**** BEGIN LOGGING AT Mon Jun 04 03:00:00 2018 Jun 04 15:36:08 is the best to insert image-specifc configs into the rootfs, using ROOTFS_POSTPROCESS_COMMAND? I don't necessarily want to mod (bbappend) the configs in their respective packages (if any) because it's image-specific, not necessarily distro or feature. just wondering if there's a better alternative out there Jun 04 17:28:04 armpit, do you care about meta-oe/rocko? Jun 04 17:28:35 yes Jun 04 17:28:49 apparently protobuf dropped the branch we use Jun 04 17:29:01 master is 3.5.2 and rocko is 3.4.x Jun 04 17:29:05 so fetch fails Jun 04 17:29:19 also sounds like just removing branch works Jun 04 17:29:28 can I send a commit removing branch Jun 04 17:29:35 not needed elsewhere Jun 04 17:29:41 sure Jun 04 17:29:43 sinc 3.5.x already in Jun 04 17:30:30 I think I must have a tarball locally since I do not see it :) Jun 04 18:34:30 armpit, sent Jun 04 18:34:51 if I was going add a lapack recipe to meta-oe, what subdir should it go in? Jun 04 18:48:50 what is layer path? devtool finish --force-patch-refresh Jun 04 18:51:57 The github exodus is going to make SRC_URIs really fun for a while I see, sigh Jun 04 18:52:14 https://pleroma.site/objects/34237850-73d3-4c95-84cc-384d5efdce3d (pkgconf moving) Jun 04 18:56:41 jesus Jun 04 18:56:54 ugh, stupid. Jun 04 18:56:58 these people are idiots Jun 04 18:57:01 yep Jun 04 18:57:09 and gitlab infrastructure is pretty rugh Jun 04 18:57:11 oh ms is evil, just because they're ms. bite me Jun 04 18:58:07 id fortran broken? Jun 04 18:58:37 layer path is the path to the layer where you want the changes to live.. Jun 04 18:59:01 ker relative to to where you are, or top Jun 04 19:00:37 They're far from the only people running away from github Jun 04 19:00:43 So there will be SRC_URI fun. Jun 04 19:12:48 Crofton|work: i don't understand the question Jun 04 19:12:52 it wants to know what layer to put the files in Jun 04 19:12:55 give it hte path to the layer Jun 04 19:13:00 not the path from the layer to the reciope Jun 04 19:14:28 hmm Jun 04 19:28:40 its not putting that path in the files, it needs to know where ti write the patches and recipe or bbappend, iirc, just like the update-recipesubcommand does Jun 04 19:32:38 hmm, fortran is broken ... | x86_64-oe-linux-gfortran: error: libgfortran.spec: No such file or directory Jun 04 20:07:08 and can't find that file in the tmp dir anywhere Jun 04 21:06:46 Hello, Is there anyone who is admin of http://layers.openembedded.org/, I am trying to reset my password but not getting reset email. Jun 04 21:18:28 bluelightning, ^^^ Jun 04 21:44:08 Crofton|work: that sort of thing would generally be halstead Jun 04 21:44:54 I can see what people's email addresses are in the system though so I can fix if that's the issue (although it would have had to be correct to set up the account in the first place) Jun 04 21:45:03 I was wondering about that Jun 04 21:45:21 I'll look at the logs once I find the e-mail address. Jun 04 21:52:52 bluelightning, The only recent e-mail I see in the logs is from Anuj and it looks like it was accepted. Did you find mjaggi_pc's e-mail? Jun 04 21:53:09 He left Jun 04 21:53:41 * halstead nods. Jun 04 21:54:17 no matches for "jaggi" or "manish" (based on mjaggi_pc's /whowas) Jun 04 21:54:46 maybe there's no password reset email because the account doesn't exist ? Jun 04 21:56:46 Yeah, that's what it looks like to me too. Does the celery server log each action with details? I see things like Jun 04 21:56:53 Task layerindex.tasks.send_email[f9addd63-2991-40eb-914f-6bd5b74ed512] succeeded in 0.7063022274523973s: None Jun 04 21:57:10 I think that's all you get currently Jun 04 21:57:38 Okay. I think I'm stuck until mjaggi_pc comes back. Jun 04 23:13:41 bluelightning: thanks for your input on my devtool patch, i was hoping you'd notice and chime in :-) Jun 04 23:18:09 bluelightning: i have further updated the patch such that the on-device filenames include the optional package name too, so a user can install a bunch of single packages from a recipe and they'll all co-exist happily (without clobbering each other during deploy) Jun 04 23:18:37 I've also added the --package option to undeploy too, to support this new filename scheme Jun 04 23:20:12 bluelightning: i'm confused by your comment, however, about 'not being able to use "devtool build ..." since it only runs up to do_populate_sysroot/do_packagedata' Jun 04 23:21:08 that's okay, isn't it? if i want to push something in ${WORKDIR}/packages-split/ to the target, doesn't do_packagedata update the items found there? Jun 04 23:31:18 tlwoerner: hi, no problem Jun 04 23:31:31 bluelightning: i'm replying the to the email thread now Jun 04 23:31:41 tlwoerner: so, no - do_package is what does the package splitting and that's actually after do_packagedata Jun 04 23:32:14 tlwoerner: actually - wait, that's completely wrong Jun 04 23:32:25 do_packagedata does depend on do_package Jun 04 23:32:41 so we're probably OK, false alarm - I'll reply to correct that Jun 04 23:32:55 :-D **** BEGIN LOGGING AT Tue Jun 05 02:04:39 2018 **** ENDING LOGGING AT Tue Jun 05 03:00:14 2018