23:00:31 #startmeeting Sponsor 31, October 22nd 2019 23:00:31 Meeting started Tue Oct 22 23:00:31 2019 UTC. The chair is gaba. Information about MeetBot at http://wiki.debian.org/MeetBot. 23:00:31 Useful Commands: #action #agreed #help #info #idea #link #topic. 23:00:35 the pad is here: https://pad.riseup.net/p/IRjuGtQkWNKZLc7qskDb 23:00:42 hi catalyst 23:00:47 hihi! 23:00:47 do we have nickm and teor in the room? 23:00:51 hi! 23:01:48 There is a little over a month to close this sponsor. We will have the next meeting in 2 weeks. I added a few things to the agenda, related to the work that is happening now and in the next two weeks. 23:01:51 teor was online not too long ago... 23:02:13 do people have anything else to add to the agenda? 23:02:32 Sorry, my time it out of sync? 23:02:47 ok 23:03:21 I assume that people will add anything to the agenda if you have anything else to talk 23:03:44 first let's check on the 3 areas that people were coordinating/leading/working on :) 23:04:04 any update? 23:05:03 on documentation, i've been working ot improve the state of our doxygen stuff, and bring doc/HACKING/design up-to-date. I think some stuff will move between them. 23:05:06 *to 23:05:26 catalyst: around? 23:05:37 are we going to start meetbot? 23:05:44 <-- already started it 23:05:45 nickm: i made some children of #29215; do you want to take #32211 or similar? 23:06:02 Sure, assign it to me? 23:06:29 I'm hoping that doc/HACKING/design is a reasonable place to put that 23:06:58 ok 23:07:05 yeah i'm not quite sure about what kind of naming scheme to use for the "target architecture" stuff 23:07:43 I've given up for now on finding the best possible naming/organization for this, and trying to just get it all down somewhere. :) 23:07:52 I figure this stuff is easy to move around once it's written 23:07:55 true, we can rearrange it later 23:08:04 and "poorly organized" is better than "not written down" 23:08:27 up to a point, in my experience 23:08:33 and for text, "poorly organized content" is maybe still better than "a good outline with no content in it" 23:08:36 that is about documentation. So do you both think this will be done in the next 2 weeks? 23:08:50 catalyst: true 23:09:05 what's the end date for sponsor31? 23:09:12 november 30th catalyst 23:09:12 end-of-november. 23:09:35 I think that if we both write stuff this week and next, we can get enough done so that everything else we do on this deliverable is good-to-have. 23:09:46 sounds good 23:10:01 I think we _should_ work on all our remaining deliverables up to end-of-nov; they are structured to be improvable. 23:10:07 catalyst: does that sound right to you too? 23:10:26 nickm: sure 23:10:41 catalyst: what part/parts are you planning to write first? 23:11:18 was going to start with #32206 23:11:29 sounds great. 23:11:52 I am excited about your vision there :) 23:12:14 thanks :) 23:12:15 there is quite a bit of documentation in the 'must-tickets' but I assume their estimation is low 23:12:18 https://trac.torproject.org/projects/tor/query?status=accepted&status=assigned&status=merge_ready&status=needs_information&status=needs_review&status=needs_revision&status=new&status=reopened&keywords=~s31-docs&sponsor=%5ESponsor31-must&col=id&col=summary&col=keywords&col=status&col=owner&col=type&col=priority&order=priority 23:12:23 :) 23:12:55 ok, it seems we are fine with documentation 23:13:05 what about 'Modularization to do in 0.4.3' ? 23:13:47 I'm moving on config refactor again; teor has started on relay-code modularization 23:14:23 I estimate I've got another week or so till other people can modularize configuration objects. 23:14:27 We merged #32123, so we have a minimal --disable-module-relay 23:14:41 Right. It doesn't actually remove any code yet; that's the next steps. 23:15:24 it removes authmode.c 23:16:20 ah, true. 23:16:23 It doesn't remove any options yet, and neither did --disable-module-dirauth 23:16:24 sorry there 23:16:47 We have #32139 to remove all the options, but that depends on the config.c refactor 23:17:01 So I want to open a ticket to remove: 23:17:17 (well, add specific config code to disable) 23:17:28 dirauth: AuthoritativeDirectory 23:17:42 relay: DirPort, DirCache, ORPort 23:18:06 Have I missed any options? Is DirCache strictly necessary? Should I set ClientOnly to 1 ? 23:18:37 I think setting ClientOnly to 1 is reasonable. Is that ticket a subticket of #32139 ? 23:18:39 That is, have I missed any options that let you run a dirauth or relay/dircache 23:18:48 I don't think there are other options; I'll look. 23:19:05 I was going to make it a subticket of #31851 23:19:54 It is now #32213 23:20:00 ok 23:20:07 ah, I understand the relationship between those tickets now 23:20:35 Ideally, we revert #32213 when we have a better way of disabling all the options 23:20:58 will we have time to "modularize configuration objects" in one week? 23:21:15 we don't have to. 23:21:34 Anything we can do on that is extra; the configuration modularization itself is enough to help us down the road. 23:22:00 ok 23:22:08 It would be nice to implement at least one example, so we know it works 23:22:13 agreed. 23:22:38 teor: how do you want to proceed on relay modularization after #32213 ? 23:23:26 We talked about making lists of things to disable/dependencies to cut, and taking them on one by one. 23:23:31 I think we can do #32162 and #32163 independently, we should also try to pick other files to disable 23:24:00 I would be happy if anyone on the team wanted to pick a file and disable it 23:24:26 But perhaps we should make a list in order of "easy to disable, as far as we can tell" ? 23:24:26 It might be convenient in some cases to do this in some way other than one file at a time. 23:24:40 yeah. Do you think you can draft a list like that? 23:24:55 Sure, I think "one lump of code at a time" is a reasonable strategy 23:24:58 Also, we want to not only disable, but to "modularize" -- we want to make the boundaries more clear between this code as we do it. 23:25:41 Part of that process is code movement into directories and files 23:25:56 also, Should there be a new parent ticket for #32163 for "improve modularization of directory authority code"? 23:26:05 It seems like it only sorta belongs under #31851 23:26:15 and like there are other dependencies to cut for the directory authority modularization 23:26:38 I can make those tickets if you like 23:27:03 If you want to split #32163, #32139, and #32213, feel free. But I wonder if it is easier to track these things in the same place. 23:27:37 if you think it is, it doesn't hurt to keep them all under #31851. 23:28:00 the third area is "Discussion on new C style. - nickm coordinating" . How are we doing with this one? 23:28:09 teor: Do you want to try to draft a list of code-chunks that we could disable, or should I do the first draft, or what? 23:28:21 gaba: I've sent out the survey and gotten replies from everybody but asn, who is out. 23:28:22 I think it's fine that way. I want to be able to see all the dependencies. If it gets too big, we can split. 23:28:36 ok 23:28:37 nickm: I can draft the list. But I think you might know the code better? 23:29:02 either way is okay with me; I could do it, but i don't want to grab stuff if you'd rather. 23:29:08 I won't have time to do it till tomorrow at the earliest. 23:29:23 I'll try today, and see how I go. 23:29:23 gaba: when asn is back, I'll hopefully get a reply from him too, and then crunch some numbers and see where we have consensus. 23:29:28 sounds good 23:29:37 gaba: in parallel, I'm playing with a couple of c code formatting tools to see which we should standardize on. 23:29:51 awesome 23:30:09 gaba: right now the leaders are (in no order) clang-format, uncrustify, and astyle. 23:30:28 ok, let's add that to the notes 23:31:57 The last topic that I added is the board. I updated it looking at what I thought you were working on. Can you take a look and see what you are working on is up to date and what the next 2 weeks will be (column 'next') is up to date: https://dip.torproject.org/torproject/core/tor/-/boards?scope=all&utf8=%E2%9C%93&state=opened&label_name[]=Sponsor%2031 23:33:25 there are two instances of #29215 ? 23:33:48 oops, maybe 23:34:07 there's nothign for improving tor-guts afaict? 23:34:45 * gaba closed one 23:34:58 these tasks do not really correspond well to what I need to do to get my work done. 23:35:13 nickm: I wrote up the relay modularisation we discussed at: https://trac.torproject.org/projects/tor/ticket/31851#comment:9 23:35:26 great 23:35:27 Can you check it when you get time? 23:36:10 Yes; I'll try to get to it tomorrow. 23:36:27 I expect to be lightly impeded tomorrow by a doctors appointment, but I should be okay for most of the day 23:36:33 gaba: I put #30901 on the backlog, but it got moved back 23:36:39 nickm: what is missing? 23:36:46 move back? 23:36:48 let me check 23:37:13 gaba: there is nothing for my "tor-guts" rewrite work -- the documentation of tor's current architecture and layout 23:37:30 which ticket nickm? 23:37:38 #29214 23:37:58 Also I don't see anything here for #31851 ? 23:38:07 ah, I see #29214 now. 23:38:10 never mind :/ 23:38:22 teor: are you saying that you are not going to work on #30901 in the next weeks/month? 23:38:35 How many weeks? 23:38:44 next 2 weeks 23:38:51 or in the sponsor (in the next month) 23:39:12 (Wow, that's confusing, let's just use one time period?) 23:39:49 i'm using the column 'next' to mark what we are going to do until next meeting (2 weeks) 23:39:52 and the backlog for the rest 23:40:11 if this is not something to include in the sponsor we remove it or move it to the icebox 23:40:24 Ok, then it belongs in "Next" or "Backlog" depending on how much work I get done on #31482 and #31851 23:40:43 ok. I'm going to move it to the backlog and then we check again next meeting 23:41:00 teor @teor removed Doing label 1 day ago 23:41:00 teor @teor added Next label 1 day ago 23:41:31 Do you want to manage my tickets on the board? Or do you want me to? 23:41:32 Is there anything else to talk in this meeting? Everybody ok? 23:41:55 teor: you can move them and i check with you 23:41:59 * catalyst is good for now 23:42:28 Next meeting is on November 5th at 23UTC. ok? 23:42:35 #31511 doesn't make sense 23:42:39 It's a TODO item 23:42:55 It isn't actually a unit of work, it's a reminder 23:43:34 I'd prefer to close that TBH. There is no apparent work to be done for it. 23:43:39 Same with #31516 23:43:52 yes, let's close what is not work 23:44:02 ok w you teor? 23:44:09 yes, that's fine 23:44:37 * nickm closes 23:45:01 ok. I'm going to end the meeting. Any other discussion can go outside :) 23:45:06 #endmeeting