ugexe if it is indeed related to . and .. in the path somewhere i wonder if we could update this conditional - github.com/MoarVM/MoarVM/blob/dd98.../io.c#L448 - to naive do the cleanup of those two things 01:18
.cleanup doesnt do anything with .. and .resolve doesnt do anything if part of the path doesnt exist yet (since it depends on checking if the file is a link etc i guess) but that will fail because the path is too long and wont be a valid long path in moarvm due to the .. 01:21
melezhik . 11:50
hey patrickb , maybe you're interested in test failure for Rainbow - sparky.sparrowhub.io/report/sh-pat...nbow/23033
patrickb Definitely interested. Can you maybe create a ticket at todo.sr.ht/~patrickb/Rainbow 11:53
melezhik. todo.sr.ht/~patrickb/Rainbow/9 13:26
patrickb Cool. Thank you! 14:33
tbrowder ugexe: i'm stuck with a bad zef ugrade for module PDF::Font::Loader. Unfortunately I uninstalled the previous working version and i can't seem to install the previous version even with --force-install. is there any way to force such an install? 16:31
timo what is preventing the install of the previous version from working? 16:32
in other words, what error are you getting?
tbrowder i get no installation error, but zef will not install the specific version i want (the older version). zef always installs the latest version regardless of my input. maybe i'm using the wrong syntax. i have tried: "zef install 'PDF::Font::Loader:version<0.8.10>:auth<zef:dwarring>'" (and variations) to no avail. 16:59
i mean "ver" 17:00
timo ok, so you can see in the output of the "zef install" call that it's actually grabbing a newer version? 17:01
the first line shows "Searching for: PDF::Font::Loader:ver<0.8.10>:auth<zef:dwarring>"? 17:03
tbrowder my bad BIG typo, but no warning from zef. i was using 'v<>' instead of 'ver'
timo oh, ok 17:04
tbrowder zef ignored the v part in all instances
my end is working now, but...i'll try --debug... 17:05
ok, no warning, but zef completely ignored the 'v<0.8.10>' part 17:12
lizmat could actually be a CURI issue
tbrowder and that confused me old eyes that kept missing the typo 17:13
now if i can just 'member that weirdism 17:14
ugexe you use the same syntax as if you were loading the module 17:17
you wouldnt do `use Foo:v<0.8.10>` 17:18
you would do `use Foo:ver<0.8.10>;`
zef also allows one to use :version<...> although raku does not look at e.g. `use Foo:version<0.8.0>` (to raku this is still :ver<*>) 17:19
but the easiest way to remember it is to just use the syntax you would use to actually use the module you want 17:20
timo is a check for wrong adverbs missing in zef, or in rakudo?
ugexe i dont think it is missing in rakudo because that wouldnt forward compatible or whatever 17:21
arguably zef could emit a warning, but generally it doesnt as it doesnt make sense to warn when parsing some bad name in an ecosystem when it generates its initial index or some such. so it would need to be targeted towards the parsing of user input specifically 17:22
ramman Are all PCRE regexs valid Raku regexs? 18:07
[Coke] Not intentionally? 18:13
I would certainly not expect them to mean the same things all the time
Why do you ask? 18:15
timo most pcre regexes are not valid raku regexes 18:19
er, unless it's okay that they don't mean the same thing 18:20
ramman Cool, thanks.
timo you can put a literal < in a PCRE right?
so for example, every pcre that matches raku-regex / <[a..z A..Z 0..9]>* '<' <[a..z A..Z 0..9>]* / is valid pcre but invalid raku regex 18:25
that was supposed to be 0..9]>* rather than 0..9>]* 18:26
because < in raku are a construct that requires a matching > to go along with it
but /[abcdef]/ are valid in both pcre and raku, but they match different things 18:28
ramman Okay, I understand. I was basically just asking if you need to learn a different way of doing regexs to use them in Raku. 18:30
coleman It's not 100% new, but it is different 18:31
timo there is limited backwards compatibility in raku for p5 regexes 18:32
m: say "abcd" ~~ m:P5/[abcd]/;
camelia ｢a｣
timo m: say "abcd" ~~ m/[abcd]/;
camelia ｢abcd｣
tbrowder ugexe: i know that, but i finger-fumble a lot and my eyes don't catch such errors. a warning or failure from zef would be very helpful. and i know the old saying something like: " doing the same thing over and over and expecting different results is the defintion of insanity" 18:53
librasteve ramman: not sure if you know the history, but Larry put regex into perl and it became the universal standard - thus "Perl Compatible Regular Expressions" == PCRE. But come the rediesign and Larry knew that PCRE were broken, particularly in respect to high level composition of regexen such as Grammars - and so with perl6 (now called raku) he fixed it. 18:56
this section of the perl => raku guidance docs.raku.org/language/5to6-nutshe.../_regexp_) has some tips for the level of change which is fairly low overall 18:59
ramman What didn't he like about pcre?
lizmat not being composable 19:00
prematurely huffman coded
librasteve tbrowder: take a sec ... you are asking for zef to consider all the ways you are typing ver<...> incorrectly and then to warn you ... err this is your own lookout 19:01
timo github.com/Raku/raku.org/blob/main...po/A05.pod this might help, that's from long long ago 19:02
tbrowder well, yes, but just tell me it can't handle it and why. i do that with my progs. 19:03
raku does that.
librasteve github.com/Raku/old-design-docs/bl...-regex.pod 19:04
tbrowder most good programs do that
esp. with "--debug" 19:05
timo i explicitly did not link to the synopsis but the apocalypse 19:06
tbrowder librasteve: ^^
librasteve sorry timo - I didn't realise you had beaten me to it ... ramman ... please start with timos link 19:07
tbrowder i love raku regexes, it was a joy redoing my old perl uses into raku 19:11
librasteve amen to that 19:12
timo the synopses are often no longer correct due to changes to the language that happened after they were no longer considered the "source of truth" 19:13
though i don't know what has changed in the regex one
librasteve timo: thanks for the link to the apo - it is a very good read 19:14
timo but the apo document goes through specifically what larry didn't like about PCRE and what the RFCs were that went into the design
librasteve yeah 19:15
We're frogs who are getting boiled in a pot full of single-character morphemes, and we don't notice. <== PCRE 19:27
timo i should probably point out that there are also still things in the apo document that are no longer how the regex design ended up, even when there is an "Update:" there those updates themselves are also possibly outdated 19:35
xinming releasable6: status 20:46
releasable6 xinming, Next release in ≈6 days and ≈22 hours. There are no known blockers. Changelog for this release was not started yet
xinming, Details: gist.github.com/2ca98afc000c40e424...ae85989ac5
coleman Today I learned (or re-learned, because I never use them) that "git trailers" metadata is a thing alchemists.io/articles/git_trailers 21:10
timo oh this is the first time i've heard of branch descriptions 21:17
antononcube They remind me of conventional commits: www.conventionalcommits.org/en/v1.0.0/ 21:52

Powered by Cro and the Raku® Programming Language.

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

Thank you!