| Time |
Nick |
Message |
| 00:00 |
|
lhofhansl joined #luanti-dev |
| 00:00 |
lhofhansl |
Merging #16307 in a few |
| 00:01 |
ShadowBot |
https://github.com/luanti-org/luanti/issues/16307 -- Fix Camera's far clipping plane for large zooms by lhofhansl |
| 00:11 |
lhofhansl |
Done... |
| 01:06 |
MTDiscord |
<josiah_wi> Interesting. I always just assumed the rule to avoid non-const references was about out-parameters in general. |
| 02:44 |
|
aliasalreadytake joined #luanti-dev |
| 04:00 |
|
MTDiscord joined #luanti-dev |
| 06:26 |
|
YuGiOhJCJ joined #luanti-dev |
| 06:39 |
|
PtiLuky joined #luanti-dev |
| 08:01 |
sfan5 |
i thought so too |
| 08:01 |
sfan5 |
if you have int a; some_function(a, b, c); it's not clear if a, b, or c will be modified |
| 08:01 |
sfan5 |
unless you avoid using non-const references |
| 08:41 |
|
Desour joined #luanti-dev |
| 08:55 |
|
SFENCE joined #luanti-dev |
| 11:26 |
sfan5 |
any more reviews on #14135? |
| 11:26 |
ShadowBot |
https://github.com/luanti-org/luanti/issues/14135 -- Add persistent unique identifiers for objects by sfence |
| 12:44 |
|
sugarbeet joined #luanti-dev |
| 12:44 |
sfan5 |
can someone explain this https://github.com/flathub/org.luanti.luanti/pull/13 |
| 13:25 |
MTDiscord |
<luatic> doesn't really make sense. the commit it wants to pin it to is the exact same as the tag is already pointing to, and the other change is just a cosmetic change? |
| 13:28 |
sfan5 |
i guess we can merge it then before it keeps opening that same PR |
| 13:31 |
MTDiscord |
<luatic> it should be harmless to merge |
| 13:37 |
[MatrxMT] |
<grorp> as long as it doesn't result in an update for users |
| 13:46 |
sfan5 |
https://github.com/flathub/org.luanti.luanti/pull/12 which I merged earlier will anyway |
| 13:55 |
|
Desour joined #luanti-dev |
| 13:58 |
|
SFENCE joined #luanti-dev |
| 14:17 |
|
SFENCE joined #luanti-dev |
| 14:50 |
|
SFENCE joined #luanti-dev |
| 14:53 |
|
guset joined #luanti-dev |
| 14:57 |
|
redundantcc joined #luanti-dev |
| 15:24 |
|
SFENCE joined #luanti-dev |
| 15:24 |
|
RudyaFunkerPixVo joined #luanti-dev |
| 15:25 |
|
RudyaFunkerPixVo left #luanti-dev |
| 16:11 |
|
SFENCE joined #luanti-dev |
| 16:56 |
|
SFENCE joined #luanti-dev |
| 17:11 |
|
SFENCE joined #luanti-dev |
| 17:12 |
|
PtiLuky joined #luanti-dev |
| 17:15 |
PtiLuky |
Regarding the non-const ref, the argument is that what a method does is clear from *its signature*, not just from its call. |
| 17:15 |
PtiLuky |
And this brings the possibility to bring the information about "is nullptr a valid parameter" instead |
| 17:17 |
PtiLuky |
A method should be documented throught its signature, not the call. |
| 17:17 |
PtiLuky |
For instance a constructor "ComplexNumber(float real, float imaginary)" is clear. |
| 17:17 |
PtiLuky |
The call "ComplexNumber myNum(1.2, 2.3)" is not clear and is not expected to be, you'd look for this information in the signature |
| 17:18 |
PtiLuky |
But in the end, this is a coding style choice, both sides have valid argument and I'm not sure we'd want to change the decision (just wanted to explain why I said I preferred this way) |
| 17:23 |
MTDiscord |
<luatic> I think the callsite visibility of references is fine if you use something like clangd. |
| 18:26 |
sfan5 |
if nobody minds I plan to import weblate later |
| 18:51 |
|
PtiLuky joined #luanti-dev |
| 19:44 |
|
Rggg90 joined #luanti-dev |
| 19:46 |
|
Rggg95 joined #luanti-dev |
| 19:46 |
|
Rggg93 joined #luanti-dev |
| 19:51 |
|
Rggg1 joined #luanti-dev |
| 20:57 |
|
PtiLuky joined #luanti-dev |
| 21:59 |
sfan5 |
https://github.com/luanti-org/luanti/commits/ci/ commits |
| 22:17 |
sfan5 |
will merge once ci passes |
| 22:32 |
|
panwolfram joined #luanti-dev |
| 22:38 |
sfan5 |
someone else will have to run updatepo some time |
| 23:05 |
|
Eragon joined #luanti-dev |
| 23:23 |
|
YuGiOhJCJ joined #luanti-dev |
