| Time |
Nick |
Message |
| 00:37 |
|
stormchaser3000 joined #minetest |
| 00:56 |
|
Trifton joined #minetest |
| 02:02 |
|
stormchaser3000 joined #minetest |
| 02:12 |
|
TheSilentLink joined #minetest |
| 02:14 |
|
nopjmp joined #minetest |
| 02:20 |
|
Noisytoot joined #minetest |
| 02:21 |
|
nopjmp joined #minetest |
| 03:26 |
|
stormchaser3000 joined #minetest |
| 04:00 |
|
MTDiscord joined #minetest |
| 04:17 |
|
Verticen joined #minetest |
| 04:18 |
|
fluxionary joined #minetest |
| 06:17 |
|
tarsovbak joined #minetest |
| 06:19 |
|
gregon joined #minetest |
| 06:28 |
|
tarsovbak joined #minetest |
| 06:41 |
|
dibesfer joined #minetest |
| 06:46 |
|
Verticen joined #minetest |
| 07:00 |
|
TomTom joined #minetest |
| 07:21 |
|
lemonzest joined #minetest |
| 07:25 |
|
Hanicef joined #minetest |
| 07:59 |
|
gregon joined #minetest |
| 08:01 |
|
tarsovbak joined #minetest |
| 09:07 |
|
Noisytoot joined #minetest |
| 09:15 |
|
Noisytoot joined #minetest |
| 09:22 |
|
Tito joined #minetest |
| 09:57 |
|
zxandris75 joined #minetest |
| 10:02 |
|
ireallyhateirc joined #minetest |
| 10:18 |
|
zxandris75 joined #minetest |
| 10:47 |
|
tarsovbak joined #minetest |
| 10:49 |
|
gregon joined #minetest |
| 12:06 |
|
gregon1 joined #minetest |
| 12:11 |
|
Noisytoot joined #minetest |
| 12:19 |
|
Noisytoot joined #minetest |
| 12:22 |
|
tarsovbak joined #minetest |
| 12:42 |
|
Thermoriax joined #minetest |
| 12:58 |
|
amfl2 joined #minetest |
| 13:10 |
|
sparky4 joined #minetest |
| 13:27 |
|
lemonzest joined #minetest |
| 13:30 |
|
mrkubax10 joined #minetest |
| 13:34 |
|
mrkubax10 joined #minetest |
| 13:43 |
|
Noisytoot joined #minetest |
| 14:01 |
|
f_ joined #minetest |
| 14:19 |
|
Noisytoot joined #minetest |
| 14:37 |
|
Noisytoot joined #minetest |
| 14:42 |
|
s20 joined #minetest |
| 14:46 |
|
definitelya joined #minetest |
| 14:49 |
|
sparky4 joined #minetest |
| 14:50 |
|
lemonzest joined #minetest |
| 14:50 |
|
amfl2 joined #minetest |
| 14:51 |
|
Thermoriax joined #minetest |
| 15:06 |
|
Noisytoot joined #minetest |
| 15:12 |
|
___nick___ joined #minetest |
| 15:19 |
|
ireallyhateirc joined #minetest |
| 15:19 |
|
f_ joined #minetest |
| 15:51 |
|
Noisytoot joined #minetest |
| 15:55 |
|
Verticen joined #minetest |
| 16:07 |
|
___nick___ joined #minetest |
| 16:09 |
|
fluxionary joined #minetest |
| 16:11 |
|
___nick___ joined #minetest |
| 16:11 |
|
Noisytoot joined #minetest |
| 16:12 |
|
Talkless joined #minetest |
| 16:21 |
|
s20 joined #minetest |
| 16:22 |
|
fluxionary joined #minetest |
| 16:29 |
|
Noisytoot joined #minetest |
| 16:33 |
|
kamdard joined #minetest |
| 16:37 |
|
turtleman joined #minetest |
| 16:39 |
|
Leopold joined #minetest |
| 16:49 |
|
est31 joined #minetest |
| 16:49 |
|
Noisytoot joined #minetest |
| 17:01 |
|
grorp joined #minetest |
| 17:54 |
|
shaft joined #minetest |
| 17:55 |
shaft |
I don't understand why my PR was closed and seemingly no one discusses it anymore. The mod is clearly not working as intended with the current release of MT 5.8. https://github.com/minetest-mods/ts_furniture/pull/20 |
| 17:58 |
|
Glaedr joined #minetest |
| 17:59 |
grorp |
shaft: regarding that PR, my response didn't mean that there is no bug, but that the bug might be in MTG instead of the engine |
| 18:00 |
shaft |
It requires default. It's a mod for MTG |
| 18:01 |
grorp |
wrong. default is not a mod for MTG, but a mod from MTG / part of MTG. |
| 18:02 |
grorp |
ah lol just misunderstood |
| 18:03 |
shaft |
How am I supposed to test where the bug is from? Mineclone doesn't have default and I'm not sure which games have default and nothing else from MTG |
| 18:06 |
grorp |
I came to this conclusion because the bug didn't go away when using old engine versions, but it did go away when using old MTG versions |
| 18:07 |
shaft |
Why didn't you say so earlier. Thank you. That sounds very conclusive. |
| 18:09 |
grorp |
sorry, I thought it was obvious from my comment. I should have been more explicit. |
| 18:10 |
MTDiscord |
<luatic> grorp, shaft: it's a feature :P, see https://github.com/minetest/minetest_game/blob/master/mods/player_api/init.lua - different animations get different eye heights in newer MTG versions, so adjusting it yourself should be unnecessary |
| 18:10 |
MTDiscord |
<luatic> (as in, you may not adjust it yourself, otherwise you're applying it twice) |
| 18:11 |
shaft |
And other games don't? |
| 18:11 |
MTDiscord |
<luatic> it depends on the game |
| 18:11 |
MTDiscord |
<luatic> i don't know how other games manage eye height / eye offsets |
| 18:11 |
MTDiscord |
<luatic> many are probably fine leaving it entirely static |
| 18:12 |
shaft |
How is anyone supposed to make mods work with multiple games? |
| 18:12 |
shaft |
I don't like this philosophy tbh |
| 18:12 |
MTDiscord |
<luatic> making mods work with multiple games is inherently a nontrivial problem |
| 18:12 |
MTDiscord |
<luatic> games can be arbitrarily different, not everything will be a MTG-esque mod soup |
| 18:13 |
shaft |
Would someone reopen my issue and merge my fix if I check for MTG and version number? |
| 18:14 |
MTDiscord |
<luatic> just account for eye height |
| 18:14 |
shaft |
MTG version > 5.5.0 |
| 18:14 |
MTDiscord |
<luatic> no need for a version check |
| 18:14 |
MTDiscord |
<luatic> the code is currently assuming a certain eye height |
| 18:14 |
MTDiscord |
<luatic> newer MTG versions break this assumption by having variable eye height |
| 18:15 |
MTDiscord |
<luatic> simply take the difference between the expected and actual eye height and add it to the offset |
| 18:19 |
MTDiscord |
<luatic> commented (the same) on the issue |
| 18:19 |
|
jaca122 joined #minetest |
| 18:22 |
shaft |
Games other than MTG don't necessarily have player_api either |
| 18:23 |
MTDiscord |
<luatic> so? |
| 18:23 |
MTDiscord |
<luatic> use player_api if present, otherwise don't |
| 18:23 |
MTDiscord |
<luatic> games other than MTG needn't have a player model |
| 18:23 |
MTDiscord |
<luatic> or the animation you want |
| 18:23 |
MTDiscord |
<luatic> they may map frame ranges entirely differently |
| 18:24 |
MTDiscord |
<luatic> ts_furniture depends on default tho |
| 18:24 |
MTDiscord |
<luatic> so it seems to be a mod tailored for MTG (default is the "core" MTG mod) |
| 18:24 |
MTDiscord |
<luatic> it should explicitly depend on player_api as well though |
| 18:25 |
MTDiscord |
<luatic> imma open the feature request for randomizing load order |
| 18:26 |
MTDiscord |
<warr1024> Generally if you want your mod to work with every game, you are going to have to actually add support for every game. There isn't necessarily a point you will reach where you've covered every possible game. |
| 18:26 |
MTDiscord |
<luatic> it's pretty much trivial to implement, just need to make sure people are fine with it |
| 18:27 |
|
sparky4 joined #minetest |
| 18:33 |
shaft |
I'm sorry to have you spoonfeeding me but how do I get the world eyeheight and eyeheight of the model? |
| 18:38 |
MTDiscord |
<luatic> There is no world eyeheight. There is the eyeheight object property of the player. You can get that via player:get_properties().eye_height. |
| 18:55 |
shaft |
I understood now. |
| 18:58 |
jonadab |
I still think default: has way too many things in it. It should have, like, air and stone, and everything else should be from other modules (even if they're ones included in MTG by default). |
| 18:59 |
jonadab |
Though it's far too late to change that now; it would invalidate too many add-on modules. |
| 19:02 |
Krock |
don't depend on default if you don't want it |
| 19:03 |
ireallyhateirc |
yeah, its only value is supporting mods that were written for it back when it was official |
| 19:03 |
ireallyhateirc |
most features from MTG can be redone in 1 month if you're motivated |
| 19:03 |
ireallyhateirc |
because it's mostly node definitions and biome definitions |
| 19:05 |
shaft |
How do I fiddle the changes upstream back into my PR repository? There is probably a super easy solution I don't know about. |
| 19:22 |
|
kamdard joined #minetest |
| 19:25 |
jonadab |
Sure. |
| 19:34 |
shaft |
Okay but I don't know how, so I did it the stupid way now. Is someone going to reopen my PR or should I make a new one? |
| 20:00 |
|
silverwolf73828 joined #minetest |
| 20:04 |
|
liceDibrarian joined #minetest |
| 20:05 |
|
Helenah joined #minetest |
| 20:21 |
|
dumpster_lord joined #minetest |
| 20:21 |
dumpster_lord |
I encountered a weird bug where vector.equals returns false for vectors that look equal |
| 20:22 |
dumpster_lord |
I have this code: https://codeberg.org/lord_of_the_dumpster/perfect_city/src/commit/e96930bc92e1c1050e9500ef9ee90065f2105809/mods/pcity_mapgen/roads_layout.lua#L234 |
| 20:22 |
dumpster_lord |
and this condition: https://codeberg.org/lord_of_the_dumpster/perfect_city/src/commit/e96930bc92e1c1050e9500ef9ee90065f2105809/mods/pcity_mapgen/roads_layout.lua#L241 |
| 20:25 |
dumpster_lord |
and here is the output from console: https://paste.debian.net/1316214/ |
| 20:27 |
dumpster_lord |
asking here before it turns out I'm doing something wrong |
| 20:29 |
|
grorp joined #minetest |
| 20:30 |
grorp |
does comparing the components yourself give the same result? |
| 20:31 |
dumpster_lord |
I'll check, gimmie a sec |
| 20:34 |
dumpster_lord |
I used this: if vec.x ~= vec2.x or vec.z ~= vec2.z or vec.y ~= vec2.y then |
| 20:34 |
dumpster_lord |
and still the same effect |
| 20:36 |
dumpster_lord |
The code in builtin/common is trivial: https://github.com/minetest/minetest/blob/1aba7f1fde101819f15ee2c28d52096605ce5253/builtin/common/vector.lua#L72 |
| 20:36 |
dumpster_lord |
Possibly something buggy hiding in the metatable stuff? |
| 20:37 |
dumpster_lord |
All vectors are made with vector.new |
| 20:45 |
grorp |
unlikely. sounds more like floating point imprecision. does adding a small tolerance give the same result? |
| 20:47 |
dumpster_lord |
I'll try forcibly running math.floor on them |
| 20:48 |
MTDiscord |
<savilli> It definitely looks like a floating point issue. But dump is supposed to write numbers with the maximal precision. |
| 20:51 |
dumpster_lord |
doing vector.floor helped |
| 20:51 |
dumpster_lord |
I'm puzzled why dump() didn't show it though |
| 20:53 |
dumpster_lord |
So I guess I should stop trusting console prints and remove all floats from my code |
| 20:54 |
sfan5 |
bad news: lua only has float numbers |
| 20:54 |
dumpster_lord |
Is dump() provided by lua or minetest? |
| 20:55 |
dumpster_lord |
can we somehow make it at least print the numbers are different? |
| 20:57 |
grorp |
by minetest |
| 20:57 |
dumpster_lord |
So I guess I can start an issue for this on github? |
| 21:02 |
MTDiscord |
<luatic> dumpster_lord, grorp: basic_dump in builtin/common/misc_utils.lua seems to tostring numbers, which loses precision. it should use something like ("%.17g"):format(num) instead. |
| 21:03 |
MTDiscord |
<luatic> (though that will result in much more verbose output for some cases..) |
| 21:04 |
dumpster_lord |
by "it" you mean dump() code or my debug prints? |
| 21:04 |
MTDiscord |
<luatic> dump() code. your debug prints too. |
| 21:04 |
MTDiscord |
<luatic> print tostrings behind the scenes. tostring on numbers does not print enough significant digits; it rounds. |
| 21:05 |
MTDiscord |
<luatic> "remove all floats from code" is not a solution anywhere. yes, floats are imprecise. no, they aren't always imprecise. you need to know when they are how imprecise (very roughly), but also when they are precise. for example using floats as integers is precise in certain (relatively large) ranges. |
| 21:06 |
dumpster_lord |
I'll write my own vector.sloppy_equals for that |
| 21:13 |
|
chilledfrogs joined #minetest |
| 21:14 |
MTDiscord |
<luatic> ideally don't "pollute" the vector table |
| 21:15 |
MTDiscord |
<bastrabun> I am looking at a broken set_eye_offset problem. Would appgurueu be among us? |
| 21:22 |
|
TheSilentLink joined #minetest |
| 21:25 |
MTDiscord |
<luatic> sure, what's the matter? |
| 21:28 |
MTDiscord |
<bastrabun> I'm an idiot, I should have connected Lars == Lars. I tried to repro the issue and found skinsdb and 3d_armor lack a certain line, you were the last to try and repro in ts_furniture. I had a question, which is now obsolete, I found the solution. Looks like I need to bug both skinsdb and 3d_armor to add that line to their register_model call |
| 21:30 |
MTDiscord |
<luatic> I suppose they leave eye height static when registering models, while player_api has it dynamic by default? |
| 21:30 |
MTDiscord |
<bastrabun> You proposed in ts_furniture https://github.com/minetest-mods/ts_furniture/pull/20 adding the eye offset to the affected mods. Would it not be smarter to add the definiton to the offending mods skinsdb and 3d_armor? |
| 21:31 |
MTDiscord |
<bastrabun> No, its about the model definiton |
| 21:31 |
MTDiscord |
<bastrabun> Last comment in https://gitea.your-land.de/your-land/bugtracker/issues/6760 |
| 21:31 |
MTDiscord |
<luatic> yes, i'm aware |
| 21:32 |
MTDiscord |
<luatic> note that player_api registers eye_height overrides for the animations while skinsdb does not |
| 21:32 |
MTDiscord |
<luatic> specifically for sit and lay |
| 21:33 |
MTDiscord |
<bastrabun> beds and boats have the same issue, its not about emote or ts_furniture |
| 21:33 |
MTDiscord |
<luatic> yes |
| 21:33 |
MTDiscord |
<luatic> so the problem with ts_furniture (and beds and boats) is, it has code that sets an eye offset to lower the effective eye height. but this may not be done now that player_api already does that based on the animation. |
| 21:34 |
MTDiscord |
<luatic> now what further complicates the issue is that 3d_armor and skinsdb override the model def, undoing these per-animation eye height overrides |
| 21:35 |
MTDiscord |
<bastrabun> If I add the line regarding eye_height to player_api.register_model defintion, at least boats, emote and beds work as expected. |
| 21:35 |
MTDiscord |
<luatic> yes |
| 21:35 |
MTDiscord |
<luatic> one way to fix everything for the newest versions of player_api would be: give 3d_armor and skinsdb these missing overrides (and the eye_height field), remove the set_eye_offset from ts_furniture |
| 21:36 |
MTDiscord |
<bastrabun> Ok, so there is already a solution planned and I did a lot of stuff otehr had already done : D Meh 😛 |
| 21:36 |
MTDiscord |
<luatic> idk whether i would call this "planned" :P |
| 21:36 |
MTDiscord |
<luatic> i'm not maintaining 3d_armor or skinsdb or ts_furniture, just saying what one way to fix it would be |
| 21:37 |
MTDiscord |
<luatic> (if you want to support older versions of player_api too, you need to be a bit cleverer, because then you may still need the set_eye_offset. you could do something like a player_api version check by feature, or you could just get the eye height and force the eye height you expect via an eye offset.) |
| 21:39 |
MTDiscord |
<bastrabun> I see, that's where the problems start. Anyways, that's the decision of the maintainers then |
| 21:39 |
MTDiscord |
<luatic> let me comment on your bugtracker... |
| 21:40 |
MTDiscord |
<luatic> how do you manage mods? do you run forks / patchsets? |
| 21:42 |
MTDiscord |
<bastrabun> We mirror upstream, then fork them, then PR changes in master to our stable branch |
| 21:43 |
MTDiscord |
<bastrabun> All I need to know is that I'm not sitting on someting that needs to go upstream |
| 21:44 |
MTDiscord |
<bastrabun> When upstream deals with it and knows what's up, I'm good, duty done. They will fix in their due time, hopefully until we deliver the next modset update |
| 21:45 |
|
TheSilentLink joined #minetest |
| 21:47 |
MTDiscord |
<luatic> commented: https://gitea.your-land.de/your-land/bugtracker/issues/6760#issuecomment-81251 |
| 21:48 |
MTDiscord |
<bastrabun> Thanks 🙂 |
| 21:48 |
MTDiscord |
<luatic> np |
| 21:49 |
shaft |
https://github.com/minetest-mods/ts_furniture/pull/21 if anyone wants to test |
| 22:32 |
|
panwolfram joined #minetest |
| 22:35 |
|
Lunatrius joined #minetest |
| 22:46 |
|
Lunatrius joined #minetest |
