style Oct 08 07:34:16 LetoThe2nd: qschulz : how does it look now, better? Oct 08 07:35:27 ndec: thanks, looks better :) Oct 08 07:35:48 I moved the generic description to the playlist 'description', and added a short summary in the titles. Oct 08 07:35:50 ah. good! Oct 08 07:36:50 i'm surprised that we accumulated way over 2k watches by now Oct 08 07:37:24 LetoThe2nd: about your comment issue, that's indeed a problem.. i suppose if you upload the video, it would be fine, no? Oct 08 07:37:30 e.g. you would get notifications? Oct 08 07:37:43 LetoThe2nd: what I'm actually surprised about is that it's mainly the first video collecting all views Oct 08 07:38:41 ndec: probably. yet, don't spend too much time on it. i plan to discuss the future of the sessions in lyon, hopefully with you, behan, whoever feels like speaking up. then we'll see. Oct 08 07:39:42 qschulz: now thats something that doesn't surprise me in the least. people always start with the first, and with "getting started". then only those who actually like the format and feel like needing more information will continue. Oct 08 07:40:22 qschulz: so based on the number of the follow up views, i'd guess that we have somewhere in the 100-range of actually active viewers. Oct 08 07:45:35 which on the other hand is quite ok, i'd say. Oct 08 07:46:51 LetoThe2nd: sure, let's discuss that in Lyon! Oct 08 07:48:10 * LetoThe2nd checks schedule for the yocto bof Oct 08 08:03:33 What do you think about suppress WKS_FILE and change image class logic to produce as many images as items WKS_FILES contains? The use case could be the support of the SD cards of various size. Oct 08 08:04:17 A patch made toward that direction could be ever accepted? Oct 08 11:44:49 I'm start to think that I have just said a very stupid thing... :) Oct 08 12:01:18 alessioigor: I'm way to new to yocto to be able to answer your question. But I've had the same problem using raspbian. The usually way raspbian solves this is to have a small filesystem that will expand on first boot to take the whole sd card in use Oct 08 12:01:50 a good solution for raspian, but I like your suggestion better if you actually know which hardware you will run on in before Oct 08 12:03:29 alessioigor: it makes sense and we'd like such feature Oct 08 12:14:46 fredrigu: the mechanism adopted by raspbian is very smart and adapts to every device Oct 08 12:15:58 shouldn't be tooo hard to write a early in first boot expand file system script Oct 08 12:21:53 mckoan: yes it it, and for raspbian it's a very good solution. Oct 08 12:22:28 mckoan: for a system where the sd card size is known beforehand it might be nice to not have a special case for the first boot. Oct 08 12:24:35 fredrigu: Is the raspian approach realiable? Oct 08 12:33:12 I have already seen that type of script (which repartition and resize the filesystem at boot) in few VM oriented distro *but* in that context if a VM doesn't start you can launch another VM... Oct 08 12:37:35 RP: not sure if you are around these days .. but see the more email on the strace issue on lkml Oct 08 12:38:16 kanavin, you mentioned that v5.2 and strace was ok on what distro ? opensuse ? if so, do you happen to be able to get the .config ? Oct 08 12:41:42 hi there, i am new to yocto and so far love the concept. i currently go through some sample builds for a raspberry pi. Oct 08 12:42:04 during taht i discovered a compilation problem and also found a fix for that. Oct 08 12:42:32 now i am starting to go the 'devtool' way to bring in the changed and apply them to my current yocto/bitbake build Oct 08 12:42:41 so i started with Oct 08 12:47:45 devtool modify libgpg-error Oct 08 12:48:23 now i cannot modify libgpg-error-native anymore as long as i am testing changes for libgpg-error... even though it's the same recipe Oct 08 12:48:35 is that supposed to be like this? Oct 08 12:51:32 as they're the same recipe, changes to the file will affect both builds Oct 08 12:52:03 that's what i thought, but as long as i am testing my modifications it seems not to be like that Oct 08 12:52:29 i mean, once i created the bbappend patch and go for the build, it works for both variants Oct 08 12:53:27 but as long as i am testing my modifications, i kept running into errors and did not understand.. till i found that i have to run the update-recipe first Oct 08 12:53:43 was kind of weird, since i didn't expect it to be like that Oct 08 12:54:29 zeddii: I am around, yes Oct 08 12:55:00 RP: just replying to lkml .. I have a fix. Oct 08 12:55:14 well, Oleg sent a fix. but I just confirmed it about a minute ago Oct 08 12:55:32 zeddii: ah, great! Oct 08 12:55:45 zeddii: in that case I will defer M4 rc1 build until we get this sorted out Oct 08 12:56:07 yup. I'll make my own patch for linux-yocto, and let Oleg/Roman do the wrangling upstream. Oct 08 12:56:38 zeddii: sounds good. We're literally just in time as I planned to roll rc1 this afternoon :) Oct 08 12:56:53 zeddii: now we have a fix in sight its better to wait Oct 08 12:57:04 yah. I'll do one and send you SRCREVs in about an hour. Oct 08 12:57:44 zeddii: nice to have a resolution/understanding Oct 08 12:59:23 indeed Oct 08 12:59:35 and also another point of evidence that we can find kernel issues :D Oct 08 13:02:08 zeddii: yes, a rather nice one! Oct 08 13:03:31 one more thing regarding the devtool approach... is it normal that bitbake build more packages than necessary if i try to build one specific package? Oct 08 13:04:41 zeddii: just reading how a perl bug I found and reported is now, 9 months on causing issues in debian :) Oct 08 13:05:13 it looks like bb builds stuff that might be interesting for the whole image at a later point... even though i expilcitly called Oct 08 13:05:13 bitbake libgpg-error Oct 08 13:05:46 heheh Oct 08 13:06:18 creich, if things depend on libgpg-error, they'd be rebuilt if you patch/change it. Oct 08 13:06:39 ah ok. thx Oct 08 13:07:04 even thoguh i never built them before? Oct 08 13:09:06 no, that really shouldn't happen. I know that I don't always notice some things building the first time. but if you pastebin what built you didn't expect, someone might know why. Oct 08 13:09:26 you can also dump dependencies for a package, but I always have to lookup the syntax when I need it. Oct 08 13:12:37 zeddii, ok thx. i'll find out how to dump the dependencies Oct 08 13:12:59 zeddii, that was opensuse, yes - do you still need the .config? Oct 08 13:13:32 kanavin. nope. I have a patch that fixes it now, via the lkml thread. Oct 08 13:13:33 now that i think about it, it doesn't make sence that it should be packages depneding on the lib, since they started building already before the lib itself was finished ^^ Oct 08 13:13:50 kanavin, actually. I'd still like to see it if you can get it. Oct 08 13:14:01 as another confirmation on the patch Oct 08 13:14:10 but at least the fix works. and i get the main idea of the devtool. Oct 08 13:14:14 thx guys :) Oct 08 13:14:41 zeddii, I cannot obtain it right now, as it is my home machine Oct 08 13:15:18 kanavin, no worries then. I have a full fix in test now. that'll show me if it is right or not. Oct 08 13:17:31 creich: it might be stuff that it needs to build packages. like if you are building rpm packages you need rpm-native and you might have caused a rebuild of that Oct 08 13:17:36 RP: what perl issue? Oct 08 13:18:14 rburton, ok. thx for that info Oct 08 13:18:32 rburton: the $) one Oct 08 13:19:04 obviously purging anything perl from my memory Oct 08 13:19:44 rburton. not purged. it was write-once. just like anything perl. never read again. Oct 08 13:19:50 lol Oct 08 13:20:04 rburton: the setgroups one from 3 months ago (I'm clearly also trying to push it back in time) Oct 08 13:25:49 alessioigor: I can't say, I've only used it on a few occasions. Oct 08 13:44:09 Hi Guys. I have a question. I want a task to rerun based on a python function which dynamically gets a build info. For example a git revision. If this git revision changes, I want a certain task to rerun. I currently store the output of this function in a variable and I can see it in the task signature expanded correctly. The problem is that if I Oct 08 13:44:09 externally change the revision, even though the env shows the correct new value, the task is not rerun. Oct 08 13:46:35 https://www.yoctoproject.org/docs/2.6.2/bitbake-user-manual/bitbake-user-manual.html#var-BB_ENV_EXTRAWHITE or https://www.yoctoproject.org/docs/2.6.2/bitbake-user-manual/bitbake-user-manual.html#var-BB_ENV_WHITELIST? Oct 08 13:48:39 qschulz that imports from external env. What I want is a python function dynamically setting a variable based on which a task is rerun or not. Oct 08 13:48:55 ``` Oct 08 13:48:58 ``` Oct 08 13:49:18 (multiline issues) Oct 08 13:50:12 alephan: question is what exactly is your issue and why you need this? Might be the wrong way to work around the issue Oct 08 13:52:50 New news from stackoverflow: Yocto: How do I export environmental variables and files using relative path in local.conf? Oct 08 13:53:36 Right. So I want to include the revision of a repository in a file that gets deployed on the rootfs. Also, because of this repository might change its revision manually, I'd like the task to rerun when that happens. qschulz Oct 08 13:55:34 It's similar to a revision manifest that ends up on the rootfs. With some additional metadata. `os-release` like data but with some external sources revisions. Oct 08 13:58:31 if this repository is source for only one recipe you could add a file manually in do_install with the data you need and Yocto will re-run do_install anyway when the revision changes. Oct 08 13:58:56 qschulz it can be the revision of a layer. For example/ Oct 08 13:59:13 I want to have a manifest of all the revisions of all the layers I include in the build. Oct 08 13:59:27 alephan: or your task can have the vardeps flag enabled https://www.yoctoproject.org/docs/2.6/bitbake-user-manual/bitbake-user-manual.html#variable-flags Oct 08 13:59:34 I have that already. Oct 08 13:59:57 Let me paste a snippet Oct 08 14:00:33 I'm not sure vardeps works for functions, I'm a bit confused with which flags can be applied to functions also or only tasks Oct 08 14:00:50 so if it's simply a function, I don't know if it is ignored or not Oct 08 14:06:07 def compute_test(d):```get the revision of a layers and setVar TEST```TEST := "${@compute_test(d)}"python do_compile() {```write TEST to a WORKDIR file```}do_compile[vardeps] += "TEST" Oct 08 14:06:53 https://pastebin.com/1duS3hZe Oct 08 14:07:34 Sorry for the formatting. Oct 08 14:10:23 bitbake-dumpsig -t test do_compile | grep "TEST value is"Variable TEST value is 89 Oct 08 14:13:10 No way to paste something formatted here. In this pastebin https://pastebin.com/7PhiCZTb you can see that the recipe environment gets the correct value, the existing signature includes the old value but the task is not retriggered. Oct 08 14:13:54 If I force the task, I get a `basehash value changed` error and the task runs with the new value as expected. Oct 08 14:43:13 Hello everyone Oct 08 14:45:47 The live session will start in 15 minutes (for those who are not on the tech sync call) - so time to get a drink and relax :) Oct 08 14:49:19 LetoThe2nd: great end of the day! Oct 08 14:49:56 * mckoan in the meanwhile is preparing slides for YP Summit :-D Oct 08 14:50:48 mckoan: i'm only preparing booze and trolling there. i come as a completely ordinary, paying attendee! Oct 08 14:51:28 Looking forward to it Oct 08 14:52:15 jeranio: welcome! :) Oct 08 14:52:42 Thanks this will be my first live view Oct 08 14:52:58 Going to make some coffee quickly Oct 08 14:53:58 LetoThe2nd: good luck, most likely not going to watch entirely live but supporting you :) Oct 08 14:54:32 qschulz: i doubt there will be something new for you in today. Oct 08 14:55:48 LetoThe2nd: You never know, there might be some information that you'll give like it's nothing and people actually don't know about it. Oct 08 14:56:36 For example, you presented bitbake -e very early in your streams but that something I discovered after a few months and I'm using it very often since then :) Oct 08 14:58:05 qschulz: everything is better with heavy metal. Oct 08 14:58:07 YPTM - armin is on Oct 08 14:59:25 YPTM ross joining Oct 08 15:00:06 YPTM scott joining Oct 08 15:00:20 YPTM Michael joined Oct 08 15:00:52 YPTM - Trevor is on Oct 08 15:01:25 YPTM - No Idea What's Going On joined Oct 08 15:05:06 LetoThe2nd: do you have a link please, I can't find it on the youtube channel Oct 08 15:05:10 YPTM Alex Kanavin Oct 08 15:05:13 yacar_: https://www.twitch.tv/yocto_project/ Oct 08 15:05:31 thanks ! Oct 08 15:05:55 if you have a twitch account, be sure to follow so you get notified for future twitch streams. Oct 08 15:09:01 RP: I have two Toaster patches pending, the second of last week's patch of two, and the the recent one about supporting Zeus Oct 08 15:21:44 JPEW: we had an ab failure with reprod thanks to that damned pod2man inserting versions into manpages Oct 08 15:22:06 JPEW: tempted to just sed those lines out globally.... Oct 08 15:22:24 rburton: Yep..... I though I patched that out of pod2man.... or is it the HOSTTOOL one? Oct 08 15:22:37 JPEW: right hosttools Oct 08 15:23:12 i wonder how much breaks if we remove pod2man from hosttools and mandate perl-native Oct 08 15:23:26 i hate having perl-native as a hard dependency on things though Oct 08 15:23:46 rburton: Agreed Oct 08 15:24:16 i wonder why git needs perl-native Oct 08 15:24:29 rburton: I hate the idea of perl-native as a key dependency Oct 08 15:25:08 agreed Oct 08 15:25:13 rburton: we could wrap it and kill the line... Oct 08 15:25:23 hm thats not a horrible idea Oct 08 15:25:37 YPTM - over Oct 08 15:26:23 i wonder if we can assume coreutils 8.16 yet Oct 08 15:32:38 rburton: git-email needs perl but that should be target perl Oct 08 15:32:52 right Oct 08 15:33:28 RP: I sent a v2 of perl/PACKAGES_DYNAMIC patch which does not over forcevariable and it works equally well Oct 08 15:36:58 LetoThe2nd: cheers! I'm leaving the office now, gl with the stream and see you in a few weeks at ELCE ;) Oct 08 15:39:48 amazing, git's build system is of course mental Oct 08 15:40:05 you can tell it where perl is at configure time and that means *target perl* Oct 08 15:40:14 you can't pass --without-perl as it says perl is needed Oct 08 15:40:24 but you can pass NO_PERL and it won't build anything with perl Oct 08 15:44:02 lol Oct 08 15:46:02 of course i'm not surprised in the *slightest* Oct 08 15:47:40 khem: I saw, thanks Oct 08 15:47:49 khem: will include in the next test build Oct 08 15:50:44 zeddii: Your patch doesn't apply, what am I missing? Oct 08 15:51:03 RP, when will zeus hit the layer compat? Oct 08 15:51:33 is that an rc1 thing? Oct 08 15:52:50 let me check. I had frozen my test branch a while ago. I'll rebase and send a v2. Oct 08 15:53:02 Are there linux-yocto changes in master-next ? Oct 08 15:53:17 I just started a full build against master, so I can restart it Oct 08 15:54:41 zeddii: there shouldn't be any Oct 08 15:55:33 I just checked out master-next. the -rt SRCREV bounced. Oct 08 15:55:34 will fix. Oct 08 15:56:29 zeddii: thanks. I wanted to understand why we didn't match :) Oct 08 15:58:55 hah. it's actually simpler. there's a bug in -rt that should be fixed. it is on my master, and not master-next. Oct 08 15:59:28 zeddii: right, I wondered if there was a patch that got lost Oct 08 15:59:31 so my fix is on top of that. we need both, since they are in the branches now. I'll send it to the list, and if you drop it in before my ptrace one, we are good. Oct 08 15:59:39 zeddii: sounds good Oct 08 16:01:56 RP: sent, you have the -rt bump in your inbox Oct 08 16:03:04 zeddii: thanks, all applies now :) Oct 08 16:12:23 RP, rburton I have been working to obsolete meta-gplv2, here's the outcome http://git.yoctoproject.org/cgit.cgi/poky-contrib/log/?h=akanavin/package-version-updates (socat and up) Oct 08 16:13:47 yes, this work is driven by the 'special needs' of the automotive industry :) Oct 08 16:14:21 hm i thought i did the gdbm one for python Oct 08 16:14:32 i wonder if thats in a branch somewhere Oct 08 16:15:58 i feel like i didn't break enough stuff during the seesion Oct 08 16:17:20 I will send these patches once RP opens the zeus branch Oct 08 16:17:30 awesome Oct 08 16:17:38 would be nice to deprecate the layer entirely Oct 08 16:18:18 rburton, that's the goal, we want to drop it from our local builds Oct 08 16:18:35 Does anybody have any idea about the issue above? (sorry for if barging into a conversation) Oct 08 16:18:55 how are you going to go about deprecating meta-gplv2 in cases where there's no bsd-ish replacement for a component? Oct 08 16:19:50 kergoth, by making the dependency on a component optional and then disabling it Oct 08 16:20:12 e.g. gdbm was mandatory for python, now it is optional Oct 08 16:20:51 OFFICIAL ANNOUNCE: we will have glitter stickers in Lyon :) Oct 08 16:20:56 or apparmor was packaging some bash scripts of less-than-necessary kind, so I splitted those into a PACKAGECONFIG Oct 08 16:23:36 hello folks Oct 08 16:24:01 why is CC exported to tasks differently than " " ? Oct 08 16:24:32 because CFLAGS goes in CFLAGS Oct 08 16:24:34 I.e CC seem to be set to " -m32 " Oct 08 16:24:52 export CC = "${CCACHE}${HOST_PREFIX}gcc ${HOST_CC_ARCH}${TOOLCHAIN_OPTIONS}" Oct 08 16:24:53 I thought CFLAGS is the c compiler flags, and CC is the compiler invocation line Oct 08 16:25:07 toolchain option is just where the sysroot is Oct 08 16:25:39 host cc arch is a few fundamental tune flags Oct 08 16:26:55 ah, I see in the manual there's docs for TARGET_CFLAGS and TARGET_CC_ARCH (the latter goes into HOST_CC_ARCH when building for the target) Oct 08 16:27:28 TARGET_CFLAGS: "Specifies the flags to pass to the C compiler when building for the target." TARGET_CC_ARCH: "Specifies architecture-specific C compiler flags for the target system." Oct 08 16:29:22 rburton, ah, I see now. TARGET_CFLAGS contains the architecture-independent flags. like, optimization level. Oct 08 16:29:43 so for a correct invocation of the c compiler, one should do ${CC} ${CFLAGS} , I think Oct 08 16:30:48 that makes sense. CFLAGS should not be passed when compiling a c++ file (but CXXFLAGS should be used then). but CC should be used for both C and C++ Oct 08 17:42:44 RP: looks like the strace fix is pretty much confirmed. victory. Oct 08 17:57:01 so how does zeddii do a victory dance ? Oct 08 17:59:17 ask me @ ELCe. It takes about 10 beer. Oct 08 18:00:51 hehe Oct 08 18:01:22 10 Candian beers or French beers ? Oct 08 18:04:13 wait there is such a thing as a 'French beer'? Oct 08 18:10:03 hehehe Oct 08 18:21:30 hello anybody knows if basez package exists in some meta-layer? Oct 08 18:21:47 looked in openembedded packages but didn't found anything Oct 08 18:22:07 have you tried the layerindex? Oct 08 18:22:20 armpit: what do you mean? Oct 08 18:22:25 http://layers.openembedded.org/layerindex/branch/master/layers/ Oct 08 18:22:32 yes Oct 08 18:22:36 I tried this one Oct 08 18:22:55 k. then its not in any register layer Oct 08 18:22:59 I'm trying to add it Oct 08 18:23:03 k Oct 08 18:23:08 but get issue with configure I'l autotools Oct 08 18:23:12 but get issue: Oct 08 18:23:23 | NOTE: Running ./configure --build=x86_64-linux --host=arm-poky-linux-gnueabi --target=arm-poky-linux-gnueabi --prefix=/usr --exec_prefix=/usr --bindir=/usr/bin --sbindir=/usr/sbin --libexecdir=/usr/libexec --datadir=/usr/share --sysconfdir=/etc --sharedstatedir=/com --localstatedir=/var --libdir=/usr/lib --includedir=/usr/include --oldincludedir=/usr/include --infodir=/usr/s Oct 08 18:23:23 hare/info --mandir=/usr/share/man --disable-silent-rules --disable-dependency-tracking --with-libtool-sysroot=/home/builder/build/tmp/work/armv7vet2hf-neon-poky-linux-gnueabi/basez/1.6.2-r0/recipe-sysroot --disable-static Oct 08 18:23:23 | Unknown option: --exec_prefix=/usr Oct 08 18:24:49 any ideas? Oct 08 18:26:49 I have not tried this but maybe CONFIGUREOPTS_remove = "--exec_prefix=/usr" or define your own Oct 08 18:28:48 it helps thanks Oct 08 18:28:55 I'll share recipe then on oe list Oct 08 18:42:07 armpit: With few other fixes recipe is ready ;) Oct 08 18:42:28 nice Oct 08 18:44:39 opennandra: it sounds like it might use a custom configure script, not an autoconf one. if that's the case, don't inherit autotools. check that. Oct 08 18:45:42 kergoth: not it have autogen.sh so I've inherited autotools Oct 08 18:48:06 not relevant Oct 08 18:48:33 i just checked myself, it uses a custom configure script, it does not use autoconf or automake. it has arguments that follow autoconf conventions, but it's not autoconf Oct 08 18:49:03 also, it does support passing an exec prefix, it just uses --exec-prefix, not --exec_prefix Oct 08 18:49:58 ok so not inherit autotools just ... Oct 08 18:50:54 probably best to just define a do_configure() { that calls ./configure with the expected arguments, likely copy and paste and modify from autotools.bbclass. it's not ideal, but autotools.bbclass and friends expect certain behaviors based on autoconf/automake, so better to explicitly align with this project Oct 08 18:51:28 kergoth: ok fine thanks Oct 08 18:51:32 it'd also pull in unnecessary dependencies on autoconf/etc if you inherit it, which aren't needed here Oct 08 18:51:35 not a problem Oct 08 18:51:46 I've also dropped installing man completions and others Oct 08 18:51:52 as patch Oct 08 18:52:05 couldn't find optin to disable it explicitely Oct 08 18:52:10 it's not a big deal, and we have other recipes floating around that inherit when they shouldn't and make certain assumptions.. the things do build that way, just not ideal Oct 08 18:52:15 * kergoth nods Oct 08 18:53:16 kergoth: thanks for valuable input Oct 08 18:53:23 I'll fix and test and post Oct 08 18:53:25 recipe Oct 08 18:53:28 on mailing list Oct 08 18:53:38 also other recipe I have worked on is tor Oct 08 18:53:45 I found tor recipe in meta-security Oct 08 18:53:57 but it was pulling browser not tor code Oct 08 18:54:10 tor builds fine without issues though Oct 08 19:06:10 what's the right way to install a /etc/systemd/system.conf in my image? Oct 08 19:08:14 that depends on whether it already exists or it's a new file. if it's a new file, you can eitehr use your own recipe or append systemd. if existing, only the latter, and you can use `recipetool appendfile` to override target files from existing recipes, creates the bbappend for you Oct 08 19:09:17 kergoth: neat, i will try that. thanks. Oct 08 19:09:36 i need to enable systemd's watchdog functionality, since on my hardware, the watchdog is turned on by the bootloader Oct 08 19:48:23 * RP suspects he just created a bit of chaos for layer maintainers :/ Oct 08 19:48:39 (bumping core to zeus as in the layer series name) Oct 08 19:50:20 ugh, i'm so behind on stable branches on mentor's public layers, i don't think i even have a warrior branch yet Oct 08 19:50:28 keep fighting other fires Oct 08 19:55:33 kergoth: the trouble is we're spread too thin :( Oct 08 19:56:14 kanavin: liking your anti-gpl2 patches. how much of the work is done in that branch already? Oct 08 19:56:19 I'm swamped with internal stuff lately, haven't been able to do jack shit upstream, which sucks, i hve a lot i want to do and submit Oct 08 19:56:20 hrmph Oct 08 19:57:51 OT, but it seems the 'bugz' commandline tool from 'pybugz' works with the yocto bugzilla Oct 08 20:01:10 kergoth: I should talk to you that privately but seems freenode won't let me Oct 08 20:01:20 you about that Oct 08 20:04:45 kergoth, the oil industry uses dynamite to put out fires. may want to give that a try ; ) Oct 08 20:05:14 if only.. Oct 08 20:13:21 I just start my own fires as a distraction Oct 08 20:15:54 armpit: I like that style :) Oct 08 23:00:43 does anyone know scotts handle Oct 08 23:01:06 Theres a documentation change thats required Oct 08 23:01:29 I'll file a bug but I still wanted to tell him Oct 08 23:06:22 i don't seem him here. You may be better of emailing him Oct 08 23:06:44 cc richard please Oct 08 23:07:33 aehs29: can you send a patch? :) Oct 08 23:07:54 armpit: will do thanks Oct 08 23:08:57 RP: I can actually, I'm a little afraid of breaking the docs, I dont think I've ever built them, but its just the syntax modifications on out multiconfig builds Oct 08 23:09:15 RP: I'll do that now Oct 08 23:09:47 aehs29: thanks. If we break them, we can fix them... :) Oct 08 23:10:30 look at me. I break things all the time ; ) Oct 08 23:10:44 armpit: we all do haha Oct 08 23:10:51 armpit: starting fires and creating smokescreens? Oct 08 23:11:10 I did like pyro as a release name Oct 08 23:13:47 so did it Oct 08 23:14:38 hahaha Oct 08 23:58:11 RP: will this make it on Zeus? Oct 08 23:58:33 I just saw you just created the branch Oct 09 00:07:46 I know thats a horrible question to ask lol Oct 09 00:07:55 classic Oct 09 00:43:04 RP: sent a patch but I wasnt suscribed to the list, youre gonna have to approve :( Oct 09 02:41:16 hey all **** ENDING LOGGING AT Wed Oct 09 03:01:40 2019