|
00:22
soverysour joined,
soverysour left,
soverysour joined
00:27
soverysour left
01:07
sibl joined
01:30
Aedil joined,
soverysour joined
01:35
soverysour left
|
|||
| tonyo | . | 01:45 | |
|
01:55
hulk joined,
kylese left
02:10
annamalai joined
02:15
hulk left,
kylese joined
02:31
sibl left
02:33
sibl joined
02:39
soverysour joined
02:44
soverysour left
03:09
lichtkind__ joined
03:12
lichtkind_ left
03:25
sibl left,
sibl joined
03:46
soverysour joined
03:51
soverysour left
04:13
johnjay left,
johnjay joined
|
|||
| avuserow | Here's a screenshot of my terminal emulator work-in-progress: fosstodon.org/@avuserow/116395488959003887 . It has a very minimal UI and a self-reload function so I can develop it from within itself | 04:28 | |
|
04:40
soverysour joined,
soverysour left,
soverysour joined
04:45
soverysour left
04:46
Sgeo left
05:20
soverysour joined,
soverysour left,
soverysour joined
05:21
sibl left
05:25
soverysour left
05:40
sibl joined
05:47
soverysour joined,
soverysour left,
soverysour joined
05:51
soverysour left
06:09
sibl left,
soverysour joined
06:13
sibl joined
06:16
abraxxa joined
06:22
Aedil left,
abraxxa left
06:24
abraxxa joined
06:30
sibl left,
abraxxa left
06:31
abraxxa joined
06:35
soverysour left
06:45
soverysour joined,
soverysour left,
soverysour joined
06:47
sibl joined
06:58
soverysour left
07:16
sibl left
07:20
soverysour joined,
soverysour left,
soverysour joined
07:31
soverysour left
07:32
soverysour joined,
soverysour left,
soverysour joined
07:37
soverysour left
07:44
sibl joined
07:52
dakkar joined
08:12
soverysour joined
08:20
soverysour left
|
|||
| disbot2 | <librasteve> weekly: fosstodon.org/@avuserow/116395488959003887 | 08:51 | |
| notable6 | librasteve, Noted! (weekly) | ||
|
08:55
soverysour joined
09:00
soverysour left
|
|||
| apogee_ntv | librasteve_: This was the thing I was working on raku.land/zef:apogee/Selkie -- appreciate any input. | 09:39 | |
| Retained mode, opinionated (inspired by re-frame), declarative, default-safe TUI framework with native performance. | 09:41 | ||
|
09:44
sibl left
09:50
soverysour joined
09:55
soverysour left
10:15
topnep left
10:16
topnep joined
10:32
soverysour joined,
soverysour left,
soverysour joined
|
|||
| librasteve_ | apogee_ntv: this is very nice … installed deps on homebrew painlessly and then Selkie itself … counter works … so far so good | 10:36 | |
| disbot2 | <librasteve> (except I have to go Ctrl-C to quit ??) | 10:38 | |
| apogee_ntv | Ctrl-Q and Ctrl-C should both work | 10:47 | |
| If Ctrl-Q doesn't work please let me know what terminal you're on and I'll investigate | 10:49 | ||
|
10:49
sjn left,
Aedil joined
|
|||
| Voldenet | apogee_ntv: I wrote it before, but consider asciinema examples | 10:50 | |
| apogee_ntv | Voldenet: I am definitely planning to do some video demos. | 10:55 | |
| I dont know if asciinema is better than VHS with notcurses, raw mode can be a pain to automate recordings for. | |||
| Voldenet | other thing - consider typed stores, current store is too vanilla Redux js to me, maybe something like `store X { has Int $.a; }; my $store = X.new; $store.register-handler('counter/increment', -> $st { $st.a })` | ||
| maybe even typed with customized dispatch and reducers, but I'm not sure if it's supposed to support immutable store state (like redux) | 10:56 | ||
|
11:00
soverysour left
|
|||
| apogee_ntv | Typed stores are an interesting idea for sure, I've just been typing the accessors and using Monad::Maybe for optional values | 11:00 | |
| Voldenet | or even, register-handler could be per-property somehow | 11:01 | |
| apogee_ntv | You can already register multiple handlers for 1 event IIRC so thats not impossible | 11:02 | |
| Voldenet | `$store.register-handler(*.a, -> $a, %ev { … }); $store.subscribe(*.a, $widget); $store.dispatch(*.a)` | ||
| i remember that using untyped store in Redux is huge pain | 11:03 | ||
| especially if the global state grows | |||
| apogee_ntv | Yeah, it's definitely not ideal for large app state. | 11:05 | |
| Voldenet | this has nice ideas regarding syntax fco.github.io/Sourcing-2/ | 11:06 | |
| disbot2 | <librasteve> yeah Ctrl-Q no working for me on vanilla macOS Terminal | 11:08 | |
| Voldenet | you could do something like `store MyStore { has Int $.x; method increment-counter() is mutation { $!x++ }; method get-counter() is query { $!x } }` also | ||
|
11:09
soverysour joined
|
|||
| Voldenet | that would define register-handler using simplified syntax | 11:09 | |
| disbot2 | <antononcube> Everyone is in a hurry to reinvent text-based terminals. | ||
| <librasteve> github.com/m-doughty/Selkie/issues/1 | |||
| apogee_ntv | This is what I get for only testing on kitty :D | 11:10 | |
| disbot2 | <librasteve> Ctrl-P is fine btw (dashboard example) | ||
| <librasteve> @antononcube - all code is fashion ... TUI is one of the current trends (as a way to make coders look even more geeky to the non-coder folks) | 11:11 | ||
| Voldenet | antononcube: I would be more worried if they were in a hurry to reinvent picture-based terminals | 11:12 | |
| sixels are not supported on my terminal at all though | 11:13 | ||
| disbot2 | <librasteve> on asciinema - I have been messing with it (a lot) eg here cragcli.info (too much already!) | ||
| <librasteve> never tried on a TUI | |||
| <librasteve> this site uses Air::Plugin::Asciinema ... which is built on the asciinema JS player | 11:14 | ||
| <librasteve> if I were to do that again, I would consider instead to use agg (asciinema gif generator) due to the lack of layout control the JS player offers ... | 11:16 | ||
| <librasteve> TLDR: try agg if you need to share asciinema recordings | |||
| <librasteve> wonders if you could build a TUI RAG with Selkie? | 11:17 | ||
|
11:17
simcop2387 left
11:18
simcop2387_ joined
|
|||
| Voldenet | I've seen a lot of people upload things to asciinema.org directly and then embedding things in readme | 11:18 | |
| apogee_ntv | TUI is a good trend tbh, better than Electron lol | 11:19 | |
| I am so fed up of trivial apps that take 1GB+ RAM | |||
|
11:19
simcop2387_ is now known as simcop2387
|
|||
| apogee_ntv | TUI is performant and doesnt require binding and vendoring GTK or Qt | 11:20 | |
| Voldenet | everything is better than Electron, if you're making html page, why bother repackaging it :\ | ||
| and making me run 10 separate browser engines at the same time, eh | 11:21 | ||
| once you load the obnoxiously heavy browser, js webapps can get quite small on top of it, I can run 100 at the same time for "free" | 11:22 | ||
| disbot2 | <antononcube> No, it is not that. Or, more precisely, it is not just that. The main reason is "big" LLM providers prefer to facilitate the LLM access via dedicated CLI tools. | 11:31 | |
| <antononcube> And this dictates certain fashion of text terminal adoption. (Which in many ways should be a "no brainer"...) | 11:32 | ||
| librasteve_ | Voldenet: yes asciinema has a cloud service too - good point … you can use that with Air::Plugin::Asciinema too | 11:33 | |
| apogee_ntv | librasteve_: Fixed Ctrl-Q on MacOS Terminal (and any other that sets IXON/IXOFF) in Selkie 0.2.3 | 11:38 | |
| tbh I thought notcurses did this automatically. | 11:39 | ||
|
12:06
sp19833_ joined
|
|||
| librasteve_ | all good with Ctrl-Q -tx! | 12:25 | |
| apogee_ntv | Sweet, let me know if you hit anything else. BTW there are full API docs at github.com/m-doughty/Selkie/blob/m...i/index.md -- will get a site up at some point to make this stuff easier to navigate. | 12:32 | |
|
12:36
sp19833_ left
12:37
sp19833_ joined
13:03
sp19833_ left
13:04
sp19833_ joined
13:45
sjn joined
13:46
sp19833_ left
|
|||
| disbot2 | <librasteve> is there a way to check whether a zef package is a new module release? I am thinking of diff ing 360.zef.pm/ from week to week, but that seems like quite a lot of effort and there may well be a better way?? | 14:49 | |
|
14:53
sp19833_ joined
14:55
sp19833_ left,
sp19833_ joined
14:59
sp19833_ left
15:01
sp19833_ joined
15:06
sp19833_ left,
donaldh joined
15:07
donaldh_ joined
15:11
sp19833_ joined
15:12
donaldh_ left
15:16
soverysour left
15:28
sp19833_ left
|
|||
| [Coke] | raku.land/zef:coke/App::Cal, for example, shows "7 Feb 2024" as the release date for the current version. | 15:32 | |
| but that info doesn't seem to be in 360.zef.pm/ | 15:33 | ||
| wondering if raku.land is pulling that from the git repo or its logs. | |||
| a hygenic git repo has, e.g.: github.com/coke/raku-cal/releases/tag/0.9.1 | 15:34 | ||
| (but you can't rely on authors doing that even if they do use github) | 15:35 | ||
| disbot2 | <antononcube> @librasteve The most topical new package release is this one: raku.land/zef:antononcube/WWW::Hor...erisSystem | ||
| <librasteve> just in time ;-) | 15:40 | ||
|
15:41
schbax joined
|
|||
| disbot2 | <antononcube> Didn't have time to implement the 3D trajectory plotting. (In Raku.) | 15:41 | |
| <antononcube> I plan to do it this week. | |||
| <librasteve> [Coke] thanks ... yeah I already automated grabbing the "most recent" from that page, but I am trying to avoid screen scraping the actual module page release date history drop down | 15:42 | ||
| <librasteve> I will probably just make a list of all modules manually, store as keys to a .raku hash and then append to it each week ;-) | 15:44 | ||
| librasteve_ | notable6: weekly | 15:45 | |
| notable6 | librasteve_, 6 notes: gist.github.com/e2b318fa9486dc354f...a6efa72cf0 | ||
| librasteve_ | notable6: weekly reset | 15:47 | |
| notable6 | librasteve_, Moved existing notes to “weekly_2026-04-13T15:47:15Z” | ||
| librasteve_ | bot test raku.org/tools | 15:49 | |
| MOAR#768 | |||
| linkable6 | MOAR#768 [closed]: github.com/MoarVM/MoarVM/issues/768 Buncha warnings when compiling on Win10 using Strawberry Perl suite's compiler | ||
| librasteve_ | rtfm | ||
| buildable6: status | 15:50 | ||
| disbot2 | <librasteve> .hug | 15:51 | |
| huggable6 hugs librasteve | |||
|
16:25
annamalai left,
annamalai joined
16:31
topnep left
16:33
topnep joined
16:48
dakkar left
17:02
soverysour joined
17:08
soverysour left
17:36
abraxxa left
17:38
soverysour joined
|
|||
| librasteve_ | rakudoweekly.blog/2026/04/13/2026-...gs-busses/ | 18:00 | |
|
18:03
johnjay left
18:07
johnjay joined
18:25
annamalai left,
annamalai joined
18:37
topnep left,
Ekho left
18:38
topnep joined
18:39
Ekho joined
|
|||
| ugexe | m: my $x = <1>; sub foo(Int() $y) { say $y.WHAT; $y; }; say foo($x) | 18:50 | |
| evalable6 | (IntStr) 1 |
||
| ugexe | m: my $x = <1>; sub foo(Int() $y) { say $y.WHAT; $y; }; say foo($x).WHAT | 18:51 | |
| evalable6 | (IntStr) (IntStr) |
||
|
18:58
soverysour left
19:19
abraxxa joined,
abraxxa left
19:20
abraxxa joined
19:21
abraxxa left
|
|||
| lucs | irclogs.raku.About a week ago, instead of using --force-fetch to grab a module with zef, I obtained and installed the raw.githubusercontent.com ca certificate, which permitted the fetch. | 19:31 | |
| (Oops, spurious URL at the beginning there.) | |||
| Now today, I'm getting the same problem. | |||
| Obtained the certificate, and it appears to have changed. | |||
| Does it keep doing that? | 19:32 | ||
| avuserow | raw.githubusercontent.com is signed by Let's Encrypt so it will rotate quickly | 19:33 | |
| lucs | Ah, hmm... | 19:34 | |
| avuserow | the CA cert shouldn't change that often though | ||
| lucs | I'm not sure I'm doing this correctly, but I obtain the certificate with `openssl s_client -connect raw.githubusercontent.com:443 -showcerts` | 19:36 | |
| avuserow | Let's Encrypt should be trusted by basically all OSes trust stores, for what it's worth. but I am probably missing context on why you need to verify it directly | ||
| lucs | My (limited) understanding is that once obtained, something like this is necessary (Slackware Linux): `cp the-certif.crt /usr/local/share/ca-certificates; update-ca-certificates` | 19:38 | |
| Worked last time, will try with today's certificate to see... | 19:39 | ||
| avuserow | you want to put a CA cert into your store (like from letsencrypt.org/certificates/ or equivalent), not the leaf one that's served up. then the CA cert can be used to validate any new leaf cert. (leaf cert is probably not quite the right term but the idea holds) | 19:41 | |
| lucs | I think I see what you mean. I'll aim for that. | 19:44 | |
|
20:04
librasteve_ left
|
|||
| avuserow | I'm looking for some NativeCall advice. I'm trying to send a CArray of structs to a function that's like `void doSomething(MyType *stuff)`. If I define it as `sub doSomething(CArray[MyType]) is native(...)` then it gets sent as a MyType** rather than MyType*. I'd appreciate some pointers (heh) if anyone has done this | 20:16 | |
| stackoverflow.com/questions/646842...-raku-to-c has a workaround for this situation but I was hoping there was a core-provided way since then | |||
| (otherwise I'd probably prefer to write a small C wrapper function myself since that seems safer/cleaner, at the cost of introducing a compiler step) | 20:17 | ||
|
20:42
Aedil left
22:58
johnjay left
23:00
MasterDuke joined,
johnjay joined
23:24
MasterDuke left
|
|||
| apogee_ntv | avuserow: Wrapping is the least painful option. | 23:32 | |
| Saying this as someone who writes a lot of nativecall, unless you want to abuse Buf and manually align struct fields... wrap it :D | 23:33 | ||
|
23:47
Sgeo joined
23:55
soverysour joined,
soverysour left,
soverysour joined
|
|||