00:11 Sgeo joined 00:19 japhb left 00:25 japhb joined 01:03 johnjay joined 02:39 kylese left 02:40 hulk joined 02:47 dano2 joined 02:49 dano left, dano2 is now known as dano 02:54 sibl joined 02:58 dano2 joined 02:59 sibl left, dano left, dano2 is now known as dano 03:03 thaewrap- left, thaewrapt joined, thaewrapt left, thaewrapt joined 03:15 hulk left, kylese joined 03:22 arunvickram joined
arunvickram hey, it's arun, i'm joining on my irc account 03:23
03:26 arunvickram left, arunvickram joined 03:29 arunvickram left 03:30 arunvickram joined 03:46 arkiuat left 03:48 arkiuat joined 03:57 arkiuat left, arunvickram left 03:58 arunvickram joined, arkiuat joined 04:01 sibl joined 04:03 arkiuat left 04:06 sibl left, arkiuat joined 04:07 sibl joined 04:11 arkiuat left
simon_sibl I am wondering, whats the reason behind Raku dropping the my @arr = (0,1,2); $arr[0] = 5; @arr[1,2] = (6,7); instead now its my @arr = (0,1,2); @arr[0] = 5; @arr[1,2] = (6,7); the sigils doesnt give as much detail as before 04:12
04:20 arunvickram left 04:21 arkiuat joined 04:27 arkiuat left 04:35 arkiuat joined 04:42 arunvickram joined 04:45 arkiuat left 04:52 arkiuat joined 04:54 arunvickram left 04:58 Aedil joined 05:48 kjp left 05:49 kjp joined 05:50 mc2 left, mc2 joined 05:51 kjp left, kjp joined 06:08 dpk left, dpk joined 06:50 abraxxa joined 06:57 abraxxa left 06:58 abraxxa joined 07:10 ilogger2_ left 07:11 ilogger2 joined 07:54 atcroft left 08:02 Guest11 joined 08:08 arkiuat left 08:21 sibl left, sibl joined 08:25 Sgeo left 08:26 zjmarlow joined 08:32 Guest42 joined 08:33 Guest42 left 08:38 arkiuat joined 08:42 arkiuat left 08:51 Guest11 left 09:10 arkiuat joined 09:15 arkiuat left 09:29 dakkar joined 09:44 arkiuat joined 09:53 arkiuat left 10:00 sibl left 10:05 arkiuat joined 10:06 librasteve_ left 10:10 arkiuat left 10:22 bobv joined, bobv left 10:30 librasteve_ joined
nemokosch it added confusion, especially since there was no guarantee that just because arr is an array, arr[0] isn't also an array 10:31
librasteve @simon_sibl one of the key complaints about the complexity of perl was that you need to "decline" the sigils based on what you wanted back eg print @hash{a,b}; - raku therefore "simplified" so that the sigil denotes the type of the declared variable storage, not the expected return type - according to Chatjippity this makes it <<Lexically consistent, Less context magic, More predictable, Easier to reason about statically - 10:39
Raku intentionally reduces context-dependent behavior.>>
10:40 arkiuat joined 10:42 sjn left, Nemokosch joined 10:43 sjn joined
simon_sibl For Perl it’s easy It can’t be an array It would be a ref 10:44
nemokosch not sure if I understand but I don't even want to, honestly 10:45
simon_sibl It definitely is “simpler” but having to decline the sigils clearly showed that value we would get from it But for Raku I understand the types are somewhat different with the whole container thing
10:45 arkiuat left
nemokosch Perl does a lot of things that no other language does and it usually turns out they don't do it for a good reason 10:45
simon_sibl I would argue that other languages tend to all look similar to satisfy the need to be quickly learned and productive for big corporation, which sucks Raku and Perl aren’t like other languages, and I believe that’s for the best 10:46
nemokosch there are other big languages, including arguably corporatist languages like C# or C++ to an extent 10:47
as far as I remember, it's much worse than "it shows what value you would get" 10:48
it outright changes what value you get
in Raku there isn't really this whole "distant context" thing 10:49
simon_sibl If you ask for something else, it gives you something else but the principle is simple (maybe there are edge case I don’t have in mind?) 10:50
> perl sigils make more sense than just "being there". I knew PHP before Perl and it seemed like there is no reason to have $ sigil there other than "it is what it is". In Perl, they actually have an important function. Raku seems to strip that function away, at least to some degree 10:51
10:51 arkiuat joined
copyright bbrtj o7 xD 10:52
10:53 tobs left, tobs joined
nemokosch referring to bbrtj doesn't necessarily make this case stronger 😅 10:54
at least he supports Castle Game Engine
honestly, I have spent a long time poking Raku, and much of the absolute craziest parts (dependencies are "modules" and not what you actually release; next and last propagate through function calls; that there is one properly working sigil with variable semantics: you guessed it, the one in PHP) come straight from Perl 11:01
I wouldn't want to go back to Perl merely to justify my criticism specifically in the context of Perl
even "it cannot be an array, it will be a ref" is an odd flex that rather describes what actually happens than what should 11:04
you know what else comes straight from Perl? This insiderism, this "us against the world" mindset 11:06
11:13 arkiuat left 11:15 Nemokosch left 11:27 arkiuat joined
bbrtj This was not actually a thing originally. Perl used to be a magpie language, stealing the best parts from other langs. I believe the severe attacks on it are what caused the community to become defensive 11:29
but are we particularly defensive though? I once went to Elixir discord and simply asked if they have any benchmarks to show me. They spent 3 hours trying to convince me that I don't really want benchmarks 11:30
Voldenet well, sigil is now part of the variable, so $arr @arr and %arr are separate, I think it's just easier to explain
11:31 arkiuat left
nemokosch probably it's not the only defensive place but I would say it is particular regardless 11:33
Voldenet in fact, raku does not strictly need sigils at all
m: my \thing = {$}(); thing = 3; say thing; thing++; say thing
camelia 3
4
Voldenet @ and % are just useful helpers for extremely common task
nemokosch it's not even just the user-blaming aspect 11:34
it's that any problem must be external, the whole world is some corporatist conspiration
this "thing" is definitely not going to be thread-safe 😛 11:36
Voldenet hm?
it's as thread safe as regular binding 11:37
simon_sibl In Perl they would be different variable too iirc
nemokosch it binds to a static variable
no they wouldn't, that much is for sure
sigils are not a part of the variable name 11:38
in Raku they are, you can have @foo and %foo as two different variables in the same scope
simon_sibl and in Perl as well 11:40
glot.io/snippets/hfp6wbof5p
11:40 sibl joined
antononcube Can we ban comparisons with Perl and inspirations by Python here?! That will remove the possibilities of bringing a lot of backwards, inconsistent, and incomplete ideas to Raku. 11:41
Might make the discussions, more boring, though…
Voldenet Yes, they are separate, but `$$x[0]` is really unpredictable
nemokosch I'm not sure how this works but this is a recurring topic that Perl doesn't have separate symbols for these 11:42
Voldenet > my $x; $x = 1; $$x[0] = 2; dd $x; # 1
nemokosch the name of the variable is just foo
Voldenet > my $x; $$x[0] = 2; dd $x; # [2]
simon_sibl that wouldnt run with use strict 11:44
or it does
hmm
Voldenet yes it wouldn't 11:45
simon_sibl for the first example yeah 11:46
nemokosch in Raku, $ is idempotent (probably)
Voldenet well, it depends
my $x; sub a { $x = 1; } sub b { $$x[0] = 2; }; a; b; dd $x # this would run
nemokosch there really are too many things that "depend" 11:47
lizmat weekly: dev.to/lizmat/positional-methods-439i
notable6 lizmat, Noted! (weekly)
nemokosch it doesn't run with Rakudo v2025.12
Voldenet perl is very good language for "write once never revisit again" code :P
nemokosch not only because of the missing semicolon
Voldenet (it's perl)
`$$` is a subtle hint ;> 11:48
nemokosch the previous snippet wasn't Perl
simon_sibl oh cmon, its possible to write readable and maintainable Perl code
nemokosch that's a common cope
Voldenet It is possible to write readable and maintainable asm code 11:49
the question is whether it is practical and easy
nemokosch it's just as possible as write any userspace application in C
Voldenet or rather how hard it is to make a deadly mistake
nemokosch "possible" is not enough
simon_sibl lot of them are 11:50
the most criticals one are
Voldenet many people don't know that "C" is actually an abbreviation for "CVE"
nemokosch the second is an overstatement
simon_sibl omg, such bait xD
nemokosch and nothing follows from the first
simon_sibl but I will steal that one from you xD CVE xD
antononcube Yes, it has to also look exclusive, “magical”, and something to be smug with or about. 11:51
11:51 arkiuat joined
nemokosch nah, said no one ever 11:51
Voldenet perl is a language where it's not hard to make things neat and tidy, but the language doesn't help you with that at all 11:54
bbrtj yeah, pretty much 😄
antononcube @Voldenet No, “C” is the first letter or “CPU”. 11:55
11:56 arkiuat left
This might be correct and insightful, but I don’t think anyone outside of Perl’s relm would associate Perl with “neat and tidy”. 11:58
simon_sibl people cant read code if thats not Python or Javascript anymore xD 11:59
11:59 tadzik left, tadzik joined
Voldenet heh, I can't read code if it's python, my eyes start bleeding from verbosity ;) 11:59
nemokosch you have not seen a lot of languages, then 12:02
if Python is verbose then we have to eliminate 80% of programming languages instantly
simon_sibl I saw a list comprehension on 3-4 lines few days ago, I lost faith /o\ 12:03
Voldenet well, python is actually more verbose than C, java and C#
it's actually insane how limited it is 12:04
nemokosch it isn't
this is a broken claim beyond repair
simon_sibl any counter arguments ?
nemokosch that would need an argument for starters 12:05
not just a hot take
Voldenet macros in C make you able to define two statement lambdas… :P
python is the single language that's unable to do that
I have no idea if I've seen anything with similar limitation
nemokosch that doesn't seem very relevant though
I have seen Pascal and Ada, for what it's worth 12:06
but anyway, that's a deliberate limitation as something that you shouldn't need too much
Voldenet python has a lot of these deliberate limitations 12:07
antononcube @Voldenet I was like that for long time. At some point, because some (younger) people kept saying about me “he has to upgrade his skills” — which meant “do data science with Python” — that I am a much better Python programmer than them.
nemokosch well, you have said one
bbrtj Pascal is far more readable than Python. At least it disallows unreadable neat tricks which pythoneers do all the time
nemokosch I think a lot of slander of Python comes from mere jealousy 12:08
I mean I can relate, that's how I feel about Go
it's a braindead language that doesn't really offer anything
Voldenet very weird formatting limitations are a huge part of what makes python annoying
nemokosch and yet people use it and praise it
antononcube I “grew up” with Pascal.
simon_sibl Go and Python serve the same purpose 12:09
being able to hire and fire people as much as you can as long as the code works
nemokosch Python is a fairly sensible language as opposed to Go
it's not deliberately dumb
Voldenet I won't start going on about coroutines in python, they're not even implemented
antononcube Yes, Python is very “spread out.” Unfortunately, for most LLM and AI things is the language to use, because of features and speed. 12:10
Voldenet python has no goto
nemokosch master baiter
Voldenet I know :>
simon_sibl feature and speed provided by... 12:11
C
nemokosch you should start a morning show on Twitch
holmdunc Most of this sounds like raging egotistical crap about strawman users
nemokosch like the moustache guy and friends
antononcube I think Python is dump by design. Guido van Rosum wanted Python to be a “second programming language” to many programmers . 12:12
nemokosch it's relatively simple but I don't think that equals dumb
the language has something to offer, over "just copy and paste more code" 12:13
bbrtj Python makes it harder to use the most important factor of code readability, the vertical whitespace. By using it it's easier to lose track of indentation, which in turn affects the code execution path. That goes against their claim that the language helps readability
antononcube I would say — at this point — Python is “deceptively simplistic.” It is a fairly complicate, old language.
nemokosch okay, the obsession with syntax in the Perl world is understandable
but it's still so unproductive 12:14
lo and behold, your forever favorite language: Ruby
Voldenet Ruby is fine
simon_sibl xD
Voldenet ;>
nemokosch that's your place if you are obsessed with syntax
the redemption arrived some 30 years ago
simon_sibl Raku is my Ruby with Perl falvor, I love it
nemokosch you won't have to touch Python ever 12:15
their cult even made it to Erlang and created Elixir
they made it into JS and created Coffeescript (it collapsed after ES6 though)
simon_sibl some of my colleagues use Ruby daily
as sysadmin
nemokosch they also created Groovy and Crystal 12:16
simon_sibl but nothing beat Perl for script you can send to somebody
and I dont think anything will reach Perl level
nemokosch forever sweet dreams in Ruby land
antononcube @nemokosch Perl combines the readability of line noise with maintainability of write-only code. Perl is just executable punctuation.
nemokosch for the rest of the world, syntax is not a primary concern
Voldenet "you hate python because it's popular, also syntax is not important" 12:17
nemokosch even with shell (short for syntax hell), the main problem is actually the semantics
Voldenet Okay then
antononcube To rephrase : Perl offers the power of a scripting language with the elegance of a cat walking across your keyboard. 12:18
Voldenet I dislike python because when it makes choices, it always chooses simplicity over elegance 12:19
antononcube It is easy to hate Python because of the people who use it.
simon_sibl well at least perl wont complain about space or tab being different for indentation when copy pasting a script from someone 12:20
bbrtj Perl will rarely complain about anything to be honest 😄
simon_sibl perl is like a good friend, never complains
always tries its best
12:21 arkiuat joined
antononcube Many Python programmers — who turn out to be just Python users — have told how great Python is while they do not know any other programming language. 12:21
simon_sibl I am a sysadmin, but I enjoy asking my CS colleagues (graduated from high rank university ?!?) about basic pointer or data structure or memory handling and they are always unable to answer 12:22
python braindead them
antononcube @simon_sibl Perl combines the flexibility of regular expressions with the clarity of regular expressions. (Hence, its complaining qualities.)
simon_sibl regex are amazing 12:23
nemokosch I'm surprised that somebody hasn't learned for years spent with Perl and Raku that elegance does not exist
it's not a real thing
it's like the easter bunny or the holy grail
Voldenet due to its simplicity, python has performance problems that aren't easy to solve – coroutines and enumerators use exceptions for example
simon_sibl it does, I spend time aligning my = with multiple ligne variable assignements !!
Voldenet you can't simply "suspend" a coroutine, it's not a thing
nemokosch I wouldn't be sure about that 12:24
simon_sibl does the bridge support image or I need to send a link ?
nemokosch but anyway, it's perhaps the biggest design mistake of Python that iterators are backed by exceptions
antononcube @simon_sibl If you post an image the IRC people would get a link. 12:25
Or, so, I have been told.
simon_sibl Few days after I discover regex, how readable it is !
cdn.discordapp.com/attachments/633...892b8&
Voldenet exceptions are one of the "simplicity vs elegance" choices I dislike 12:26
12:26 arkiuat left
nemokosch exceptions in themselves wouldn't be such a choice 12:26
simon_sibl I am on the error as value team 12:27
nemokosch and for the proliferation of exceptions for no real tangible reason, I simply consider that a mistake
not really a big philosophical choice
12:27 bbrtj joined
I have never seen a rationale and I see no reason to read much into it 12:28
Voldenet well, it is very easy to implement that way: `try { while True: body(it.__next()) } except StopIteration { }` and that's it
pure simplicity
nemokosch it would be very easy to implement with hasNext as well - in fact arguably easier
it would even be easy to implement with a sentinel value 12:29
antononcube @Voldenet Do you see a link for Simon’s image?
simon_sibl I saw from my IRC client
link to discord image
nemokosch does it open for somebody not logged in?
Voldenet Yes, I see the link :D
sibl 07
tellable6 2025-11-28T14:09:46Z #raku <SmokeMachine> sibl: maybe you want a proto?
Voldenet I see nothing bad, it's perfectly readable
simon_sibl what is this: tellable6 12:30
nemokosch a bot
simon_sibl why does it randomly show a message like this xD
nemokosch because somebody sent you a message on IRC a long time ago
simon_sibl I know right ?
nemokosch I sorely miss the whitespaces from the regex 12:31
Voldenet well, the "hasNext, moveNext" seem to be different in most languages despite looking simple
nemokosch pretty sure discord responses aren't bridged in a way that would be easy to follow
in general it's best to assume everything you send here will show as plain text 12:32
antononcube Hm… it is a good idea to see links shown in the IRC logs.
nemokosch reactions also cannot be bridged, of course
antononcube > Perl offers the power of a scripting language with the elegance of a cat walking across your keyboard.
SmokeMachine sibl: I’m sorry… I have no idea what I was talking about… :( 12:34
Voldenet people will hate on perl being able to process a log in one line
sibl it rings a tiny bell in my mind, but I am sure that whatever that was, I no longer have the issue xD
Voldenet while in other languages it'd take 10 lines to do the same things
nemokosch I mean, Perl is an improvement over two or three languages 12:35
Bash, awk, and sed if you want to consider it
Voldenet in fact it's easy to make the above 1-liner readable just by adding whitespace to it
nemokosch but as a general purpose programming language, it's kinda half-assed
Perl is its own second system 12:36
we'd have to say Raku is already a third system
bbrtj Perl is a VM, CPAN is the language
or so they say 12:37
antononcube I was fairly impressed by the Raku one-liner of doing factorization of Gaussian integers.
Voldenet I still pick perl over raku for simple text processing, not only because of speed, but because perl is less formal
nemokosch to be quite fair, now that nobody is really interested in Perl, they actually modernized the language quite a bit
sibl I was surprised that with pack/unpack in Perl I was able to reach the same speed of a Python script using a very optimized C module 12:38
the code was obviously a bit less readable and more C level like
nemokosch if Raku were fast, the underlying code could be asm for all I care 12:39
Voldenet i hear that perl has added `given/when` in perl then deprecated it
antononcube codegolf.stackexchange.com/a/185311
nemokosch performance is a high-level non-functional feature
Voldenet being slow is a feature of a language, because it lets you make coffee when it compiles 12:40
nemokosch chainsmoker logic
bbrtj Voldenet: yes smartmatch/given/when is on the way out
Voldenet You can't smoke inside and outside you get fresh air while you smoke, therefore it's healthy for lungs 12:41
this is not an investment advice
nemokosch XDD
bbrtj a pack a day keeps cancer away
nemokosch cdn.discordapp.com/attachments/633...53e6c& 12:42
cdn.discordapp.com/attachments/633...0967e&
interesting
that image was familiar somehow 😆
12:42 arkiuat joined
antononcube - If compilation is slow enough for coffee, debugging will be slow enough for retirement. - Which brings me to one of my favorite statements / observations that the average age of Raku programmers is 58. 12:44
simon_sibl I am 23, people say its a 100 years language, hopefully I will retire with Raku 🦋 12:45
lizmat simon_sibl++
antononcube I do not believe you. 12:46
Voldenet Well, if you're 23, then it only means that some 93 years old programmer averages this out
antononcube • @Voldenet I prefer languages that let me make coffee because I want to — not because I have to.
nemokosch how humorless of you
12:47 abraxxa left
simon_sibl must been really hard in the past, programming and having to wait the next day to see the result 12:47
12:47 arkiuat left
I heard Rust programming language try to mimic this for nostalgia 12:47
antononcube This is how I made my PhD. (Using supercomputers.) 12:48
But, also, this is how I started programming with punchcards — you wait for the next day.
Forced us to program really well, at least syntax-wise. 12:49
simon_sibl I wonder what would happen if one would time traver and bring Raku to Turing How long would it take him to learn the language and break enigma
12:49 abraxxa joined
antononcube It wasn’t Turing who broke Enigma. 12:50
sibl I wanted to send s/traver/travel but now Discord understand regeex
regex*
antononcube But, probably, he would stick to his guns.
simon_sibl wasnt it ?
12:50 arunvickram joined, arunvickram left
Voldenet meh, it'd only work if you took modern computer with you 12:50
and modern computer would be bigger gift than any language 12:51
nemokosch Discord has this feature for quite long I think
simon_sibl I would bring my phone and they would be blown away
Voldenet literally, by lithium battery?
simon_sibl I dont have a Raku interpreter on it tho
12:51 arunvickram joined
lmao 12:52
sibl some of you went to clown university it seems, no way average brain generate those jokes 12:53
antononcube It was the Polish — both the idea of breaking it and the “breakthrough.”
12:54 abraxxa left
sibl I thought Turing took inspiration from them ? 12:54
like they had a kind of machine but not complete or something
12:56 abraxxa joined
sibl btw I was trying to find for a while, any idea where the recording for Raku/Perl from FOSDEM 2026 is ? 12:57
lizmat there was no dev room ? 12:59
or do you mean the BoF meeting? 13:00
simon_sibl yeah this one: fosdem.org/2026/schedule/event/ZRG...-cpan-bof/
not sure if there is anything interesting there, but any Perl/Raku content is good to take xD
nemokosch so now Raku.land is a community 13:01
that's a bit odd
lizmat sjn ^^ any idea what's going on `/
?
.oO( note to self: don't type when you're trying to eat baklava )
 13:02
sibl google what is baklava 13:03
nemokosch XDD
lizmat en.wikipedia.org/wiki/Baklava
nemokosch it's this ultra sweet Turkish strudel
sibl looks nice ! 13:04
nemokosch it has walnut or something like that
13:05 arkiuat joined
lizmat pistachio 13:05
nemokosch I still feel resentment because of the raku.land "incident" 13:10
13:10 arkiuat left 13:11 sibl left
I basically got accused of insinuation and time has proven my concerns 13:13
SmokeMachine lizmat: I’m curious, are you also going to talk about custom modules that also use the CAPS methods “notation’? 13:19
lizmat didn't really plan on that: looking at 100+ uppercasey things in Raku right now... 13:20
but perhaps... what did you have in mind?
13:23 arkiuat joined 13:26 _________ left 13:27 _________ joined 13:28 arkiuat left
simon_sibl What happened 13:31
nemokosch github.com/Raku/problem-solving/pu...1807187587 13:37
lizmat
.oO( sometimes I wish people's egos were smaller )
 13:39
nemokosch I made this point - all previously expressed - that the old modules.raku.org site did have some redeeming qualities over raku.land and that perhaps it wouldn't be a good idea to give a site this important from the organization to volunteers 13:40
13:41 arkiuat joined
then for unrelated reasons modules.raku.org died and got redirected to raku.land 13:41
lizmat we're all volunteers... the raku.land's code is open source
nemokosch that's not what I'm talking about
visibly nobody is interested in contributing to raku.land, and nobody can take it over, either 13:42
that's exactly what I had in mind back then, and it was made out to be "you exclude us from the community" and all that
lizmat and why is it that nobody can take it over? because of the domain?
nemokosch because everybody took it as "it's their project" so nobody has either the knowledge, nor any sort of legitimation 13:43
it's not me who has to answer these questions, I didn't make these decisions 13:44
lizmat this applies to many aspects of many open source projects 13:45
nemokosch and I also didn't insinuate anyone
lizmat ok, it appears you want to have the final word on this: so you shall
13:46 arkiuat left
nemokosch if it's egoistic that I feel resentment for getting called out, and then time proving my concerns exactly, then I'll have to take it 13:47
antononcube I agree with lizmat that Open Sources projects are self-selecting and/or self-electing. But Nemokosch was making the valid point that a big and important for the community project should not be like a house of cards — you pull a person or two out, and the whole thing becomes collapses. (Metaphorically at first, then actually.) 13:53
That is just common software engineering sense. I was surprised that people had hard time to see that. 13:54
lizmat but that same software engineering sense is applicable to *all* of Raku's infrastructure
that fact that raku.land has a bus factor of 2 is way above average!
most of the raku community modules have an effective bus factor of 0 13:55
simon_sibl Zef is decentralized tho right ? Can point to a git repo and that would work ? Also why it’s not possible to use CPAN (since we have Inline Perl and they have the same for Raku) ? Or to make a very similar alternative for Raku ? 13:56
antononcube @lizmat Agreed. That is why I made the qualification “big and important for the community.”
lizmat so the point made about raku.land could be made general as well.. and as such, the two people of raku.land felt targeted when that was made an issue 13:57
and they were right to feel targeted 13:58
antononcube Ok. It was long time since I read the discussion, and I might have missed some nuances. (“Nuances” for me, but more important to others.) 14:00
nemokosch and I wasn't right that I suddenly got attacked for something I openly held and explained for long?
does 1 feel and 1 real not make at least as much as 2 feels? 14:01
the only thing that has changed in this time is that now it's plain to see that there was a legitimate concern 14:03
14:10 arkiuat joined
lizmat My analysis: 14:11
1. 2+ years ago someone was worried about the raku.land project being run by only 2 people
2. these people indicated that they were real raku community people willing to put in work and money into maintaining the raku.land site 14:12
3. that same someone was upset and decides to bring it up 2+ years later because they felt wronged 14:13
my conclusion: get over it, there are more important things to work on than discussing this
14:14 bobv joined
antononcube I agree with the last point. But I also re-read big part of that GitHub discussion. 🙂 14:14
nemokosch I get it, you don't mind because you weren't offended and had your sympathy elsewhere anyway
antononcube So, after re-reading the comments in that GitHub discussion Nemokosh points still seems valid to me. I am not sure about why the raku.land creators feel “targeted.” Seems to me to be an immature reaction of people who seem to be mature age- and career-wise. 14:15
nemokosch but to me, it was like, there were valid concerns about raku.land, including but not limited to ownership 14:16
I was transparent about them and I added a very slight and in my book fairly diplomatic note in a good-faith presentation
14:16 arkiuat left
and then these ever-so-humble people lash out at me, apropos the shutdown of modules.raku.org, even 14:17
2 years later, the situation of raku.land haven't improved a slight bit, and we are used to not having anything else 14:18
lizmat this discussion was restarted because of your statement: "14:10:04] <disbot13> <nemokosch> I still feel resentment because of the raku.land "incident"" 14:19
[14:13:13] <disbot13> <nemokosch> I basically got accused of insinuation and time has proven my concerns
nemokosch yes, I do feel a kind of bitterness about why all these accusations were necessary
lizmat so this is not about the situation re raku.land, but it is about your ego 14:20
14:20 [Coke] left
nemokosch how can you separate them? 14:20
lizmat so please, let's stop this discussion: your worry has been noted
nemokosch drawing this false dichotomy. The outcome ("situation") of raku.land, among other things, reminds me that I'm expected to suck it up, as apparent 14:22
lizmat that would be helpful, indeed 14:23
nemokosch and if this is all you wanted to add, you really didn't have to, and it would have been easier for everybody
talking about diplomacy
14:23 arunvickram left
it also didn't help the case that one side gets "they felt targeted and they were right to feel that way" while the other side gets "get over it, ego issue" 14:32
patrickb For what it's worth I plan to spend some time on raku.land in the hopefully not so distant future. 14:39
14:44 arkiuat joined 14:49 arkiuat left 14:52 arunvickram joined 14:53 atcroft joined 14:57 Guest11 joined 14:58 Guest11 left 15:00 arkiuat joined, arunvickram left 15:01 Guest11 joined 15:03 Guest11 left 15:06 bobv left 15:07 arkiuat left, arkiuat joined 15:14 arkiuat left 15:22 arkiuat joined 15:25 simcop2387 left
nemokosch you are kind of an unsung hero with the obscure challenges you pushed forward 15:26
15:27 simcop2387 joined 15:32 arkiuat left 15:36 arkiuat joined
librasteve I do not want to share this space with egos who rake up arguments from years ago, I do not think that we as a small community have the time to waste on policing such discussions instead of working on solutions, please don't do it again 15:39
nemokosch I don't think you can just ask that 15:40
I mean you can but not with the expectation that it will be followed 15:41
15:41 arkiuat left
and for this ego this, ego that: again, I get it, you haven't been basically singled out, it doesn't matter to you 15:46
but at least consider how it would feel if you were singled out, turned out you were right, and then people basically shut you down saying that they don't like to see "these arguments" 15:47
if you don't care about it then don't engage but more lectures and attacks of character are the last thing I need about this thing 15:49
15:56 arkiuat joined
bbrtj so much drama XD 16:15
16:20 sftp joined, sftp left, sftp joined
librasteve 🦆 16:21
16:27 Guest11 joined 16:28 Guest11 is now known as quiddicit 16:29 quiddicit left, quiddicit joined 16:30 quiddicit is now known as guestll 16:32 arkiuat left 16:34 guestll left 16:41 sftp left 16:44 arkiuat joined
lizmat nemokosch if you really want modules.raku.org be up again: why don't you pull github.com/Raku/modules.raku.org, set up a server somewhere and then request a DNS change ? 16:45
16:48 arkiuat left 16:52 arkiuat joined
lizmat or create Pull Requests at gitlab.com/raku-land/raku-land to get the features that you're missing implemented there 16:56
from: drewdevault.com/2021/06/14/Provide...ranty.html
"FOSS is what you make of it. You have the right to make the changes you need from the software yourself, and you are the only person that you can reliably expect to do it. You aren’t entitled to the maintainer’s time, but you are, per the open source definition and free software definition, entitled to change the software, distribute your changes to others, and to sell the software with or without those changes."
nemokosch perhaps I don't "really" want modules.raku.org be up again - from what I understood, it broke at some point and nobody found any point of resurrecting it 16:59
not all work is "well spent". CIAvash made a whole new raku.org site, I made a new stack for the old raku.org site, none of that was ever useful to anybody 17:00
now both the old modules site and raku.land itself are way more complex
lizmat ok, than let's just please stop this discussion 17:01
please 17:02
pretty please
nemokosch and for raku.land: raku.land/zef:dwarring/Pod::To::PD...es?v=0.1.7
almost sure this is a relatively simple escaping issue
lizmat then it should be simple to make a PR for it 17:03
nemokosch it would be quite weird to ask me of all people to clean after them
lizmat "FOSS is what you make of it. You have the right to make the changes you need from the software yourself, and you are the only person that you can reliably expect to do it. You aren’t entitled to the maintainer’s time, but you are, per the open source definition and free software definition, entitled to change the software, distribute your changes to others, and to sell the software with or without those changes."
and again:
Voldenet huh, deja vu
lizmat "It is nice when a maintainer offers you their time, but by no means are they required to." 17:04
nemokosch anyway, I also have other things to do, and other things than to have this non-argument
lizmat thank you 17:05
nemokosch it's not my biggest hobby to get these personal kind of remarks, either 17:07
17:11 obarb is now known as brabo 17:45 dakkar left 17:50 ds7832 joined 17:51 ds7832 left, ds7832 joined
SmokeMachine lizmat: I was just remembering my own projects that has all caps methods, and I thought it would be cool if you described a suggestion to module authors to use that. 18:08
lizmat sure... it would help if there's a page somewhere describing them :-) 18:09
nemokosch only infix operators have associativity, right? 18:19
I found this: github.com/rakudo/rakudo/issues/4779 18:23
Daniel made a good point 18:24
--$x++ could have multiple meanings with different associativities
korvo Reading through the scrollback, I'm kind of surprised how few engineering issues are considered at the PLDI/PLT level of Raku. Python isn't chosen for aesthetics but for getting things done. Case in point: last week I ported a Raku script to Python 2 so that I could statically compile and link it. 18:26
18:26 librasteve_ left
nemokosch mind if I ask why Python 2? 18:27
18:27 abraxxa left
korvo Because Python 2 has a really nice toolchain for generating native code, JITs, etc. rpython.readthedocs.io/en/latest/ 18:28
18:29 notna joined
nemokosch crazy, they never made it out of Python 2? 18:29
korvo I took notes if you want more details. Prior to considering Python, I also considered Ada, OCaml, and Rust. This really was about static linking and generating an initramfs-compatible executable, not about anything in Raku's design proper.
nemokosch it's interesting for sure 18:30
I have been using Ada as my main dayjob language for the last 2 years (probably about to end)
korvo Python 3 isn't worth the engineering effort for us, no. There is an implementation of Python 3 in RPython, PyPy, but the guts of PyPy and RPython are mostly distinct.
Oh, I actually have to paste the link, whoops. gist.github.com/MostAwesomeDude/9a...128d221cbc 18:31
nemokosch oh you are the advent calendar guy 18:32
korvo Yeah! This is a continuation of that work. 18:33
SmokeMachine `"a:b:c".split: ":", 2` returns ["a", "b:c"]... would it make sense if passing -2 instead of 2 (`"a:b:c".split: ":", 2`) it returned ["a:b", "c"]?
18:33 ds7832 left
nemokosch what if you do *-2 ? 18:34
one never knows with Raku
lizmat m: "a:b:c".split: ":", *-2
camelia Cannot resolve caller Real(WhateverCode:D: ); none of these signatures matches:
(Mu:U \v:: *%_)
in block <unit> at <tmp> line 1
lizmat m: dd "a:b:c".split: ":", *-2
camelia Cannot resolve caller Real(WhateverCode:D: ); none of these signatures matches:
(Mu:U \v:: *%_)
in block <unit> at <tmp> line 1
nemokosch wait 18:35
2 here is an amount, right?
not an offset
1 would be "split it into one piece", 2 "split it into 2 pieces", 3 "split it into 3 pieces"
lizmat m: dd "a:b:c".split(":", 2, :end) # meh 18:36
camelia ("a", "b:c").Seq
SmokeMachine yes... another thing it would make sense is to add a :invert, or something like that (edit: I like :end) 18:37
nemokosch I also like :end fwiw
korvo: > Go is famously not even as nice as Algol 68, which itself doesn't have good-enough libraries
lizmat similar to :end in .first
nemokosch I love this! 🤣 18:38
it's not really an engineering consideration but let it slip
korvo Oh yeah, my personal comfort was much more important than any kind of objective engineering. There's not a team of people that will help me maintain this. There's just my personal time which I don't want to waste. 18:39
nemokosch have you ever considered Swift? or is that also on some sort of index? 18:41
korvo Swift's got a couple really nasty design bugs and their toolchain kind of sucks AIUI. Infamously, danielchasehooper.com/posts/why-swift-is-slow/ 18:43
Same with Mojo, except Mojo's not even FLOSS. After Google v. Oracle, during which I was working at Google, I've realized that using non-open languages is unacceptable. Doubly so in the initramfs of a Linux distro.
nemokosch okay, this is something tangible 18:44
it came to my mind because they have a pretty good linking story 18:45
18:45 vasko4535586 left
korvo Another good example of that sort of thing is Julia. I don't think there's a single good summarizing blogpost, but Julia's type system leads to lots of easy mistakes, notably mixing 0-based and 1-based indices for lots of fencepost errors. 18:47
nemokosch I don't really know Julia but the complaints I have seen were... too familiar
a big wild west language with an ecosystem just like that 18:48
18:55 librasteve_ joined 19:10 abraxxa joined 19:12 abraxxa left 19:13 abraxxa joined 19:35 notna left
Voldenet may I offer to you a solution to all your problems: "rewrite it in rust" ;) 19:35
korvo Voldenet: Two highlights from my notes. First, when I switched from Rust to RPython, it took me one afternoon to replicate the previous two days' work; Rust just has that much ceremony. Also, I had such cryptic rustc errors that an upstream dev opened a usability bug for one of them: github.com/rust-lang/rust/issues/152064 19:38
Voldenet I was just replaying the old meme – jokes aside when encountering a fork in the road, you can't ever go both ways – rust is strict and very mean about it :\ 19:44
librasteve I have burnt some cycles on Rust (see Dan::Polars if you are interested) - got tired of typing safe() 19:45
Voldenet and saying "x is better than y" is always wrong
19:57 vasko4535586 joined
nemokosch right is always better than wrong 20:08
... except for the GPU
Voldenet sometimes left is right and right is wrong 20:14
nemokosch when nothing goes right, you have to go left 20:16
but when nothing is left, you have to turn right
Voldenet when nothing is left you can't go right, because right is also not left 20:18
nemokosch when nothing is left then in fact everything is right 20:19
Voldenet yes, but also nothing is left obviously so nothing is right either 20:20
nemokosch that's all wrong 20:21
Voldenet yes, that's not right
nemokosch alright 20:23
20:47 simcop2387 left 20:53 simcop2387 joined
antononcube Julia is like Obama -- got an award for just showing up. 21:04
I "Julia" is translation target of some my DSLs, but it is last one in the priority list and I do it only for illustration purposes. 21:05
21:23 arkiuat left 21:35 arkiuat joined 21:45 arkiuat left 22:16 arkiuat joined 22:21 arkiuat left 22:37 arkiuat joined 22:55 arkiuat left 23:06 librasteve_ left 23:35 abraxxa 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!