Time Nick Message 09:49 Krock I'd like to do a meeting this evening again. Who's got time to attend? 10:07 sfan5 me most likely 15:16 Krock ill merge #10239 in 10 minutes 15:16 ShadowBot https://github.com/minetest/minetest/issues/10239 -- Fix GCC class-memaccess warnings by pauloue 15:25 Krock merging 15:26 rubenwardy Me 15:26 Krock cool :) 15:26 Krock with quite some luck, either ANAND, nerzhul or paramat will pop up 16:43 Krock meeting around 18:00 UTC? (1h 15min) Would that be okay? 16:52 Krock added new points: 2 MTG, 2 organisation 16:55 * sfan5 looks 16:57 Krock most of the PRs are just about reviewing, organisation points might take more time 18:32 Krock rubenwardy, sfan5 - ready? 18:33 rubenwardy yeah 18:33 sfan5 sure 18:33 Krock alright :) first up "Theme of 5.4.0" 18:34 Krock any particular focus? 18:34 Krock there are a few PR in the milestone, so I guess that's it? 18:34 sfan5 those are probably just there because they were "ready" during feature freeze 18:35 Krock though the ContentDB PRs look quite promising 18:35 Krock might review them later (again?) 18:35 Krock > New core developers 18:35 Krock could this be a leftover from previous meetings? 18:36 rubenwardy no, we still need core developers 18:36 rubenwardy ANAND left 18:36 Krock though I think the problem still persists - ANAND became a core dev, but is not active 18:36 Krock oh? 18:36 Krock that was quick 18:36 rubenwardy yeah, he left Minetest permanently 18:36 Krock sad 18:36 Krock so the search continues 18:38 GreenXenith v-rob and Desour (DS-Minetest) are very active with engine contribution 18:38 GreenXenith Sort of specialized to certain areas, but still active 18:38 Krock also still thinking of Df458 who provides solid work 18:38 GreenXenith He cant legally contribute 18:38 GreenXenith AFAIK 18:38 rubenwardy he can 18:39 Krock GreenXenith: He, who? 18:39 rubenwardy I discussed this with df. He is limited in time, though 18:39 Krock ah 18:39 GreenXenith Oh, I suppose the allowance extends past PRs 18:39 GreenXenith That's good then 18:40 Krock yes, I've also read something about that. As long it doesn't get chaotic it would be fine for me - althogh it might not make much sense when time is a concern 18:40 Krock PR reviewing is tiring and requires quite some time, thus it might not bring too much of a benefit for him 18:41 Krock Desour is also in the loop for quite a while now, though not so active recently 18:43 Krock I'd love to bring pyrollo into the team, but I don't know what was the situation there 18:45 sfan5 hm 18:45 Krock rubenwardy: any ideas? not sure whether I already asked that... 18:45 rubenwardy I don't think his involvement with kidscode is a problem for us, in terms of conflict of interest 18:46 rubenwardy he would be a good addition 18:46 Krock I'll write him an email - let's see what happens :) 18:46 rubenwardy it becomes a problem when you have more than one person employed by the same company 18:47 Krock yes, but we're aware of that to not let it escalate 18:47 rubenwardy yeah 18:48 Krock > minetest-mods.github.io needs a library update 18:49 Krock I got an alert about one of the dependencies, thus the gemfile needs updating 18:49 GreenXenith You know what else needs an update? Doxy >.> 18:49 sfan5 doesn't dependabot make a PR or something like that? 18:49 Krock now it can't, because some dependencies seem to be incompatible with the new version 18:51 Krock I know this is neither Minetest engine nor game, but this channel first best given the current activity within this group 18:52 Krock sofar: FYI ^ 18:53 Krock weird. to me it looks like github-pages itself depends on both of the conflict-causing dependencies, hence I doubt we can do anything? 18:54 sfan5 honestly security vulnerabilities in a static site generator are pretty much irrelevant 18:54 Krock alright. postpone then 18:55 Krock > Minimal/Devtest policy 18:56 Krock this now became quite a big discussion, hence I think it would be best to discuss it here quickly and decide on the next steps 18:56 Krock At least proposals 1 + 2 seem to be accepted in #e quickly and decide on the next steps 18:57 Krock * #9645 18:57 ShadowBot https://github.com/minetest/minetest/issues/9645 -- [Policy] Commitment on new Minimal development test 18:58 sfan5 then, just write down points 1 and 2 somewhere to make it a "policy"? 18:58 Krock the README could be linked to the code style guideline wiki 18:59 Krock the only other policies I could find are here: https://dev.minetest.net/How_to_communicate#Contributing_to_Minetest 19:00 Krock closely related: the "Direction of Minetest" proposal which currently stalls 19:01 Krock it also includes some "how to contribute" aspects which could be put into the wiki (or wherever) 19:02 Krock because you worked on it, rubenwardy: any preference? where should those "General Contribution Guidelines" go? 19:02 rubenwardy contributions.md? 19:04 Krock alright, so next up is a PR for rules 1 + 2, and potential for being expanded later 19:04 Krock I'm not a fan of the "hidden" README file, but that can be linked as well 19:05 pgimeno as for #9645 point 4, addition of a test can not possibly cause breakage to the engine, even if the test is not trivial, so I'd say it's a policy to consider 19:05 ShadowBot https://github.com/minetest/minetest/issues/9645 -- [Policy] Commitment on new Minimal development test 19:06 Krock minimal-only PRs are rare, and for those case I still think the 2-approval-policy should be kept 19:07 Krock my reason for that is consistency, and also ensuring the quality (even if the requirements are lower) 19:08 pgimeno so, if I get this right, each PR should include the corresponding test in devtest when possible, from now on? 19:08 Krock if that makes sense and is needed - yes. 19:08 Krock depends on the case 19:08 pgimeno uh, tests are never needed 19:09 Krock missing sarcasm (/s) ? 19:09 Krock > With an up-to-date Minimal it should also be faster to reveal bugs and—even more important—regressions. 19:09 Krock main focus on avoiding regressions 19:10 GreenXenith Good code style and PR reviewing isnt needed either, but it is so helpful it would be stupid to not do it 19:10 pgimeno no, they are not a strict necessity to have something that works, it can work without the need of having unit tests. Minetest has lacked unit tests for a long time. 19:10 Krock pgimeno: indeed, depends on the case. no need to do that for copy&pasted code, for example 19:10 pgimeno Basically, people tested by hand, not by writing a test 19:11 GreenXenith People arent perfect 19:11 GreenXenith Granted, neither is the code they write. But it is certainly faster than they are and more likely to find something they wont. 19:12 pgimeno my point here is that "needed" is too subjective. I'd say if the code change is suitable for a test, then adding one should be mandatory. 19:12 GreenXenith Im .. pretty sure that was the implication 19:14 Krock pgimeno: yes, "suitable for a test" is likely a better expression 19:14 GreenXenith Minimal guidelines for when a change is test-suitable might be worth thinking about 19:15 Krock test code is already written in most PRs, thus it might not make much of a difference to just put it into minimal once done 19:16 pgimeno by suitable I mean, it's possible to write an automated test for it 19:16 Krock pgimeno: how about testing new HUD elements? a demo mod would be helpful 19:17 Krock visuals cannot be automated, but tests can be added anyway 19:17 pgimeno right 19:17 Krock well "manual tests" 19:18 pgimeno I guess those manual tests should include a description of what to look at when invoked 19:19 pgimeno but note that manual tests that need to be run by the developer will either not be run often (if at all) or place a big burden on the tester to check every test 19:21 Krock the existing tests within minimal require manual steps to start a specific one 19:21 Krock I think that's better than running everything at once (unittest-style) 19:28 GreenXenith More discussion on the details probably belongs in the policy PR 19:29 GreenXenith There is still much to discuss, according to the meeting outline, and I am particularly interested in the next item 19:29 Krock > LuaEntitySAO API #9717 19:29 ShadowBot https://github.com/minetest/minetest/issues/9717 -- Extended motion mechanics API for LuaEntitySAOs by sorcerykid 19:29 Krock I'd like to pick up the concepts of this PR 19:29 GreenXenith Er, an item in the next section* (no particular order, apparently) 19:30 sfan5 do you think it can't be merged as-is? 19:30 Krock GreenXenith: the other have lower priority 19:30 GreenXenith Im not complaining 19:30 sfan5 I have only glanced at the replies to my reiew but they don't seem unfixable 19:31 Krock API functions must be unique and generic 19:31 Krock hence I'm not sure about a few functions which seem to re-implement a similar API again 19:32 Krock mainly set_speed_lateral or set_speed (done using vector rotation functions) 19:33 Krock also on_step arguments should be kept slim IMO - no information contained that could be retrieved using API functions 19:34 Krock sharing my opinions here to see what your impressions are - assuming overall good impressions from sfan5 19:37 Krock well, currently the PR is unmergable due to the on_step changes 19:38 Krock moveresult must be param #3 19:38 ShadowBot https://github.com/minetest/minetest/issues/3 -- Furnace segfault 19:39 Krock okay, the left over points in the meeting are PRs that look good and only need one more approval, or such that lie around for quite a while without any attention 19:40 Krock please feel free to pick one to review - so that the PR and issue count may begin to decrease again 19:42 GreenXenith Does #9079 just need a secondary review or does it need changes to be merged? 19:42 ShadowBot https://github.com/minetest/minetest/issues/9079 -- Minimap in HUD, minimap showing texure, control minimap modes from Lua by EvidenceBKidscode 19:43 Krock secondary review 19:43 GreenXenith Rubenwardy already did a bunch of reviewing, most of which is resolved *nudge nudge* 19:46 GreenXenith Once that PR is merged and the server has control over the minimap, ive considered doing a PR to add minimap icon definitions to entities, because I want full control over how players and objects are displayed on the map 19:51 pgimeno mainly set_speed_lateral or set_speed (done using vector rotation functions) <==== not without #8515 19:51 ShadowBot https://github.com/minetest/minetest/issues/8515 -- Add matrix helpers to builtin by Desour 19:53 Krock GreenXenith: that's a feature request for a PR that builds onto that one 19:54 pgimeno would unit tests for 8515 help in getting it merged? 19:54 Krock pgimeno: not with this? #9572 19:54 ShadowBot https://github.com/minetest/minetest/issues/9572 -- Some vector functions useful for working with rotations by NetherEran 19:55 Krock yes, it would clearly help, though I wonder whether a mod would be better suitable for such functions (if they're too specific for most mods) 19:55 pgimeno uh right... now there's a redundant function there 20:04 GreenXenith Krock: Sorry, what? 20:04 Krock (invitation sent) 20:04 Krock GreenXenith: nvm. yes, that PR would be a good idea to allow further customization 20:05 GreenXenith Alrighty :] 23:54 paramat hi, my comments on the meeting. i could not attend, the meeting time is usually a bad time for me