**** BEGIN LOGGING AT Tue Oct 08 02:59:57 2019 Oct 08 17:45:11 Crofton|work: had a go at scipy lately? Oct 08 18:03:31 moto-timo, don't push his buttons Oct 08 18:22:50 no I haven't Oct 08 18:23:21 I need to make meta-fortran-stuff and put lapack there and colelct historical scipy things Oct 08 18:23:57 there may be some interest in 'fortran stuff' in ADAS users.. just an FYI Oct 08 18:24:24 (ADAS - Automotive Driver Assistance .... basically autonomous driving uses some fortran stuff for image recognition) Oct 08 18:24:32 that is nuts Oct 08 18:25:00 moto-timo, you agree the thing to do is create a meta-fortran as a home for scipy and its freinds? Oct 08 18:25:19 I know I made an image that I could native build scipy and have it pass many tests Oct 08 18:26:00 I was lead to believe the fortran stuff is heavy math computations (mostly matrix math from the sounds of it) Oct 08 18:28:27 then it should be meta-stuff-one-wants-to-forget Oct 08 18:30:36 I am being polite Oct 08 18:31:51 its with my other layers meta-stuff-oe-would-not-take Oct 08 18:32:33 even oe has standards Oct 08 18:32:46 and meta-armpit-did-not-understand Oct 08 18:33:06 all in favour of me creating meta-fortran on github and pushing my lapack recipe say eye Oct 08 18:33:15 +1 Oct 08 18:33:51 sorry, 'eye' Oct 08 18:36:50 Whats the deal with patchwork.openembedded.org? I was checking on the status of a patch I sent: https://patchwork.openembedded.org/patch/165312/ when I noticed a ton of stuff is still marked New. Is stuff just not getting marked properly or ...? Oct 08 18:38:23 armpit: you mean meta-fun-stuff-no-one-wants-to-forget :) Oct 08 18:39:08 hehe Oct 08 18:39:36 georgem, which repo ? Oct 08 18:39:42 armpit: oe-core Oct 08 18:39:59 ask the maintainer of oe-core Oct 08 18:42:25 RP: Any idea why so many patches are showing up as New here (and on page 2)? https://patchwork.openembedded.org/project/oe-core/patches/?page=1 Oct 08 18:42:39 Looks like the state might not be getting changed when things are merged. Oct 08 18:49:34 When I send a patch and it doesn't go in after a couple of weeks like https://patchwork.openembedded.org/patch/165312/ it's nice to be able to see if other patches sent around the same time have made it in or not so I can chalk it up to a busy queue vs getting lost in the ether or some miscommunication on the ML. Oct 08 19:20:04 georgem: I'm afraid we're struggling finding people to help maintain these things. I don't know why its not working :( Oct 08 19:21:03 RP: understood, just asked in case someone didn't notice or something. thanks Oct 08 19:22:15 georgem: to be honest it upsets me as it would be nice if everything worked and I realise it impacts people. Sadly I don't have enough hours in the day to fix everything :( Oct 08 19:24:36 I totally understand. Oct 08 19:27:48 RP: Sandy can take a look now that she's back from Golden Week holiday. We're also working on a backlog of defects but I think she can spare an hour or two. Oct 08 19:28:06 * vmeson copies IRC and sends it to Sandy Oct 08 19:28:33 vmeson: help is appreciated as ever. Given your other comments I wouldn't like to see you overextended... Oct 08 19:29:03 RP: I'll CC her manager. :) Oct 08 19:33:28 vmeson: :) Oct 08 19:33:29 I heard there someone had some scrips to update patchwork when things got merged. Oct 08 19:33:41 you may want to ping Joe MacDonald Oct 08 19:36:06 armpit: I think there is a post commit hook which is supposed to trigger something like that Oct 08 19:36:24 ah Oct 08 19:37:25 I noticed patches being marked 'superseded' for cve patches that are sent for each branch Oct 08 19:37:44 I had to go back through and manually change the status Oct 08 19:50:19 khem, ping is the tcpdump update going to land in master before zeus is branched ? Oct 08 19:53:33 yes Oct 08 19:54:39 cool. thanks Oct 08 19:55:42 this lot on master-next will land today into master before zeus is branched out probably tonight Oct 08 20:01:19 good as the tcpdump update has a lot of critical cve fixes needing to be backport to stable branches Oct 08 22:14:38 Crofton|work: there are those that feel language based layers are odd (should be functional), but most of the reason for using fortran these days is scientific software. Oct 08 22:15:54 Crofton|work: so it seems to make sense... I also thought about meta-scientific or something stupid like that... then it could house python recipes related to numerical processing Oct 08 22:17:48 Crofton|work: but in short (too late): eye Oct 08 22:18:59 tgamblin: meta-stuff-you-learned-in-school-that-would-be-useful-if-you-remembered-it-now Oct 08 22:19:33 or meta-stuff-i-was-an-expert-at-20-years-ago Oct 08 22:22:07 the common thread here is you need to enable fortran Oct 08 22:27:25 moto-timo: what did you do in scientific computing 20 years ago? Oct 08 22:35:29 the heat characteristic of tube technology? Oct 08 22:46:59 vmeson: theoretical crystal structures of 3 and 4 element compounds Oct 08 23:09:35 Crofton|work: indeed that is the common thread :) Oct 08 23:13:07 lumping by fortran makes sense in that you need to set variable to anble the tool chain Oct 08 23:13:23 I'd rather the user only do that if the one layer is invovled Oct 08 23:13:38 clearly I am hoping this is a small set of SW Oct 08 23:56:58 fortran still being used makes me feel relevant since this was my second language I learned as a kid Oct 09 01:44:19 khem, you make me laugh Oct 09 01:44:35 My third I think after BASIC and 6800 asm **** ENDING LOGGING AT Wed Oct 09 03:01:40 2019