| Time |
Nick |
Message |
| 00:20 |
|
FeXoR joined #luanti |
| 01:39 |
|
SFENCE joined #luanti |
| 02:56 |
[MatrxMT] |
<Blockhead256> eh? that's a bug or bad build, I regularly switch to German to test things |
| 03:05 |
cheapie |
Air conditioner is bringing in plenty of fresh air: https://cheapiesystems.com/media/2025-12-13%2021-03-02.webm |
| 03:07 |
[MatrxMT] |
<Blockhead256> for the unfamiliar, that texture is built into the engine for when you have air in your inventory |
| 03:43 |
cheapie |
Random screenshot of the translation work I was doing on celevator a bit ago: https://cheapiesystems.com/media/images/Screenshot_20251213_213830.png |
| 03:44 |
cheapie |
I just did a Spanish locale for now (and probably not a very good one, it's a mix of machine translation and what bits of Spanish I picked up in high school, but good enough for testing) |
| 03:45 |
cheapie |
There's a user on VE-C who has offered previously to do a German translation of it, I'll have to see if he still wants to do that |
| 03:46 |
MTDiscord |
<wsor4035> ive sometimes wondered since this is open source, doing shitty machine/ai/llm/whatever translations to bait people into making them better. basically the old reddit/stackoverflow/forum thought of post a question, answer it wrong and youll get a bunch of engagement of people telling off the wrong person |
| 03:47 |
MTDiscord |
<wsor4035> (answering it wrong with another account) |
| 03:48 |
cheapie |
I have to wonder if that bakedclay mod is trying that |
| 03:49 |
[MatrxMT] |
<Blockhead256> people definitely complain about poor l10n, but how much translates (heh) into action I'm not sure. It's not as easy as typing a diatribe on reddit |
| 03:50 |
cheapie |
Although the particular way in which the translations in that one suck suggests someone is doing something like S("Black")..S("Baked Clay") in there somewhere |
| 03:50 |
MTDiscord |
<wsor4035> im just happy when tenplus1 isnt increamenting the license violation counter, my bar is that low 🤷 |
| 03:51 |
cheapie |
I mean, they clear a higher bar than that, they at least make some sense if you think about them a bit |
| 03:51 |
MTDiscord |
<wsor4035> now yes (hopefully, havent checked in a while), the past no |
| 03:51 |
cheapie |
I mean these specific translations, not tenplus1 as a whole |
| 03:52 |
MTDiscord |
<wsor4035> ahhh |
| 03:53 |
MTDiscord |
<wsor4035> blockhead: fair |
| 03:53 |
cheapie |
Like the black baked clay says "Negro Barro Cocido", which reads approximately as "cooked mud black" |
| 03:53 |
MTDiscord |
<wsor4035> i kinda want to try it for kicks just to see what happens tho. scientific method and all |
| 03:54 |
MTDiscord |
<wsor4035> course im not sure how you would measure to make sure you actually have a trash translation vs somehow a good or good enough one that no one bothers |
| 03:54 |
*Â cheapie |
gives wsor4035 a pile of rejected machine translations from celevator to use, things like "car status" but with "car" as in "automobile" |
| 03:54 |
[MatrxMT] |
<Blockhead256> altman et. al. have already unleashed enough experiments on us. The LLM has definitely changed the way we speak – or punctuate.. |
| 03:55 |
user333_ |
an LLM's favorite character: – |
| 03:55 |
MTDiscord |
<wsor4035> course the latter would be kinda ideal as a base line/init if it works - then can be usable to people who dont know English and doable by people who dont know other langauges |
| 03:55 |
mrcheese |
user333_: yea lol |
| 03:56 |
MTDiscord |
<wsor4035> before someone runs off the deep end, not proposing the shit Mozilla is doing, just for init |
| 03:57 |
MTDiscord |
<wsor4035> (at least as far as i understand, they are overwriting human translations, etc) |
| 03:57 |
[MatrxMT] |
<Blockhead256> when you use firefox ESR like us debian tortoises, you tend to forget about mozilla until the next ESR lifecycle |
| 03:58 |
MTDiscord |
<wsor4035> iirc this is there web? documentation? so the ancient software you use isnt relevant? (citation needed) |
| 03:58 |
MTDiscord |
<wsor4035> now i must check |
| 03:59 |
[MatrxMT] |
<Blockhead256> oh the MDN machine translated? ehhhh |
| 03:59 |
user333_ |
is there any good browser nowadays? chrome/chromium are too google-d, edge is too microsoft-ed |
| 03:59 |
MTDiscord |
<wsor4035> i wouldnt recommend it, but im rooting for servo to go places |
| 03:59 |
[MatrxMT] |
<Blockhead256> no, not even one. They're all either too popular or not popular enough; too bleeding edge or don't have the features |
| 04:00 |
user333_ |
make your own ig \_('_')_/ |
| 04:00 |
MTDiscord |
<wsor4035> > Japanese support articles https://tech.yahoo.com/ai/articles/mozilla-faces-backlash-introducing-ai-154356642.html |
| 04:00 |
[MatrxMT] |
<Blockhead256> and you're also going to get fingerprinted one way or another |
| 04:00 |
MTDiscord |
<wsor4035> so not mdn, firefox? support articles i guess |
| 04:01 |
[MatrxMT] |
<Blockhead256> ah, look, giving the bots unfettered push access. What could go wrong? |
| 04:01 |
user333_ |
(a lot) |
| 04:02 |
[MatrxMT] |
<Blockhead256> fetters are exactly what the bots need |
| 04:02 |
[MatrxMT] |
<Blockhead256> there's your vocabulary for the day |
| 04:02 |
MTDiscord |
<wsor4035> always maintain that ai/llm is a useful tool, used in moderation/by a skilled human its useful. its not a hammer where everything becomes a nail |
| 04:04 |
[MatrxMT] |
<Blockhead256> ai: so advanced it can replace your annoying co-workers who push without review |
| 04:04 |
MTDiscord |
<wsor4035> lel |
| 04:04 |
MTDiscord |
<wsor4035> depending on your coworkers it might produce better code than them 😛 |
| 04:05 |
user333_ |
or worse, depending on the model |
| 04:05 |
MTDiscord |
<wsor4035> practicality, ruining the implication. but not wrong |
| 04:06 |
[MatrxMT] |
<Blockhead256> back in the bad old days, I remember my mate was using GPT-3. He'd start a greentext and the model would finish it |
| 04:06 |
[MatrxMT] |
<Blockhead256> actually must have been GPT-2, pre Chat |
| 04:07 |
user333_ |
back when LLMs though people had 6 fingers |
| 04:07 |
MTDiscord |
<wsor4035> only thing so far that i "trust" ai enough to just rip on is tests. mostly since its guardrailed into the test framework and my existing code |
| 04:07 |
MTDiscord |
<wsor4035> course i still review its output, first step being do they pass lol (and if not, is there a valid reason) |
| 04:43 |
|
Verticen joined #luanti |
| 05:00 |
|
MTDiscord joined #luanti |
| 06:10 |
|
SFENCE joined #luanti |
| 06:29 |
|
sys4 joined #luanti |
| 06:42 |
|
SFENCE joined #luanti |
| 07:01 |
|
SFENCE joined #luanti |
| 07:07 |
|
SFENCE joined #luanti |
| 07:24 |
|
SFENCE joined #luanti |
| 07:43 |
|
SFENCE joined #luanti |
| 08:30 |
|
FeXoR joined #luanti |
| 09:13 |
|
YuGiOhJCJ joined #luanti |
| 09:18 |
|
YuGiOhJCJ joined #luanti |
| 09:22 |
|
Warr1024 joined #luanti |
| 09:26 |
|
mrkubax10 joined #luanti |
| 09:47 |
|
Warr1024 joined #luanti |
| 10:14 |
pgimeno |
cheapie: may I PM you about possible improvements to the celevator Spanish translation? |
| 10:46 |
|
mrcheese joined #luanti |
| 10:49 |
|
panwolfram joined #luanti |
| 10:53 |
|
jaca122 joined #luanti |
| 11:12 |
|
panwolfram joined #luanti |
| 11:34 |
|
panwolfram joined #luanti |
| 11:36 |
|
mrcheese joined #luanti |
| 11:53 |
|
panwolfram joined #luanti |
| 11:56 |
|
panwolfram joined #luanti |
| 12:04 |
|
mrkubax10 joined #luanti |
| 12:15 |
|
panwolfram joined #luanti |
| 13:40 |
|
tzenfore joined #luanti |
| 14:05 |
|
PoochInquisitor joined #luanti |
| 14:17 |
cheapie |
pgimeno: Sure, or if you want to send reasonable patches I can merge those too |
| 14:35 |
|
ireallyhateirc joined #luanti |
| 15:24 |
|
turtleman joined #luanti |
| 15:24 |
|
silverwolf73828 joined #luanti |
| 16:03 |
|
Desour joined #luanti |
| 16:09 |
|
jaca122 joined #luanti |
| 16:29 |
|
Trifton joined #luanti |
| 16:43 |
|
mrkubax10 joined #luanti |
| 16:44 |
MinetestBot |
[git] sfan5 -> luanti-org/luanti: Update MinGW libraries and toolchain 22b5c93 https://github.com/luanti-org/luanti/commit/22b5c939dc2b0576f1d50d781d1bd2703c89ab8a (2025-12-14T16:42:24Z) |
| 16:44 |
MinetestBot |
[git] sfan5 -> luanti-org/luanti: Apply some general fixes to porting code 5b05fcf https://github.com/luanti-org/luanti/commit/5b05fcf8e1dfbbf66527acd257cb4b32dcc37fef (2025-12-14T16:42:42Z) |
| 16:44 |
MinetestBot |
[git] sfan5 -> luanti-org/luanti: Fix problem with _LIBCPP_ENABLE_ASSERTIONS 5dee294 https://github.com/luanti-org/luanti/commit/5dee294a72d906878b46baad220127ed222159e1 (2025-12-14T16:42:42Z) |
| 16:44 |
MinetestBot |
[git] siliconsniffer -> luanti-org/luanti: Restore server description in menu (#16742) 9c37b46 https://github.com/luanti-org/luanti/commit/9c37b46e1f0514a3cb2b5e5d844f42fa8e044f84 (2025-12-14T16:43:00Z) |
| 16:52 |
|
SFENCE joined #luanti |
| 17:09 |
|
chilledfrogs joined #luanti |
| 17:37 |
|
tzenfore joined #luanti |
| 18:13 |
|
SFENCE joined #luanti |
| 18:32 |
cheapie |
Maybe setting these things to recirculate wasn't the greatest idea https://cheapiesystems.com/media/2025-12-13%2020-23-30.webm |
| 18:59 |
|
SFENCE_arch joined #luanti |
| 19:19 |
|
Pokey joined #luanti |
| 20:03 |
|
FeXoR joined #luanti |
| 22:10 |
cheapie |
can't have shit in Bonnthal :P https://cheapiesystems.com/media/images/Screenshot_20251214_160820.png |
| 23:08 |
|
mrcheese joined #luanti |
| 23:08 |
|
mrcheese joined #luanti |
| 23:15 |
|
YuGiOhJCJ joined #luanti |
| 23:32 |
|
panwolfram joined #luanti |
| 23:49 |
|
chilledfrogs joined #luanti |
