Time  Nick          Message
06:38 nerzhul       oh there is a new thing in the github ci
11:18 rubenwardy    #7253 #7302
11:18 ShadowBot     https://github.com/minetest/minetest/issues/7253 -- Add updating to online content repository by rubenwardy
11:18 ShadowBot     https://github.com/minetest/minetest/issues/7302 -- Add screenshots to online repository browser by rubenwardy
11:31 nerzhul       rubenwardy, nice to see this project evolving :)
11:36 nerzhul       when i get time i think i really should backport my lua c++ convertors model to MT, our raw and non safe low level lua C++ conversion model is just... anoying :D
11:42 rubenwardy    is using u32 instead of int an actual code style convention?
11:42 rubenwardy    u32 in places other than networking or serialisation seems weird
11:42 nerzhul       we use s32 and u32 everywhere, having raw int is not conventional in our code
11:43 nerzhul       the only misconvention is the f32/f64 & float/double which remains special, but our integer convetion is quite uniform
11:44 rubenwardy    should be in the code style guide lines then
11:44 rubenwardy    if it's not already
11:44 rubenwardy    !devwiki Code_style_guidelines
11:44 rubenwardy    no MinetestBot :(
11:44 nerzhul       if i remember someone, maybe c55 or kwoelr told me 3 years ago to follow this model
11:45 rubenwardy    yeah, it's not mentioned
11:45 rubenwardy    if it's not mentioned it's not a rule ;)
11:45 nerzhul       do like you want, but i tend to prefer u32/s32 as they are used everywhere in the code
11:45 nerzhul       i know running on a ZX spectrum is not our goal :p
12:15 rubenwardy    Done
14:35 Fixer         or is it?
14:37 Fixer         imagine running minetest on zx in a pervert way
14:38 Fixer         damn, that windows stealing focus in a weird way is very annoying
14:39 Fixer         wrong channel, damn it
14:46 longerstaff13 lol
17:21 Krock         hello nerzhul. I see your point in not having to use LIMIT when fetching a primary key. However, "read" uses "WHERE name = ?", which isn't primary. https://github.com/minetest/minetest/pull/7279/files#r186629734
17:23 Krock         so do you agree that LIMIT 1 might be used for that single statement?
17:23 nerzhul       no
17:23 nerzhul       i asked for a UNIQUE index
17:23 nerzhul       this guarantee the username unicity
17:24 nerzhul       a UNIQUE index on name column is sufficient
17:24 nerzhul       if you use limit 1 where a field is unique you should just stop to use mongodb or any nosql-moreshit
17:26 Krock         oh I see the suggestion now. Sure, UNIQUE solves that too
17:26 nerzhul       it's unicity + index then the query is just correct in terms of performance
17:26 nerzhul       without that you do a full scan over the whole table on this query, LIMIT 1 doesn't solve anything on that it's a filtering after a full scan on the table
17:28 Krock         thanks, I'll have to correct my thought that LIMIT increases the execution speed if the entire table is scanned anyway
17:29 sofar         <nerzhul> if you use limit 1 where a field is unique you should just stop to use mongodb or any nosql-moreshit
17:29 sofar         lol
17:35 nerzhul       Krock limit is (on all SGDB) the last operation done when the result set was calculated
17:41 rubenwardy    updated #7253 and #7302
17:41 ShadowBot     https://github.com/minetest/minetest/issues/7253 -- Add updating to online content repository by rubenwardy
17:41 ShadowBot     https://github.com/minetest/minetest/issues/7302 -- Add screenshots to online repository browser by rubenwardy
17:46 Krock         I find it somehow funny that you'll have to resolve merge conflicts caused by your one of your own pulls :P
17:47 rubenwardy    nah, that's easy
17:49 rubenwardy    trivial PR #7307
17:49 ShadowBot     https://github.com/minetest/minetest/issues/7307 -- Fix luajit include not being found by rubenwardy
17:50 rubenwardy    will merge in 15
17:54 rubenwardy    so, remove secure.enable_sandbox? XD
18:01 Krock         rubenwardy, the preview images seem to load forever
18:01 rubenwardy    only one exists
18:02 Krock         then shouldn't it display the "no screenie available" image?
18:02 rubenwardy    it does?
18:02 rubenwardy    that's the empty frame
18:02 Krock         ah I see. Saw the [x] image but that's for the download error
18:03 rubenwardy    awards should be the only photo there
18:03 rubenwardy    is it showing as X?
18:05 rubenwardy    merging #7307
18:05 ShadowBot     https://github.com/minetest/minetest/issues/7307 -- Fix luajit include not being found by rubenwardy
18:05 Krock         no, awards is loading correctly
18:06 Krock         are the images optipng'd/crushed yet?
18:06 rubenwardy    ah no
18:07 Krock         just asking. IIRC there was something about that we will optimize everything before release
18:07 rubenwardy    done and pushed
18:08 rubenwardy    why did you get an error?
18:10 Krock         I saw the [x] image in the diff/PR, not in the mainmenu
18:11 rubenwardy    aah
18:11 rubenwardy    good
18:11 rubenwardy    is it obvious that it's an error?
18:14 Krock         yes.
18:14 rubenwardy    cool
18:15 rubenwardy    interesting opened times: https://i.rubenwardy.com/5bE2v.png
18:15 rubenwardy    one PR was opened 24x longer ago than the other
18:16 Krock         at some point all PRs will somewhen be existing 24x longer than another
18:17 rubenwardy    well, #7307 won't have been
18:17 ShadowBot     https://github.com/minetest/minetest/issues/7307 -- Fix luajit include not being found by rubenwardy