|
00:14
arkiuat joined
00:16
DavidSch left
00:22
arkiuat left
00:27
arkiuat joined
00:42
arkiuat left
|
|||
| disbot4 | <.landyacht.> Voldenet - good shout on .close being bugged, awaiting .socket-host got me the effect I wanted. Thanks! | 00:43 | |
|
01:10
arkiuat joined
01:13
vasko4 left,
vasko4 joined
01:16
arkiuat left
01:22
kylese left
01:23
kylese joined
01:27
arkiuat joined
01:32
arkiuat left
01:45
arkiuat joined
01:50
arkiuat left,
oodani left
01:56
oodani joined
02:03
arkiuat joined
02:07
arkiuat left
02:15
kylese left,
kylese joined
02:28
arkiuat joined
02:37
arkiuat left
02:51
arkiuat joined
02:56
arkiuat left
03:05
apogee_ntv left
03:07
apogee_ntv joined
03:15
arkiuat joined
03:22
arkiuat left
03:24
arkiuat joined
03:25
stanrifkin_ left,
stanrifkin_ joined,
stanrifkin left
03:26
stanrifkin joined
04:12
Aedil joined
04:14
deoac left
04:28
Aedil_ joined
04:29
Aedil_ left
04:50
kylese left,
kylese joined
05:20
arkiuat left
05:31
arkiuat joined
05:40
stanrifkin_ left
05:42
stanrifkin left
06:56
melezhik joined
07:28
Sgeo left
07:51
wayland76 joined
08:07
dakkar joined
08:08
kaskal left
08:09
kaskal joined
08:17
arkiuat left
08:25
arkiuat joined
08:30
arkiuat left
08:56
arkiuat joined
09:01
arkiuat left
09:15
arkiuat joined
09:20
arkiuat left
09:32
arkiuat joined
09:37
arkiuat left
09:56
arkiuat joined
10:02
arkiuat left
10:11
arkiuat joined
10:16
arkiuat left
10:28
arkiuat joined
10:32
arkiuat left
10:38
apac joined
10:55
lichtkind joined
10:59
arkiuat joined
11:03
arkiuat left
11:16
arkiuat joined
11:24
arkiuat left
11:27
librasteve_ left
11:36
arkiuat joined
11:40
arkiuat left
12:00
arkiuat joined
|
|||
| lizmat | weekly: github.com/Raku/problem-solving/issues/491 | 12:01 | |
| notable6 | lizmat, Noted! (weekly) | ||
| disbot4 | <antononcube> "DOM::Tiny" has not been updated for ≈6 years. Any recent experiences to report of how reliable or useful it is? | 12:04 | |
|
12:05
arkiuat left
|
|||
| lizmat | in my experience Sterling Hanenkamp is a good author, so I'd say there's a good chance it's reliable | 12:06 | |
| I'm not sure about its usefulness though | |||
| disbot4 | <antononcube> OK. | 12:07 | |
| lizmat | it could use an update, if for the lack of using the Raku name :-) | ||
|
12:13
apac left
12:36
arkiuat joined
12:42
arkiuat left
12:47
apac joined
|
|||
| disbot4 | <librasteve> I nearly used DOM::Tiny yesterday (to read some HTML ^^) on the advice of an LLM ... but it failed to zef install so I ended up with quite a bloated HTML::Parser::XML => XML chain ... would fill a gap between fast and dirty getting eg a table as values and precision XML parsing | 12:48 | |
| lizmat | ah, seems to have been reported recently: github.com/zostay/raku-DOM-Tiny/issues/19 | 12:53 | |
| disbot4 | <librasteve> yeah sorry was too lazy to raise an issue | 12:54 | |
| <antononcube> @librasteve I will make html-table-data-extraction available in "Data::Translators". | |||
| <librasteve> that'd be cool - I'm actually going to lean in to LLM a bit on this hobby project | 12:56 | ||
| <antononcube> See the MVP here: github.com/antononcube/Raku-Data-T...ction.raku | 12:59 | ||
| <librasteve> nice! | 13:01 | ||
|
13:01
arkiuat joined
|
|||
| disbot4 | <antononcube> No dependencies, just regexes. | 13:01 | |
| <antononcube> I need to make it more robust and write tests for it. | 13:02 | ||
| <librasteve> feature request to specify which table (if > 1 on the page) either by table class="xxx" or by just an index (0,1,2,...) | |||
| <antononcube> Hmm... that is on the robust side, I think: if no header spec is found then indexes re used. | 13:04 | ||
|
13:06
arkiuat left
|
|||
| disbot4 | <antononcube> Agh, no, you are asking for something (much) more. | 13:12 | |
| <jubilatious1_98524> @librasteve unix.stackexchange.com/a/769560/227738 | 13:15 | ||
| <librasteve> LLM must have been trained on that then | 13:16 | ||
| <antononcube> Yeah but you cam always add to the prompt: > use only built-in functionalities, do not use 3rd party modules. | 13:18 | ||
| <jubilatious1_98524> Right. LLMs are the world's greatest plagiarists. | 13:19 | ||
| <antononcube> More importantly, in that answer, did you parrot the statistical parrots yourself? | 13:22 | ||
| <jubilatious1_98524> @antononcube I don't use LLMs. I taught myself Raku by solving Unix&Linux StackExchange problems. Some I solved de novo. For others I translated awk and perl one-liners into Raku code. | 13:26 | ||
| <antononcube> I will stage some (LLM) tests to verify that claim... | 13:27 | ||
| <jubilatious1_98524> Do your worst (best)! 🤣 | 13:28 | ||
| <antononcube> This actually, is a hard problem, both methodologically and technically, with Raku code. | 13:29 | ||
| <antononcube> I refer to the obvious chicken and egg question. | 13:30 | ||
| <jubilatious1_98524> "Dear LLM, please prove your plagiarism of StackExchange Raku code. Thx." | |||
| <antononcube> The other part, is for being able to ingest stack-exchange posts/answers of a particular user. | 13:31 | ||
| <antononcube> I think that latter one was recently discussed here... | 13:32 | ||
| <jubilatious1_98524> No egg-and-chicken. My first Raku answer posted on U&L StackExchange was in 202. What LLMs were around then? unix.stackexchange.com/a/581448/227738 | 13:34 | ||
| <antononcube> Yeah, but this one is post LLM-revolution. | 13:35 | ||
|
13:35
arkiuat joined
|
|||
| disbot4 | <antononcube> Honestly, gathering those answers might make a nice book(lette). | 13:38 | |
| <jubilatious1_98524> Are you seriously suggesting that an LLM installed and tested SEVEN Raku modules to come up with a verifiable StackExchange answer? | |||
| <antononcube> This is the so called programming language LLM-plugins do/did since 2023. | 13:39 | ||
|
13:40
arkiuat left
|
|||
| disbot4 | <antononcube> For Python it is done by default, always. As far as know, by OpenAI, and I assume others. (Gemini, Claude, etc.) | 13:41 | |
| <antononcube> But, I doubt it is done for Raku. | 13:43 | ||
| <jubilatious1_98524> I'd love to see you test it, Anton. Along the way I'd love for you to discover the idiosyncratic phases I use when posting my Raku answers, such as "...formerly known as Perl_6". | 13:44 | ||
|
13:45
apac left
|
|||
| disbot4 | <jubilatious1_98524> @antononcube Wouldn't it be funny if you get better LLM answers asking for one-liner code written in Perl_6 rather than Perl6? ‼️ ❓ | 13:45 | |
| <antononcube> These sort of small changes should not matter that much with "serious" LLMs. | 13:49 | ||
| <antononcube> The problem with many LLMs is the same as with Raku -- there is no argument for specifying a random seed when the queries are made. Hence, how to make this kind of experiments. | 13:50 | ||
|
14:02
arkiuat joined
14:10
hvxgr joined
14:12
hvxgr left
14:13
hvxgr joined
14:37
tgt joined
14:38
apac joined
14:44
tgt left
|
|||
| disbot4 | <dr.shuppet> ChatGPT still doesn't know Coq has been renamed to Rocq. Hopefully the situation is better with Raku, which was renamed much earlier | 14:46 | |
|
14:47
arkiuat left
|
|||
| tonyo | . | 14:50 | |
| ugexe | the problem with chatgpt is it always refers to raku as "Raku (formerly Perl6)" | ||
| disbot4 | <dr.shuppet> I tried to teach it that Coq is now named Rocq, it saved it to its memory | 14:53 | |
| <dr.shuppet> But then it claimed to me when I used the term "Rocq" that it has been renamed into Coq | |||
| <antononcube> Yeah, the same with "Wolfram language, formally known as Mathematica." | 14:56 | ||
|
15:01
arkiuat joined
15:06
arkiuat left
|
|||
| [Coke] finally switches over to duckduckgo for majority of his mac browsing (getting rid of both chrome and firefox) | 15:12 | ||
| (getting the OS level 1password integration was key since there aren't extensions, apparently?) | 15:13 | ||
|
15:35
arkiuat joined
15:40
arkiuat left
|
|||
| melezhik | . | 15:54 | |
|
16:00
arkiuat joined
16:02
stanrifkin joined
|
|||
| [Coke] | .. | 16:04 | |
|
16:08
arkiuat left
|
|||
| melezhik | … )) | 16:14 | |
| disbot4 | <antononcube> @librasteve See : raku.land/zef:antononcube/Data::Translators | 16:23 | |
| <antononcube> See this file how to get HTML tables data into Raku datasets: github.com/antononcube/Raku-Data-T...n.rakutest | 16:25 | ||
|
16:30
arkiuat joined
16:34
dakkar left
16:39
arkiuat left
17:06
arkiuat joined
17:11
arkiuat left
17:27
arkiuat joined
17:30
oodani left
17:39
oodani joined
|
|||
| disbot4 | <librasteve> =b | 18:06 | |
|
18:07
librasteve_ joined
19:15
Aedil left
19:16
vasko4 left
19:25
vasko4 joined
21:11
arkiuat left
21:20
arkiuat joined
21:24
arkiuat left
21:28
stanrifkin left
21:54
arkiuat joined
21:55
melezhik left
21:59
arkiuat left
22:03
wayland76 left
22:08
lichtkind left
22:12
arkiuat joined
22:17
arkiuat left
22:21
DavidSch__ joined
22:25
guifa joined
22:29
arkiuat joined
22:30
apac left
22:52
arkiuat left
22:53
arkiuat joined
22:55
Sgeo joined
22:58
arkiuat left
23:27
arkiuat joined
23:32
arkiuat left
23:47
arkiuat joined
|
|||