00:46 schbax left 00:52 schbax joined 01:17 topnep left 01:19 topnep joined 01:36 johnjay left 01:40 johnjay joined 01:51 schbax left 01:52 hulk joined 01:53 kylese left 01:57 schbax joined 02:15 hulk left 02:16 kylese joined 02:28 schbax left 02:34 schbax joined 02:36 johnjay left, johnjay joined 02:47 johnjay left, schbax left 02:48 schbax joined 02:49 johnjay joined 02:53 schbax left 02:55 johnjay left 02:57 johnjay joined 02:58 schbax joined 03:06 lichtkind_ joined 03:08 lichtkind left 03:13 johnjay left 03:18 schbax left 03:23 topnep left 03:24 schbax joined 03:25 sibl joined 03:27 topnep joined 03:29 schbax left 03:35 schbax joined 04:08 Aedil joined 04:12 abraxxa1 left, abraxxa joined 04:20 abraxxa1 joined, abraxxa left, abraxxa1 left, abraxxa joined 04:22 abraxxa1 joined 04:37 schbax left 04:43 schbax joined 04:47 schbax left 04:52 schbax joined 05:07 abraxxa1 left 05:12 soverysour joined, soverysour left, soverysour joined 05:22 soverysour left 05:31 topnep left 05:32 topnep joined 05:33 schbax left 05:39 schbax joined 05:42 sibl left 05:45 sibl joined 05:47 schbax left 05:54 schbax joined 06:02 schbax left 06:07 schbax joined 06:09 soverysour joined 06:10 schbax left 06:14 soverysour left 06:15 schbax joined 06:25 schbax left 06:31 schbax joined 06:39 soverysour joined 06:44 soverysour left 06:45 soverysour joined, soverysour left, soverysour joined 06:54 sibl left 06:55 sibl joined 06:58 soverysour left 07:03 schbax left 07:06 sibl left 07:07 sibl joined 07:08 schbax joined 07:09 schbax left 07:15 schbax joined 07:21 schbax left, schbax joined 07:30 Sgeo left 07:59 dakkar joined 08:00 Aedil left, schbax left 08:06 schbax joined 08:11 schbax left 08:17 schbax joined 08:19 schbax left 08:24 schbax joined 08:27 schbax left 08:33 schbax joined 08:35 schbax left 08:41 schbax joined
apogee_ntv Notcurses::Native and Vips::Native now fallback to source compile automatically on old libc. Fixed intermittent UB on aarch64 due to stack allocation of variadic arguments. 09:20
patrickb: If you want an example of MacOS being much, much worse at dependency resolution than Windows: github.com/m-doughty/Vips-Native/b...s.yml#L183 09:22
This is what happens when you don't have sibling-first search paths and you try to dynamically open libs. 09:23
09:31 schbax left 09:36 schbax joined 09:42 topnep left 09:43 topnep joined 09:48 schbax left 09:54 schbax joined 09:56 schbax left 09:57 sibl left 10:02 schbax joined 10:05 _________ left 10:20 schbax left 10:26 schbax joined 10:35 soverysour joined, soverysour left, soverysour joined 10:44 librasteve_ left 10:52 soverysour left 11:03 _________ joined 11:05 schbax left 11:08 librasteve_ joined 11:10 schbax joined, schbax left, soverysour joined, soverysour left, soverysour joined 11:15 schbax joined 11:17 schbax left 11:18 soverysour left 11:23 schbax joined 11:42 soverysour joined, soverysour left, soverysour joined 11:52 soverysour left 11:57 schbax left 12:03 schbax joined 12:10 schbax left
lizmat
.oO( what would the libs think of that :-)
 12:10
13:13 annamalai left 13:14 annamalai joined 13:34 soverysour joined, soverysour left, soverysour joined
disbot2 <librasteve> thanks to all for the webhook comedy performance btw 13:35
13:40 abraxxa left
disbot2 <librasteve> ;-) 13:42
13:44 soverysour left 13:53 topnep left 13:54 topnep joined
librasteve_ o/ please can someone with chops go `fez org invite cro member librasteve` 14:19
14:33 soverysour joined 14:38 soverysour left 15:01 merp left
apogee_ntv Selkie v0.4.0 now has a bunch of graph/chart widgets built in. 15:18
disbot2 <librasteve> cool 15:22
<librasteve> post? 15:36
15:54 Aedil joined
apogee_ntv None yet 15:56
16:04 merp joined
disbot2 <antononcube> I have not heard of Notcurses before, but it seems to be of very niche interest. 16:31
<antononcube> If I can convince it to work with an AI-agent or two, I might be more interested in it. 16:32
<antononcube> My biggest problem is that all videos about Notcurses seem chaotic and incoherent. So, hopefully, the Raku related posts / videos are better. 16:33
16:35 soverysour joined 16:36 dakkar left
apogee_ntv They're going to be working at the C level, I don't know how familiar you are with C but that would make them pretty incoherent. 17:11
The goal with Selkie is to essentially have the framework deal with all of the pain and you just write a declarative TUI application.
Posted on reddit.
As for getting it to talk to LLMs I already have a WIP client using LLM::Chat & LLM::Character. 17:12
17:19 Sgeo joined 17:30 soverysour left 17:37 soverysour joined, soverysour left, soverysour joined 17:42 soverysour left
disbot2 <antononcube> The content of the videos is incoherent and unclear. I do not see what C has to do with it. 18:16
<antononcube> The videos I think/thought that might be informative are 3÷5 years old. (So, maybe too old.) 18:17
[Coke] do we have a URLencoder in the ecosystem? 18:23
I know I can roll my own but was hoping for something off the shelf.
ah, URI::Encode 18:25
I have a run() command that calls $*e 18:55
$*EXECUTABLE - if I don't pass :out, :err, it runs and the output goes to out. If I add :err, :out, then $proc.out.close(:slurp).say; (the result of run is in $proc), I get Invocant of method 'close' must be an object instance of type 18:56
'IO::Pipe', not a type object of type 'IO::Pipe'
.. but If I do a simple test on the command line with something like $*EXECUTABLE, '-e', 'say 3' ... it works. 18:57
m: my $p = run($*EXECUTABLE, "-e", "say 3", :out); $p.out.slurp(:close).say
evalable6 3
[Coke] ... any ideas why I'm getting that weird type error otherwise?
Working off the examples in docs.raku.org/type/independent-routines#sub_run , search for "raku is great" 18:58
OOF 19:05
I had "run (*$EXECUTABLE"
m: my $a = run
evalable6 (exit code 1) Cannot resolve caller run(...); n…
[Coke], Full output: gist.github.com/fba904ee9db8ab6f22...5466b60ce7
[Coke] m: my $p = run ($*EXECUTABLE, "-e", "say 3", :out); $p.out.slurp(:close).say
evalable6 (exit code 1) 3
Invocant of method 'slurp' must be an object instance of type
'IO::Handle', not a type object of type 'IO::Pipe'. Did you forget a
'.new'?
in block <unit> at /tmp/9P0GsWSO2n line 1
[Coke] there it is. LTA error 19:06
lizmat m: my $p = run($*EXECUTABLE, "-e", "say 3", :out); $p.out.slurp.say 19:19
evalable6 3
[Coke] Yup, just needed run(), not run () 19:20
lizmat the space makes the arguments be a list, in which :out becomes a Positional Pair I guess
[Coke] ... or no parens
lizmat m: my $p = run ("echo","foo",:out); $p.out.slurp.say' 19:21
foo outTrue
evalable6 (exit code 1) ===SORRY!=== Error while compilin…
lizmat, Full output: gist.github.com/910fc04e468ee9f2de...ad74996225
lizmat m: my $p = run ("echo","foo",:out); $p.out.slurp.say
evalable6 (exit code 1) foo out␉True
Invocant of method 'slurp' must be an object instance of type
'IO::Handle', not a type object of type 'IO::Pipe'. Did you forget a
'.new'?
in block <unit> at /tmp/pb3NUUlyf0 line 1
lizmat note the foo outTABTrue
without parens it's ok 19:22
m: my $p = run "echo","foo",:out; $p.out.slurp.say
evalable6 foo
lizmat I think that's a more general gotcha 19:23
at least with routines that have a slurpy 19:24
m: sub a(*@a, *%h) { dd @a, %h }; a("foo", :out)
evalable6 ["foo"]
{:out}
lizmat m: sub a(*@a, *%h) { dd @a, %h }; a ("foo", :out)
evalable6 ["foo", :out]
{}
lizmat m: sub a(*@a, *%h) { dd @a, %h }; a "foo", :out 19:25
evalable6 ["foo"]
{:out}
patrickb Which delimiters am I allowed to use for delimiting the here doc terminator? Is that list of allowed terminators showing up anywhere else in the language? 19:43
lizmat you mean heredoc or quoting in general 19:45
the heredoc terminator needs to be an identifier 19:46
afaik
patrickb I mean the thing that delimits the terminator. q:to/TERM/ <- the slash in this case 19:48
' is forbidden 19:49
lizmat right, so that's not specific to heredocs
patrickb < is ok
lizmat afaik, anything that you can do with q//
patrickb ah
got it. Thanks! 19:50
lizmat the :to is basically just an adverb that initiates some special grammar handling
but that's really it
patrickb Whoop! Got it working. Now Rainbow gets that right also. 19:57
20:20 tester123451 joined 20:21 tester123451 left 20:27 soverysour joined 20:31 soverysour left
[Coke] if I am on windows, and I have a path like "C:\\a\rel\path", and I want only "rel/path"... what is the correct way to do that? I have cheated and feel bad. 21:46
I have a var with "C:\\a" in it, so I strip off a leading C:\\a and an optional / or \, then subst all \ to / 21:47
Surely there are methods to do what I want here by converting between IO::Paths.
ugexe m: my $path = "/a/rel/path"; say $path.IO.relative("/a");' 22:10
evalable6 (exit code 1) ===SORRY!=== Error while compilin…
ugexe, Full output: gist.github.com/e7dc74388983939c67...5a168e362e
ugexe m: my $path = "/a/rel/path"; say $path.IO.relative("/a"); 22:11
evalable6 rel/path
[Coke] yup, that was a definite miss on my part. 22:14
22:26 Aedil left
apogee_ntv Just dogfooding Selkie, I built most of a TUI version of Omnifocus (gtd app) in a day, has tags/projects/time logging/repeatable tasks etc. Flagging, forecast view, quick capture. 22:40
Uses an encrypted database by default. 22:43
22:44 librasteve_ left

Powered by Cro and the Raku® Programming Language.

Please report any issues / comments / feature requests as an issue on App::Raku::Log.

Thank you!