Time |
Nick |
Message |
06:05 |
|
fluxflux joined #minetest-dev |
06:40 |
|
NetherEran joined #minetest-dev |
06:43 |
|
proller joined #minetest-dev |
07:43 |
|
proller joined #minetest-dev |
08:00 |
|
ShadowNinja joined #minetest-dev |
09:08 |
|
calcul0n joined #minetest-dev |
09:59 |
|
NetherEran joined #minetest-dev |
10:32 |
|
Taoki joined #minetest-dev |
11:05 |
|
Fixer joined #minetest-dev |
12:43 |
|
Qiangong2[m] joined #minetest-dev |
13:04 |
|
ShadowNinja joined #minetest-dev |
13:04 |
|
basxto joined #minetest-dev |
13:05 |
|
Fixer joined #minetest-dev |
13:09 |
|
jas_ joined #minetest-dev |
13:19 |
|
turtleman joined #minetest-dev |
13:31 |
|
anon5[m] joined #minetest-dev |
13:31 |
|
kb1000 joined #minetest-dev |
13:31 |
|
Kimapr[m] joined #minetest-dev |
13:56 |
|
djenkins_ joined #minetest-dev |
13:57 |
|
dennisjenkins joined #minetest-dev |
15:05 |
|
calcul0n joined #minetest-dev |
15:47 |
|
proller joined #minetest-dev |
15:48 |
|
calcul0n_ joined #minetest-dev |
16:11 |
|
fluxflux joined #minetest-dev |
17:29 |
|
mizux joined #minetest-dev |
18:15 |
Krock |
josiah_wi: any success? |
18:18 |
Krock |
#9417 easy review |
18:18 |
ShadowBot |
https://github.com/minetest/minetest/issues/9417 -- Use "Aux" key name consistently by Wuzzy2 |
18:28 |
|
Miner_48er joined #minetest-dev |
18:37 |
|
Fixer_ joined #minetest-dev |
19:20 |
|
homthack joined #minetest-dev |
19:33 |
|
Andrey01 joined #minetest-dev |
19:33 |
Andrey01 |
New PR #10471 |
19:33 |
ShadowBot |
https://github.com/minetest/minetest/issues/10471 -- Add `table.len` method. by Andrey2470T |
19:37 |
Krock |
use-cases? |
19:42 |
Andrey01 |
read the PR description |
19:56 |
Krock |
I don't see any use-case there |
19:56 |
Krock |
it's just demonstrating what the function does, not where it could effectively be used in mods |
19:58 |
MTDiscord |
<Jordach> i overuse numerical keys because the # operator exists |
19:59 |
Andrey01 |
I don`t know, I think elsewhere where getting of table length is used, it is difficult to say |
20:00 |
Krock |
if length of the table matters, you'd use numeric indexing anyway. for non-numeric keys (mostly strings) you don't care about the amount |
20:02 |
|
proller joined #minetest-dev |
20:02 |
Andrey01 |
why do you say you don`t care about the amount? really are there not cases where you need to count amount of fields in an associative arrays? |
20:03 |
MTDiscord |
<GreenXenith> There are definitely cases where you care about the amount but index by string because it will lookup faster. In such a case, it is best to have a length key that you increment when adding values. |
20:04 |
Andrey01 |
just all problem in # operator is it will return incorrect count if counting is implemented in asscociative arrays |
20:05 |
MTDiscord |
<GreenXenith> I dont think table.getn is deprecated in 5.1 (it is later though) so for Minetest specifically you can increment an n key yourself in a non-integer-indexed table, so table.getn will work on both your table and numerically indexed tables. |
20:06 |
MTDiscord |
<GreenXenith> Or just make your own function to return table.n or #table |
20:06 |
MTDiscord |
<GreenXenith> (probably the better option) |
20:06 |
Andrey01 |
what is table.getn? |
20:06 |
MTDiscord |
<GreenXenith> google is your friend |
20:06 |
celeron55 |
i think it's good practice in Lua to not use a counting len() like that |
20:07 |
Andrey01 |
here is no such method |
20:07 |
Andrey01 |
https://www.lua.org/manual/5.4/manual.html#6.6 |
20:07 |
MTDiscord |
<GreenXenith> it doesnt exist in 5.4 |
20:08 |
MTDiscord |
<Jonathon> https://www.lua.org/pil/19.1.html or https://www.google.com/search?q=table.getn&oq=table.getn&aqs=chrome.0.69i59.2164j0j1&sourceid=chrome&ie=UTF-8 |
20:09 |
MTDiscord |
<GreenXenith> regardless getn can be replaced with a oneliner functino |
20:09 |
MTDiscord |
<GreenXenith> function* |
20:10 |
MTDiscord |
<GreenXenith> But if you know the table you are working with has a length key, you dont even need a function |
20:13 |
|
DS-minetest joined #minetest-dev |
20:13 |
MTDiscord |
<GreenXenith> you can also give your table a metatable to count for you and return the value with the # operator |
20:16 |
pgimeno |
if it is implemented, I would not call it len, maybe count |
20:17 |
DS-minetest |
GreenXenith: afaik, the # operator can't be overridden for tables |
20:17 |
MTDiscord |
<GreenXenith> __len |
20:19 |
pgimeno |
doesn't work for me |
20:19 |
MTDiscord |
<GreenXenith> works for me |
20:19 |
DS-minetest |
GreenXenith: http://www.lua.org/manual/5.1/manual.html#2.8 see the elseif type(op) == "table" |
20:19 |
pgimeno |
a=setmetatable({},{__len=function() return 9 end}) ; print(#a) gives 0 |
20:19 |
MTDiscord |
<GreenXenith> thats not how you use setmetatable |
20:20 |
DS-minetest |
you can only override operators that would otherwise not work |
20:20 |
MTDiscord |
<GreenXenith> a={}; setmetatable(a, {__len=function() return 9 end}); print(#a) |
20:20 |
MTDiscord |
<GreenXenith> this works |
20:20 |
pgimeno |
returns 0 for me |
20:21 |
pgimeno |
are you sure you're using Lua 5.1? |
20:21 |
DS-minetest |
http://www.lua.org/manual/5.1/manual.html#pdf-setmetatable setmetatable returns the first parameter |
20:21 |
DS-minetest |
@GreenXenith |
20:22 |
DS-minetest |
seems to work for lua5.4 |
20:22 |
MTDiscord |
<GreenXenith> Yeah thats what ive just figured |
20:22 |
MTDiscord |
<GreenXenith> Should work on 5.3 as well |
20:22 |
MTDiscord |
<GreenXenith> Minetest should really update |
20:22 |
pgimeno |
it is implemented in 5.2 |
20:22 |
MTDiscord |
<GreenXenith> BuT WhY wOuLD i EveR NeEd ThOSe FeATuReS?? |
20:23 |
Andrey01 |
I think it would be better to use function, not overriding an operator |
20:23 |
pgimeno |
LuaJIT is stuck at 5.1, and it's essential to run servers at a half-decent speed |
20:23 |
MTDiscord |
<GreenXenith> the override operator is just for convenience |
20:24 |
MTDiscord |
<GreenXenith> the underlying method is still incrementing a length key |
20:24 |
pgimeno |
__len is implemented in LuaJIT with a compilation option, http://luajit.org/extensions.html (search __len in the page) |
20:25 |
DS-minetest |
you could set __newindex in the metatable to increase n (first check if it was nil) |
20:25 |
MTDiscord |
<GreenXenith> right |
20:25 |
MTDiscord |
<GreenXenith> that part of the metatable is still valid |
20:31 |
|
Fixer joined #minetest-dev |
21:34 |
|
Kray joined #minetest-dev |
22:52 |
|
Okazaki joined #minetest-dev |
22:53 |
DS-minetest |
clipping of boxes doesn't seem to work anymore |
22:54 |
sfan5 |
huh? |
22:56 |
DS-minetest |
no wonder, #8676 sets the clip rect to nullptr |
22:56 |
ShadowBot |
https://github.com/minetest/minetest/issues/8676 -- Add gradients and borders to FormSpec boxes by v-rob |
23:07 |
DS-minetest |
made a fix: #10473 |
23:07 |
ShadowBot |
https://github.com/minetest/minetest/issues/10473 -- Fix box[]es not being clipped by Desour |