**** BEGIN LOGGING AT Wed Nov 11 02:59:58 2015 Nov 11 08:15:04 good morning Nov 11 09:09:41 morning all Nov 11 09:19:39 hi bluelightning Nov 11 09:35:20 hi mckoan Nov 11 10:50:04 hello Nov 11 13:03:42 koen: sort.sh/angstrom-feed-configs Q. some packages are placed in armv7ahf-vfp-neon in my feedconfig, and that is a place my Jetson does not look. I need help understanding what's wrong :) Nov 11 13:04:36 should packages not be placed there for me, or should the feed config look in that dir as well as armv7at2hf-vfp-neon? Nov 11 13:13:12 I want to change my Q to: why don't I have a libc6-dev.ipk for my arch :) Nov 11 13:14:58 iirc newer sort.sh version alias those together Nov 11 13:22:38 can you make bitbake use git am to apply patches instead of just patch Nov 11 13:22:58 heeen: you can indeed, just set PATCHTOOL = "git" in the recipe Nov 11 13:23:16 huh Nov 11 13:25:47 koen: ah. I have the one from 1.8 I think Nov 11 13:31:26 hm fails with index out of range Nov 11 13:37:41 koen: that is actually the same one as in master. Nov 11 13:48:55 heeen: ? Nov 11 13:50:26 nvm i patched it manually Nov 11 13:51:35 if you can help me to reproduce the issue I might be able to fix the bug if there is one... Nov 11 15:22:05 when taking one recipe from another repo/layer, what's the correct procedure to indicate that in a commit, setting --author to the original author and my sob? Nov 11 15:28:08 joseppc: our usual practice is to just say which layer you imported it from Nov 11 15:28:37 if you're setting the author that may not really be right; besides it doesn't cover any changes by people since the recipe was added Nov 11 15:31:50 bluelightning: ok, I wanted to somehow honor the actual author of the recipe, but indicating repo and commit should be quite close to that I guess Nov 11 15:43:15 unless the original author produced the exact diff against that repo, setting --author is disingenuous, even if it is well intentioned Nov 11 15:44:46 could always mention the author in the commit if you want to Nov 11 16:02:43 kergoth: thanks Nov 11 16:04:56 it's debatable. in the US, evne lacking a copyright statement, those lines are under the othe rperson's copyright, and if you don't mention them at all, that information won't exist anywhere if someone wanted to re-license. of course, in this case it's MIT and trivial and i doubt anyone cares, but in general... Nov 11 16:10:37 right... my comment was based on what we've done in OE up to now; e.g when bringing the stuff we've brought in from OE-Classic, we just stated that in the commit message Nov 11 18:18:22 Any idea why this doesn't show up in run.do_deploy ? Nov 11 18:18:23 do_deploy_append_ettus-300() { Nov 11 18:19:18 - is invalid in shell function names Nov 11 18:19:22 wont' be possible to emit it Nov 11 18:19:35 of co urse, if the append applies, it won't matter Nov 11 18:19:45 did you confirm OVERRIDES is what you expect? Nov 11 18:22:50 OVERRIDES looks ok Nov 11 18:23:15 but the lines from that append aren't in the do_deploy function? Nov 11 18:23:23 right Nov 11 18:23:29 dropping the machine bit Nov 11 18:24:11 since the line can work work for all cases Nov 11 18:24:36 do_deploy_append() { Nov 11 18:24:41 does what I expect Nov 11 18:25:33 weird Nov 11 18:25:43 so the - in AMCHINE can be bad? Nov 11 18:32:49 i think the - is likely replaced by _ in the override Nov 11 18:40:45 hmmm Nov 11 18:41:24 but only for shell? SRC_URI is fine? Nov 11 18:42:00 I think I used to know this Nov 11 18:45:41 hmm, maybe it was the other way around, _ replaced with - :) (remembering x86_64) Nov 11 18:45:44 * kergoth shrugs Nov 11 18:56:00 I the the - is replaced with _ in shell overrides Nov 11 18:58:35 interesting do not see it in manual Nov 11 19:09:35 the thing is, the override should never be seen in the shell, it's not a directly used function, it's just used as the conditional, so that aspect shouldn't be a factor here Nov 11 19:09:46 i misread it at first thinking itw as a custom function ratehr than an append Nov 11 19:25:21 tis odd Nov 11 19:33:20 i don't think overrides cares what's in the name, just an _ causes issues since that's the override separator Nov 11 19:48:40 yeah **** ENDING LOGGING AT Thu Nov 12 02:59:58 2015