**** BEGIN LOGGING AT Tue Jan 21 02:59:56 2020
Jan 21 07:47:05 <LetoThe2nd> poll here: https://twitter.com/YoctoThe/status/1219515995288633344
Jan 21 13:24:14 <davidlt> Is there a plan to clean up meta-openembedded from setuptools.bbclass which was removed in core (Date:   Fri Jan 17 14:16:32 2020 +0000) ?
**** BEGIN LOGGING AT Tue Jan 21 13:31:00 2020
Jan 21 14:05:31 <jonmason> LetoThe2nd: I'm working on a script to generate an "top contributors" to YP/OE for a given release
Jan 21 14:05:41 <jonmason> similar to how the kernel does it
Jan 21 14:05:56 <jonmason> https://lwn.net/Articles/679289/
Jan 21 14:06:35 <LetoThe2nd> jonmason: yeah. we don't have that many votes (yet), but it seems to be the preferred perk so far.
Jan 21 14:06:46 <jonmason> For example:
Jan 21 14:06:57 <jonmason> Most Active Developers in meta-openembedded for zeus release
Jan 21 14:06:57 <jonmason>     140  Khem Raj 
Jan 21 14:06:57 <jonmason>     123  Adrian Bunk 
Jan 21 14:06:57 <jonmason>     112  Zang Ruochen 
Jan 21 14:06:57 <jonmason>     103  Oleksandr Kravchuk 
Jan 21 14:06:58 <jonmason>      63  Kai Kang 
Jan 21 14:06:58 <jonmason>      62  Andreas Müller 
Jan 21 14:06:59 <jonmason>      59  Martin Jansa 
Jan 21 14:06:59 <jonmason>      48  Yuan Chao 
Jan 21 14:07:00 <jonmason>      43  Changqing Li 
Jan 21 14:07:00 <jonmason>      40  Hongxu Jia 
Jan 21 14:07:48 <jonmason> I'm aiming to have this done for the OE workshop, then it can be run for each release
Jan 21 14:08:59 <LetoThe2nd> it think it would be good if we had a list of "doing this for $company" names, and filter those out into a list of their own. to encourage small contributors.
Jan 21 14:09:07 <jonmason> I did the above with a hacky shell script.  I'm trying to write something more professional and extansable now
Jan 21 14:09:18 <jonmason> LetoThe2nd: I have this
Jan 21 14:09:27 <jonmason> Most Active Companies in meta-openembedded for zeus release
Jan 21 14:09:27 <jonmason>     333 gmail.com
Jan 21 14:09:27 <jonmason>     270 windriver.com
Jan 21 14:09:27 <jonmason>     170 cn.fujitsu.com
Jan 21 14:09:27 <jonmason>     123 stusta.de
Jan 21 14:09:28 <jonmason>     103 oleksandr-kravchuk.com
Jan 21 14:09:28 <jonmason>      29 linux.intel.com
Jan 21 14:09:29 <jonmason>      26 tworaz.net
Jan 21 14:09:29 <jonmason>      19 intel.com
Jan 21 14:09:30 <jonmason>      13 baylibre.com
Jan 21 14:09:30 <jonmason>      12 wdc.com
Jan 21 14:09:48 <LetoThe2nd> like anything jama, rp, khem, intel, WR - we all know they're doing awesome work, but its not exactly the point we want to convey that they are great.
Jan 21 14:10:25 <jonmason> well, we can't skip over the people doing good work
Jan 21 14:10:42 <RP> ndec: ^^^
Jan 21 14:10:56 <jonmason> I can make it 20 or whatever, which should show the non-superstars
Jan 21 14:11:03 <RP> ndec: any advice/help you can offer jonmason?
Jan 21 14:11:26 <RP> jonmason: maybe some of us should be excluded as "known" contributors?
Jan 21 14:11:43 <LetoThe2nd> jonmason: we should not skip them, of course. but we need at least *one* constructed list onto which newbies have good chance of making it.
Jan 21 14:11:57 <jonmason> RP: I don't think that is fair.  I think making the list longer or maybe even the whole list
Jan 21 14:12:14 <RP> Doing a delta and highlighting new contributors ?
Jan 21 14:12:19 <LetoThe2nd> for example.
Jan 21 14:12:24 <jonmason> that could be interesting
Jan 21 14:12:31 <LetoThe2nd> there's always ways to fine-tune statistics :)
Jan 21 14:12:39 <RP> jonmason: I know ndec has some mapping info
Jan 21 14:12:42 <jonmason> this is per release too, not historic (which could be a separate one)
Jan 21 14:13:27 <LetoThe2nd> jonmason: think of like 3 or 4 lists. total contributions, new contributions... by company name, by area...
Jan 21 14:13:56 <jonmason> I was thinking the same.  One for each meta-layer, then a total for all of OE/YP
Jan 21 14:14:08 <jonmason> person and company
Jan 21 14:14:19 <jonmason> but the idea of a new person and company is interesting
Jan 21 14:14:45 <jonmason> also, mapping it over time to show who is consistent and other trends
Jan 21 14:14:52 <LetoThe2nd> RP: i think you would happily share some of your well-earned praise if you get motivated new folks sending bugfixes in turn :)
Jan 21 14:15:36 <jonmason> Also, if we could find a way to tag bugfixes in the git log, I could look for that too
Jan 21 14:16:06 <LetoThe2nd> jonmason: thats what i meant. be creative with the database.
Jan 21 14:16:07 <RP> LetoThe2nd: right, I don't need the praise ;-)
Jan 21 14:16:07 <jonmason> NM, I see it now
Jan 21 14:16:36 <RP> jonmason: we do tag open bugs
Jan 21 14:16:37 <jonmason> I can search for "Fixes [YOCTO #13533]" or wahtever and show who does the most bug fising
Jan 21 14:16:55 <RP> jonmason: YOCTO #XXXX is the expression to look for
Jan 21 14:18:02 <jonmason> my goal is to have this done in the next couple of days.  I'll send it to the OE mailing list
Jan 21 14:18:17 <jonmason> err...send the output
Jan 21 14:20:43 <davidlt> I just BBMASKed hundreds of python2 packages for now
Jan 21 14:22:10 <ndec> jonmason: YP has a paid contract with bitergia they provide dashboard for git, email, ... contributions
Jan 21 14:22:46 <jonmason> ndec: is this something we can ask them to do then?
Jan 21 14:22:58 <jonmason> or is it just a dashboard
Jan 21 14:23:16 <ndec> we have full control on the data they process.
Jan 21 14:25:37 <ndec> jonmason: in fact, LF is working on a similar tool as well, for all LF projects. e.g. https://lfanalytics.io/projects/723f55c0-b723-4da2-965f-2b2e4a6c4f72/dashboard;v=global%2Fcode%2Fgit%2Foverview-7f26c2fe-3369-43df-ac6f-0eb6007f5edc
Jan 21 14:26:30 <ndec> one thing which is important to highlight for each release would be new contributors as well. 
Jan 21 14:26:49 <ndec> I would be happy to work with you on these things!
Jan 21 14:27:52 <jonmason> ndec: wow, this looks impressive
Jan 21 14:28:31 <ndec> it's based on the tools built from the CHAOSS project, all tools used are open source. it's a great project for community..
Jan 21 14:30:50 <jonmason> ndec: I'll look this over more and contact you offline to see how we can best do this
Jan 21 14:42:12 <LetoThe2nd> ndec: in other news, what are you using to remove the pre-/postroll from videos?
Jan 21 15:03:30 <ndec> LetoThe2nd: https://handbrake.fr/ 
Jan 21 15:03:41 <hrw> jonmason: you need database email->company as LWN has.
Jan 21 15:04:03 <Crofton|road> hrw: you mean linkedin :)
Jan 21 15:04:14 <hrw> jonmason: I remember mail from Corbet one day with "for which company you did your kernel work?"
Jan 21 15:04:14 <jonmason> hrw: there was a discussion about simply letting people self-identify
Jan 21 15:04:29 <jonmason> hrw: I got them from gregkh back in the day
Jan 21 15:05:03 <hrw> I remember times when we could just use Ohloh for it.
Jan 21 15:05:10 <jonmason> this is why a shell script won't cut it.  I need to essentially build a database of people, their commits and their info
Jan 21 15:06:15 <jonmason> I thought of trying to correlate the person and a linkedin profile, thus allowing for accurate corporate representation.  But I think that's a pandoras box
Jan 21 15:06:52 <JPEW> Is there a way to tie it into the groups.io account somehow?
Jan 21 15:07:12 <JPEW> That way, users can manage their own identity
Jan 21 15:08:10 <JPEW> Oh, right, OE isn't on groups.io. Sorry
Jan 21 15:08:42 <Crofton|road> based on the number of subject lines starting with #foo ....
Jan 21 15:10:15 <ndec> jonmason: the tools i mentioned have this database. we can update each record.
Jan 21 15:10:53 <jonmason> JPEW: was that a burn...because it felt like a burn ;-)
Jan 21 15:14:10 <jonmason> ndec: I'm going to look, I promise 
Jan 21 15:14:46 <LetoThe2nd> ndec: ah nice.
Jan 21 16:04:31 <smurray> jonmason: might be worth asking Walt Miner from AGL if he'd share his script(s), he does the same thing for the AGL repos
Jan 21 16:05:51 <smurray> jonmason: he's mentioned that having mapping for different email addresses for the same person and accent character mangling in names is a big chunk of the complexity
Jan 21 16:06:04 <jonmason> I bet
Jan 21 16:12:28 <LetoThe2nd> and one more time: please take your vote! https://twitter.com/YoctoThe/status/1219515995288633344
Jan 21 16:16:12 <kergoth> LetoThe2nd: i'd agree with rburton on that, callout in the weekly
Jan 21 16:16:42 <LetoThe2nd> kergoth: heh, its not mine to decide - i don't write the weeklies.
Jan 21 16:17:18 <LetoThe2nd> kergoth: i just read the OE TSC report mail and thought like, lets try and find out.
Jan 21 16:17:31 * kergoth nods, it's a good idea
Jan 21 16:20:51 <LetoThe2nd> kergoth: i'm totally in "i'll give all that random sh** that nobody bothered doing so far a try."-mode this year.
Jan 21 16:21:44 <kergoth> nice, i like doing that sort of thing
Jan 21 16:22:06 <kergoth> occasionallyi 'll grab something off my todo list that's been there for like 4 years and prototype it in an hour and it works, and i'm wondering whyt he fuck i waited so long
Jan 21 16:22:44 <LetoThe2nd> kergoth: yeah, we are all like that. grabbing stuff and making it work. but i'm specifically focusing on the non-technical, community, advocacy stuff.
Jan 21 16:23:19 <LetoThe2nd> for anything technical in OE/YP, there are about tens of guys far superior to me.
Jan 21 16:23:24 <LetoThe2nd> and gals, ofc.
Jan 21 17:01:08 <JPEW> jonmason, No, not a burn, I just got confused :)
Jan 21 17:03:22 <JPEW> moto-timo, kergoth I think we're going to release Pyrex 1.0.0 at the end of this week unless you can find anything earth shatteringly broken with it
Jan 21 17:09:24 <kergoth> sorry, haven't had much time to play with it lately, busy at work and son keeps getting sick, now i'm sick again
Jan 21 17:15:58 <JPEW> kergoth, Thats too bad... it's been a bad year for that
Jan 21 17:46:18 <JaMa> LetoThe2nd: my commits are my own, LGE doesn't pay me to work on upstream OE
Jan 21 18:49:06 <khem> q
Jan 21 19:33:58 <LetoThe2nd> JaMa: interesting. sorry for the misunderstanding then.
Jan 21 19:40:17 <LetoThe2nd> final reminder for (my) today: https://twitter.com/YoctoThe/status/1219515995288633344
Jan 21 20:22:29 <JaMa> ndec: any idea why https://lfanalytics.io doesn't show me anywhere? Is the list of people it parses restricted somewhere? At least when checking meta-oe repository I think I should be in top 5
Jan 21 20:23:28 <JaMa> ndec: it shows khem as top1 with 1,476 commits, but simple grep shows:
Jan 21 20:23:39 <JaMa> martin@jama:/OE/meta-openembedded$ git log  | grep -c -i ^Author:.*Martin.*Jansa
Jan 21 20:23:39 <JaMa> 1961
Jan 21 20:23:40 <JaMa> martin@jama:/OE/meta-openembedded$ git log  | grep -c -i ^Author:.*Khem.*Raj
Jan 21 20:23:40 <JaMa> 1571
Jan 21 20:24:12 <ndec> JaMa: i will check when i am at my desk tomorrow. You should definitely be there! We have some problems with the LF tool, it is in beta (or alpha..). I am using another dashboard which is usually more accurate.
Jan 21 20:26:50 <JaMa> ok, thanks, maybe the tool is confused that I've used 4 different formats of my name in Author and somehow it counts only one of them or something
Jan 21 20:31:11 <paulbarker> ndec: Is there an email address for the advocacy team or someone on the team I can contact? I don't want to join the mailing list for a one-off
Jan 21 20:31:24 <khem> so all the webtool needs is to run `git shortlog -n -s`
Jan 21 20:31:28 <khem> :)
Jan 21 20:31:52 <khem> but it could be harder than that I dont know 
Jan 21 20:32:17 <Crofton|road> @paulbarker what is the question for advocacy?
Jan 21 20:32:45 <ndec> paulbarker: send the email to the list. I will add you to approve senders
Jan 21 20:32:58 <paulbarker> On the Yocto call today we were discussing a few things but I don't think any of the advocacy folks were on the call
Jan 21 20:33:08 <khem> paulbarker: talk to Andrea or send to ml yocto-advocacy at yoctoproject.org
Jan 21 20:33:26 <paulbarker> I just wanted to point at the developer profiles that other LF projects are doing as a good idea for Yocto
Jan 21 20:33:30 <Crofton|road> non subs will be held for moderatin and apporved if reasonable
Jan 21 20:33:40 <paulbarker> e.g. https://www.automotivelinux.org/blog/2019/11/26/developer-showcase-barker
Jan 21 20:34:35 <LetoThe2nd> Crofton|road: how boring. "reasonable"
Jan 21 20:35:02 <Crofton|road> ok, not spam :)
Jan 21 20:55:06 * zeddii reads
Jan 21 20:55:13 <zeddii> yah. that dashboard is whacked.
Jan 21 21:11:42 <ndec> khem: i would tend to agree... in fact our cgit implementation on git.yp.org and git.oe.org has a 'stats' tab that just does that.. e.g. https://git.openembedded.org/meta-openembedded/stats/?period=y&ofs=10
Jan 21 21:12:23 <ndec> it only shows the last 4 'periods' which can be week, month or year.. though.
Jan 21 21:13:14 <ndec> JaMa: on the other dashboard (which I am hoping to have public soon), you are showing as #1 in the meta-oe contributors.. not sure what's wrong with the LF dashboard yet, but i confirm it's bugged.
Jan 21 21:17:52 <Crofton|road> so sip in meta-oe has py2 and py3 support. I need to updat ethe recipe, because pyqt in meta-qt5 needs more reent sip (not sure how the update was tested in meta-qt5 :)
Jan 21 21:18:26 <Crofton|road> do I merge the inc into teh sip3 recipe and drop py2? WSHould py2 go somehwere, or do I let someone else solve that problem?
Jan 21 23:21:24 <Crofton|road> grrrr, I sense the answer to my wquestion is different for zeus than master already
Jan 21 23:22:40 <armpit> Crofton|road, yes merge the inc file and drop py2.. it would be nice if the py2 patch could be sent to meta-py2
Jan 21 23:23:02 <Crofton|road> but seau has py2
Jan 21 23:23:15 <armpit> correct
Jan 21 23:23:23 <Crofton|road> zeus
Jan 21 23:23:35 <Crofton|road> for zeus I wil preserve current status
Jan 21 23:23:48 <Crofton|road> sip lives in meta-oe to serve qt4 and qt5 layers
Jan 21 23:24:34 <Crofton|road> still not sure how the qt5 update for zeus happened, since pyqt5 is unbuildable with sip in meta-oe
Jan 21 23:25:28 <Crofton|road> so I will ned too beg for a version upgrade in zeus :)
Jan 21 23:26:11 <armpit> why is that a meta-oe issue sounds like qt5 messed up
Jan 21 23:26:38 <Crofton|road> there is a dependency and I do not want to downgrade pyqt5 in meta-qt5
Jan 21 23:27:05 <armpit> so demanding ; )
Jan 21 23:27:43 <armpit> you can always send a patch and see what happens
Jan 21 23:28:21 <Crofton|road> yeah
Jan 21 23:28:35 <Crofton|road> need to talk to khem, he did update
Jan 21 23:28:48 <Crofton|road> py2 verion of pyqt5 doesn't build also
Jan 21 23:28:51 <armpit> ah, so khem broke it : )
Jan 21 23:28:56 <Crofton|road> right
Jan 21 23:29:10 <Crofton|road> , jsut not sure how it passed any build test
Jan 21 23:29:14 <Crofton|road> tomorrow problem
Jan 21 23:29:55 <armpit> k
Jan 21 23:34:06 <Crofton|road> py2 version of pyqt5 is alrady blacklisted
Jan 21 23:34:13 <Crofton|road> have to see if that is true in zeus
Jan 21 23:34:18 <Crofton|road> anyway I am off
**** ENDING LOGGING AT Wed Jan 22 03:01:07 2020