🦋 Welcome to the MAIN() IRC channel of the Raku Programming Language (raku.org). This channel is logged for the purpose of keeping a history about its development | evalbot usage: 'm: say 3;' or /msg camelia m: ... | Log inspection is getting closer to beta. If you're a beginner, you can also check out the #raku-beginner channel!
Set by lizmat on 25 August 2021.
thowe How is popularity measured on raku.land? 02:09
oh, github stars? So raku modules must be on github? 02:14
japhb thowe: I doubt that's a hard requirement. It wouldn't surprise me if that was the first thing implemented though. 05:10
lizmat So I'me reading dev.to/taw/100-languages-speedrun-...sions-4p8n 10:17
and the first half is about \d not being <[0..9]>
moon-child 'In the entire history of regular expressions, I don't know if there was even one case when someone actually wanted to match Unicode digits' 10:20
moon-child wh-wh-what? 10:20
I mean ... if you want <[0..9]>, it's very easy to write <[0..9]> 10:21
MasterDuke well, it's 6 more chars to type than \d 10:25
and 8 more keypresses (on a US keyboard at least) 10:26
moon-child Sure. How many characters is the regex for what \d is equivalent to? 10:27
also: supposing you need a lot of ascii decimals. Then you can say my regex d { <[0..9]> } and then refer to <d>, which is only one more character 10:28
MasterDuke i'm not as sure as that blog post that \d matching unicode digits is going to cause a ton of bugs and security vulnerabilities, but i wouldn't have objected to \d just matching <[0..9]> either and instead having something else for all digits 10:31
Nemokosch Welcome to a world where ASCII is still highly overrepresented as a norm 😆 10:53
Really, just to name a few development-related products, gyp, Android Studio and Lazarus are broken if I just try to use them at a normal place on my Windows system 10:55
to be fair, I don't think Unix even allows you to rely on non-ASCII folders - but if Windows allows this by default with no warnings whatsoever and people keep doing this, it's more than desirable that software tools in the 2020's handle this correctly 10:56
moon-child what is 'unix'? 10:58
most unices I know of treat paths as nul-terminated strings of octets
they do not care whether those octets represent ascii characters or not
moritz_ most OSes (or their userland tools) that I have worked with do try to render path names into characters at some point 10:59
Nemokosch maybe this is a GNU/Linux thing then
Skarsnik This guy does not understand they are lot of people using no ascii char (China + India is half the mondial population?) 11:00
Nemokosch all distros I've tried to install made it clear very quickly that they want a traditional \s+ string for the username 11:02
moritz_ that must really suck, having to learn another script just to be able to pick a login name :-( 11:28
though some display managers show the full name when logging in, at least
Anton Antonov Entertaining discussions this morning... 12:10
CIAvash ASCII users can solve their problem by using basic Latin: `[<?:Block('Basic Latin')> \d]+` 12:27
Nemokosch yes, tbh the problem is using the same tools for parsing generic unicode text and IP addresses.... 13:04
and still no proper mentions of the "vulnerabilities"
moon-child ip addresses have a very specific specification, in terms of ascii. Whatever spec will say 'an ascii digit (0-9)', and if I were implementing such a specification, I would use <[0..9]> to make it clear how the grammar corresponds to the spec, even if \d were ascii-only 13:07
Nemokosch ^this 13:35
leont I remember some IP parsing libraries have had issues because they did octal in some cases… 13:41
Woodi re vulnerabilities: few days ago was discusion about loading any module via any repository... I bet such easy in overriding used modules already had few CVE in other languages... maybe it should be limited instead left until someone have idea how to do flexibility secure ? 16:20
Woodi also last month was same CVE about bidirectional source code text. I think I remember discusion #here few years ago about such posibility... do Raku code is vulnerable to such text overriding ? 16:22
japhb Woodi: That one's about slipping in evil changes that won't be seen as such in an editor or patch viewer. 16:34
Woodi yes 16:35
Woodi re ascii: until all computer stuff from cpus to operating systems and apps are USmericans made then primary ascii systems are well deserved :) 16:38
ggoebel regarding ip addresses and ascii digits... I can recall a version os OS/2 (before warp) where in the configuration for ip addresses you had to enter it in hexidecimal... 17:45
lizmat he, at least it wasn't octal :-) 17:50
japhb .o( netmask 377.377.377.0 ) 17:51
lizmat :-) 17:54
tonyo is PTS going this year? 17:59
Nemokosch what is PTS? 18:04
tonyo perl toolchain summit
Voldenet 0x7F.0x00.0x00.0x01, 0177.0000.0000.0001 and 2130706433 are all valid representations of the same ip 18:06
i wish it wasn't true
lizmat tonyo: no idea, but I guess this is the wrong channel to ask :-) 18:09
holyghost_ You couldn't play games on OS/2 :-)
lizmat tonyo: fwiw, when in person events are a thing again, I would love to organize a RTS :-) 18:10
tonyo lizmat: really bummed i missed the last one, i was in israel - so close but so far 18:10
lizmat The one in Marlow, yeah. It was eventful indeed 18:12
2020 was officially cancelled, I don't think 2021 was ever announced, nor has 2022 afaik 18:13
tonyo ahh bummer, is ovid still running those? 18:18
lizmat ovid was never running them, book was afaik? 18:20
and the last one was mostly run by Neil Bowers? 18:21
tonyo ah 18:27
guifa moon-child I actually double checked the RFC on IP addresses. unlike virtually every other RFC i've seen, there's no strict definition for the textual representation of them 19:02
(there is, however, for the binary representation)
Interestingly, the one for IPv6 exclusively uses capital letters, but in practice, I only see folks use lowercase. To me, an western Arabic 3 vs an eastern Arabic ٣ is no larger distinction than a lower f and an uppercase F in terms of text processing. 19:04
Nemokosch where is Marlow; 19:31
lizmat www.google.com/maps/place/Marlow,+...949036,14z 19:43
El_Che guifa: my programs often have a normalize function for IPs and mac addresses :) 19:51
guifa El_Che exactly. In fact, I could almost envision programs that normalize data being potentially more secure. Someone has actually taken care to think about potentially junk data and handling it 19:52
guifa emphasizes could almost envision (statement not to be taken as endorsing anything other than rigorously tested code as being considered secure lol) 19:53

Powered by Cro and the Raku® Programming Language.

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

Thank you!