🦋 Welcome to Raku! raku.org/ | evalbot usage: 'p6: say 3;' or /msg camelia p6: ... | irclog: colabti.org/irclogger/irclogger_log/raku
Set by ChanServ on 14 October 2019.
[Coke] (assuming the decisions are based on their activity through... now) 00:00
voting closed.
(I'd push sooner, but if I push a partial, then add a vote and re-push, it won't be a secret ballot) 00:02
guifa [Coke]: thanks for organizing it all 00:03
AlexDaniel [Coke]++
timotimo who wants a rakudo appimage 00:04
[Coke] gist.github.com/coke/07f167179135f...d526d5437e - updated to note who was invited post hoc, and changed ids to links so it's easier to research if we should invite them. 00:05
00:12 mowcat left 00:14 klapperl joined
Juerd I have no idea what docs.raku.org/type/Real is trying to explain. 00:38
Or when you should ever use Real instead of Num. 00:39
Another question: is there a built-in floating point that doesn't allow NaN? That's what I expected Real to be but it's clear that it's certainly not that :) 00:40
00:42 aborazmeh joined, aborazmeh left, aborazmeh joined
Grinnz the eternal struggle: how do you put NaN in a number 00:42
Juerd Since $foo = NaN is too easy that was probably not your question. What do you mean? 00:43
Grinnz i mean semantically because NaN represents not a number 00:44
Juerd Still no idea what you mean :)
Grinnz and then the step further: inf is technically not a number 00:45
i'm just math-trolling really
Juerd It may not be a number, but isn't NaN either :)
m: Inf.isNaN;
camelia ( no output )
Juerd m: Inf.isNaN.say
camelia False
Grinnz about how ieee-754 is extremely unmathematic :P
timotimo github.com/ddmitov/perl-executing-browser - anybody interested in stealing this for raku? 00:46
Juerd In any case, passing NaN to functions taking Num rarely produces anything useful and since NaN is basically a float's exception mechanism, it would be useful to just escalate that to actual exceptions. In fact, if there was a built-in subtype of Num, where not *.isNaN, I'd probably use that most of the time even :) 00:47
s/Num,/Num/
Same for Inf actually. I kinda expected Real to be the thing that only supported real numbers, and iirc, while R is infinite, infinity itself isn't part of it... 00:50
Grinnz indeed 00:51
guifa You could actually probably modify the handling of numbers to throw exceptions when you get NaN 01:04
timotimo use NeverNaN; 01:05
Juerd Maybe a type called Actual to be the real Real :P 01:06
timotimo RealEx
Juerd timotimo: That might come across as extended real, which would include -Inf and Inf :)
timotimo ha
guifa timotimo: and here I thought that was an acual module 01:07
Juerd I hate that English has no opposite of "not". Dutch has it, "wel" is the opposite of "niet". Though Raku went with "so" in that sense, so maybe SoReal :)
guifa “indeed” 01:08
not real, indeed real
timotimo Surreal
Juerd lol
guifa how can you find individual candidates of a multi sub? 01:09
timotimo "find"?
m: .say for &say.candidates
camelia &say
&say
&say
timotimo m: dd $_ for &say.candidates 01:10
camelia Sub say = multi sub say { #`(Sub|65779248) ... }
Sub say = multi sub say (\x) { #`(Sub|65779392) ... }
Sub say = multi sub say (|) { #`(Sub|65779536) ... }
guifa timotimo: I’ve been having a lot of fun with wrapping lately
hmm 01:15
I can’t seem to wrap any of the + subs
timotimo are you sure your code isn't just getting its uses of + optimized away? 01:16
01:16 molaf left
guifa That’s certainly possible 01:16
Yup that’s what it seems like 01:17
01:18 __jrjsmrtn__ left 01:19 __jrjsmrtn__ joined
codesections guifa: that's both interesting and slightly troubling. I'm assuming that your wrapping code introduced a side effect (e.g., &say). I would have expected/hoped that that would prevent inlining 01:19
timotimo wrap happens at runtime, a part of inlining happens at compile time
guifa It’s both a good thing and a bad thing about wrap
timotimo i think there's a "soft" parameter and pragma that's supposed to control this? 01:20
codesections Why isn't wrap compile-time?
timotimo because you haven't put it in a BEGIN block
codesections (I guess I expected it to be similar to a macro)
timotimo yeah, like "use" or so 01:21
guifa m: &infix:<+>.wrap( sub (|c) { my \x = callsame; die X::AdHoc.new(payload=>"BAD MATH") if x.isNaN; x; }); sub add($a, $b) {$a+$b}; say add NaN, 8 01:29
camelia BAD MATH
in sub at <tmp> line 1
in sub at <tmp> line 1
in sub add at <tmp> line 1
in block <unit> at <tmp> line 1
01:29 molaf joined
timotimo notable6: weekly github.com/timo/rakudo-appimage/re.../2020.08.2 01:29
notable6 timotimo, Noted! (weekly)
timotimo notable6: weekly github.com/timo/moarperf/releases/...-appimage1 01:32
notable6 timotimo, Noted! (weekly)
timotimo please check out the first link there
01:35 aborazmeh left
[Coke] sees Surreal Numbers in the backlog and wonders if that's worth a module. 01:43
codesections [Coke]: would the text of the module just be 01:45
m: subset Surreal of Real where * !~~ Inf | NaN
camelia ( no output )
01:45 ex_nihilo joined 01:46 BenGoldberg joined
[Coke] no, I imagine you'd need .Str to show the set driving the Surreal #. 01:46
... or maybe .set 01:47
02:01 stoned75 left
Juerd m: subset Surreal of Numeric where -Inf < * < +Inf; say (-Inf, -9, 0, 9, Inf, NaN, 2/7) »~~» Surreal 02:21
camelia (False True True True False False True)
Juerd tbh I don't really care if it's a Rat or a Num anyway, as long as it's an actual, real, number :) 02:22
Maybe just Number. Sufficiently confusing to have Num and Numeric already, I don't see much harm in adding a third one. 02:23
Or the punnier name Numb
guifa I’d go with something like ConcreteNumber. Although concrete already exists in the number world for something else 02:26
Could go for a reduplicative approach 02:27
subset NumberNumber …
Juerd For the numeric type I'd end up using most, a long name would not be huffman coded well.
guifa NumNum ha 02:33
or even N
02:34 gnufr33dom left
Grinnz Really 02:34
Juerd Grinnz: Ooh :) 02:35
I guess there's not much use for a subset of Bool where .so, but it could be called Truly :P 02:36
guifa codesections: that subset doesn’t work, because actually a surreal number includes infinite and infinitesimal numbers
02:38 Xliff joined 02:42 stoned75 joined 02:47 stoned75 left 02:59 stoned75 joined 03:19 BenGoldberg left 03:28 stoned75 left 03:34 stoned75 joined 04:34 nativecallable6 left, sourceable6 left, committable6 left, greppable6 left, bisectable6 left, tellable6 left, releasable6 left, statisfiable6 left, reportable6 left, coverable6 left, linkable6 left, unicodable6 left, squashable6 left, bloatable6 left, notable6 left, quotable6 left, evalable6 left, shareable6 left, benchable6 left, epony left, linkable6 joined, shareable6 joined, committable6 joined, reportable6 joined, tellable6 joined 04:35 releasable6 joined, quotable6 joined, nativecallable6 joined 04:36 benchable6 joined, statisfiable6 joined, squashable6 joined, unicodable6 joined, sourceable6 joined, bloatable6 joined 04:37 evalable6 joined, greppable6 joined, notable6 joined, bisectable6 joined, coverable6 joined 05:18 aluaces joined 05:26 gnufr33dom joined 05:39 stoned75 left 05:45 molaf left 05:57 gnufr33dom left 05:59 gnufr33dom joined 06:00 cpan-raku left 06:01 cpan-raku joined, cpan-raku left, cpan-raku joined 06:09 BenGoldberg joined 06:14 jmerelo joined
jmerelo We've reached 1500 questions in StackOverflow right by the end of summer stackoverflow.com/questions/tagged/raku 06:18
Will we reach 2000 before the end of the year? 1600, more likely, but still...
06:27 ex_nihilo left 06:34 stoned75 joined 06:37 Altai-man joined
jmerelo releasable6:status 06:38
releasable6 jmerelo, Next release will happen when it's ready. 3 blockers. Changelog for this release was not started yet
jmerelo, Details: gist.github.com/040925d93942579bb7...be8abac3d2
tobs Juerd: I think there is even use for the Truly type as (:$flag where *.so) in the signature gives you a "flag" adverb. 06:41
06:44 BenGoldberg left
tobs As for Real, the hierarchy is: Numeric (lots of things) -> Real (real numbers) -> Num (IEEE representation). The Real role in the middle admits multiple representations of real numbers, like bigfloats and isolating roots of polynomials. I think that's the reason it exists, even though it does not forbid infinities or NaNs. 06:45
06:47 gnufr33dom left
moritz Real also includes things like Int, but not Complex 06:47
06:54 aborazmeh joined, aborazmeh left, aborazmeh joined, stoned75 left 07:00 stoned75 joined 07:03 sena_kun joined 07:05 Altai-man left 07:32 abraxxa joined 07:34 Sgeo left 07:37 abraxxa left, abraxxa joined 07:45 dakkar joined 08:02 ingy left 08:06 domidumont joined
El_Che jmerelo: did you post the results of the survey somewhere? 08:07
sena_kun El_Che, see github.com/Raku/survey/releases/tag/v2020.09 08:11
08:11 Kaiepi joined
El_Che thx 08:14
1 page :)
jmerelo El_Che: Just the summary, thanks sena_kun++ 08:18
El_Che: I've posted some stuff in Twitter, also. The R files and charts are in the repo, too.
El_Che the pdf mangles questions with lots of words, like about installation 08:19
probably the output was broken on the google site already
08:20 pecastro joined 08:23 aborazmeh left 08:24 aborazmeh joined, aborazmeh left, aborazmeh joined
Xliff jmerelo: Huh. I didn't know R was that nice in terms of graphing. 09:21
Wonder if I can make a Raku version that clean.
What do you think it would take?
moritz a good graphing library would be a nice start :D 09:23
JJAtria[m] Was that using ggplot? In my R days that was the good library to use, but that was a while ago 09:24
Xliff moritz: Give me a couple. 09:25
jmerelo Xliff: R + ggplot2 Nicer than matplotlib, actually. And there's a lot of data massaging, functional style, you can do.
JJAtria[m]: correct. ggplot2, actually.
Xliff looks at ggplot2
jmerelo Xliff: it's mostly written in R, though. It'd probably be easier to start Inline::R first 09:26
Nice thing about ggplot2, it's a graphing grammar, functional in design. Syntactically, wouldn't be so difficult to port. Graphing is a different history altogether
JJAtria[m] I liked the idea of ggplot more than actually using it when I had to, but I never used the second version. Back then, I ended up moving entirely to pgfplots 09:27
09:27 cpan-raku left
Xliff jmerelo: I'm not a compiler designer, unfortunately. I'm more a NativeCall hacker. 09:28
jmerelo JJAtria[m]: apparently, the first version wasn't so nice. ggplot2 really makes sense in terms of composing parts of a chart together. Also, very nice transformation primitives you can use from the get go. Plus "themes" like gg_themes that add look&feel; I use "Tufte" most often.
Xliff jmerelo: I've got 20 projects worth of language binding code that just needs a few improvements in the Raku build chain before I can start releasing. 09:29
jmerelo But leaving aside the charts and stuff, I'm slightly worried about the results.
09:29 cpan-raku joined, cpan-raku left, cpan-raku joined
jmerelo To put it mildly, it does not show a vibrant and thriving language attracting lots of new developers. 09:29
Xliff Explain 09:30
jmerelo As a matter of fact, only 20 responses say they've started using it during the last year, down from 40ish in the previous one.
Xliff Ah. 09:31
We need to get Raku taught in schoolos.
s/choolos/schools/
jmerelo The percentage of people coming from Perl hasn't changed much, either.
Xliff Even if it starts off as extra-curricular.
jmerelo Well, I do all in can in this little corner of the world to show it off in my classes. But it's not enough. 09:32
Xliff jmerelo: I think I shared this with you before -- drive.google.com/file/d/12j1Dy6zIc...sp=sharing
jmerelo Maybe, but I don't remember having checked it out. Good work.
Xliff That's an old version. I'm up to 450+kloc 09:33
jmerelo Xliff: it's your own tally of contributions to Raku, right?
Xliff Yes. Unreleased currently due to install times using current zef
jmerelo Just the G* contributions are really awesome.
Xliff So if you look at column N, that's the minimum number of seconds to install uzing zedf. 09:34
s/zedf/zef/
09:34 BenGoldberg joined
Xliff Even getting it down to the O column times would be a bit of work, thoguh much of those improvements will make it into 2020.09, I think. 09:35
jmerelo Xliff: that would be awesome... yes, zef takes some time in ocassions. That's why I create custom Dockerfiles to test, most of the time. 09:36
But, wow, 2225 is *a lot*
I guess it will improve in time... Now that we were mentioning R, some libraries also take a lot of time to install. Don't think it's a big hurdle for adoption of the language 09:38
(and that's probably not your point, either)
Xliff No. The big one is lack of feedback in current zef. 09:39
For installing these, the user needs more than what you see in the current implementation. 09:40
I'd like more text emitted during install.
Ideally I'd like to write a custom callback zef can use so that I can feed installation information back to the user.
Don't have any idea of how such a mechanism would work, though. 09:41
Next big step would be binary packaging so that end-users wouldn't need zef.
jmerelo Xliff: that would be really nice. Even simple source-based packaging would be a plus. 09:44
Xliff Yeah. Lots of plans, but no direction yet. 09:45
So that's the first thing I'd lobby the steering council for.
jmerelo Xliff: exactly what I was writing right now. And deleted. 09:46
Xliff Ah.
Why deleted?
Not the right words? Not the right time?
Is it something related to your reasoning above? About the data showing Raku as not being vibrant and thriving? 09:47
I kind of get that. However the Raku language design is playing the long game. 09:48
Much of Raku's features are ahead of their time.
jmerelo Xliff: you have said it already :-)
Xliff I think you are right in your efforts to get Raku into educational facilities.
In the meantime it will take the concerted efforts of everyone here to keep the torch lit until we start gaining users in serious numbers. 09:49
jmerelo Xliff: not nearly enough, though.
Xliff Raku is not even 5 years old yet.
jmerelo Xliff: it definitelly will need those efforts, and then some...
Xliff And compile times are putting off adoption in the efforts where Pythin, Ruby and others gained their early adoption. 09:50
For web development, Raku really needs an app-server before it can be considered.
jmerelo Xliff: you mean something like GUnicorn? 09:51
Xliff Cro provides the pieces to make such. In its current state it falls far short of apache + mod_perl, which what we need to shoot for.
Xliff atempts to find the mythical GUnicorn
jmerelo Something I _really_ miss is a coverage tool.
Xliff jmerelo: Um yeah. I do mean that.
jmerelo Green Unicorn is a PSGI front-end, which you can put in front of webservices written in Python.
I think that until we have a good coverage tool, it's going to be very difficult to get Raku into production. Without coverage tests, few people are going to be putting Raku extensively into production. 09:52
Xliff Cro is almost there. It just needs code to implement a true application server.
jmerelo Xliff: right 09:53
Xliff: at any rate, we are into the mentality of "If we build it, they will come". Not really sure about that.
Xliff Then instructions on how to serve it all using something light weight for the reverse proxy... like nginx
jmerelo: Look at the converse. "If we don't build, they won't come." 09:54
Which is absolutely true.
jmerelo Xliff: right, too. Cro is nice by itself, however. Much more powerful that other similar things like express or flask.
Concurrent, for one thing.
Xliff We need to do more than build. We need to start advertising for one thing.
jmerelo But you're right. The problem is who is "you", the proverbial "you" that build the things we need... 09:55
Xliff So a Raku blog which we can repost from would be a good thing.
09:55 aborazmeh left
jmerelo Xliff: I'm in the TPF marketing committee... Only 2 representatives of Raku, though. 09:55
Xliff jmerelo: In my mind, that's this community.
It would be nice if we got help from the greater perl community rather than disdain.
jmerelo Xliff: yep, but we're kinda overworked here. Just the docs would need like three times the effort we're devoting now. 09:56
Xliff I know.
jmerelo Xliff: and tbrowder talked about creating a Raku blog. But taking into account how hard it is to put together a * calendar, I really don't know about blogs... 09:57
sena_kun There is coverage for Raku code. A handy UI is another question, I'm using Comma UI and it's fine for me. Used it for Pod::To::HTML refactoring recently.
Xliff The community, as it stands now, is healthy. We are innovating. We are creating. It's just that seem to be doing so in am empty room.
jmerelo sena_kun: where's that tool?
Xliff That generally means we are missing a "killar app" that would draw people to the language.
jmerelo sena_kun: comma covers very well the tooling area. No complain in that area. You can also use raku-mode if you don't like proprietary software. 09:58
Xliff Comma is nice, but it has it's problems handling things larger than a typical ecosystem project script.
It falls over on p6-GLib.
sena_kun jmerelo, MVM_COVERAGE_LOG=some-dir-path in env and then process results to show nicely. 09:59
Xliff So it looks like I might need to take classes on the proper use of Comma so I can really evaluate it as a tool.
Because the current documentation isn't nearly enough.
jmerelo Xliff: don't believe too much in that, either. It worked for Ruby, with Ruby on Rails. In the broad sense, data-science did it for R, Scala, Julia...
But still. Development is incredibly broad.
Xliff jmerelo: So let's actually build some data-science into rakudo. 10:00
10:00 leont joined
Xliff The more areas Rakudo covers, the more likely it will be discovered and used. 10:00
jmerelo sena_kun: that's MoarVM coverage, right?
sena_kun jmerelo, yes?
Xliff jmerelo: I am, I think, maybe a year or two out before I have enough G* coverage to actually have enough pieces to write a desktop manager in Raku. 10:01
jmerelo sena_kun: I was looking more at white-box testing, that showed source-code test coverage. But let me check that anyway.
Xliff: that would be cool.
sena_kun jmerelo, erm, it is about source-code test coverage.
jmerelo Xliff: in data science, some good work is being done wrapping the Gnu Scientific Library by Frithnanth.
sena_kun: let me check then.
sena_kun: I vaguely remember that was not what I was looking for, but I'll check back. 10:02
Xliff I think I may look into bindings for ggplot2 10:04
MasterDuke jmerelo: it's a MoarVM feature/flag, but the reports it creates are of the higher level code. i.e., it's not showing MoarVM coverage
sena_kun have an impression this was talked over and over. I work on Comma at $dayjob whenever time allows and after that I spend my weekends and free time rewriting doc tooling, investing funds there, writing numerous posts, etc. Telling "You folks don't do enough" is not very actionable. I do agree with a stance like "if you are quietly improving a thing the people just won't know about it without marketing", but it is not a job for people who fix bugs and do
software. If RSC will become a thing, marketing/adoption can be set as a high priority question.
Xliff Got couple of weeks on refactoring GIO.
Then I have to get back to SOUP tests.
Then... maybe... ggplot2
sena_kun s/ have/I have/ 10:05
Xliff sena_kun: Maybe you can teach me Comma? :)
Or at least show me what I am doing wrong there. 10:06
sena_kun Xliff, I can try. :)
Xliff If I can start using it for G* dev, progress would improve exponentially!
sena_kun: Plus... that's one way I can get you to download all of my code. ;q
sena_kun Where does it ache?
jmerelo sena_kun MasterDuke I see now, but it's not a _tool_ proper, or a coverage tool at that
Xliff sena_kun: Debugging
sena_kun Xliff, OS? 10:07
timotimo jmerelo: hey, can you explain something about the analysis of the survey?
Xliff sena_kun: Linux
sena_kun Xliff, I see. Let me guess, your breakpoints are not stopping where they should?
jmerelo MasterDuke: you can try and create a coverage tool out of that, of course, but you would need to parse that log, check out the original source it's supposed to check, and then say: well, we have 90% coverage or whatever. That's what I mean.
Xliff I find that JetBrains internals will either crash outright, or I'll lose sync with the debugging process and then lose things like local variables and such.
sena_kun: Aha! That's a large pain point. How did you guess? :) 10:08
jmerelo timotimo: certainly, but not done a lot of analysis so far. Basically charting. But if you want me to clarify something, shoot
sena_kun Xliff, this "bug" was there from the beginning. I think it is a MoarVM level issue and we have no experts with time on board to tackle on this.
10:08 BenGoldberg left
Xliff sena_kun: Ah. Ouch. 10:09
sena_kun Xliff, re "internals crash" - I need traces to look at, it usually gives you a nice stacktrace and all on exception. :)
Xliff So... people with MoarVM experience should split time between developing new things and fixing old ones.
Of course, that means better time management for things being done away from $dayJob. 10:10
I get it.
10:10 thundergnat left
Xliff I think we need more core devs. 10:10
More people to shadow folsk like jnthn++ and niner++
s/folsk/folks/
Yeah. My fingers going dyslexic.
And lizmat++ 10:11
sena_kun In a volunteer project where people do extremely intricate things for free, yes. :]
Xliff Another thing that would help would be to get commercial interest in raku.
sena_kun .oO ( just you wait for my announce in a week or two )
Xliff It's frustrating to write something like this: github.com/Xliff/p6-Amazon-AWS-EC2 10:12
timotimo jmerelo: in many of the answers there are multiple entries for "the empty string" or something. can that be?
also, of course the parts of the charts that are scrollable are useless in the pdf :)
Xliff Only to get it into testing phase and then run into this: github.com/rakudo/rakudo/issues/3075 10:13
10:13 wbn left 10:14 Black_Ribbon left
MasterDuke jmerelo: yep. there is a script in the rakudo repo, but i think its bitrotted 10:15
10:15 wbn joined
jmerelo timotimo: that's most likely due to questions being added after the survey started. Or simply non-mandatory questions without an answer. 10:15
timotimo: can weed that out if you want. 10:16
sena_kun Xliff, I get you, really. People relatively often say "Is Raku slow, where performance, we need speed", but I found just plain, boring bugs more frustrating to deal with.
MasterDuke github.com/MoarVM/MoarVM/blob/mast..._report.p6
jmerelo MasterDuke++ 10:19
MasterDuke: couldn't that be spun off to an ecosystem module? 10:20
MasterDuke probably
jmerelo MasterDuke: it needs core.SETTING, oops, that might be hard, but probably will do with a few nice defaults. 10:21
timotimo jmerelo: does that explain multiple empty answers in the same question? i guess one for every "version" of set of answers? 10:22
and also, some things are cut off with no way (in the pdf) to get the rest of the text 10:23
jmerelo timotimo: it might. The PDF is not complete by any means. It's just a screen capture. Problem is, GDrive forms don't generate any other kind of nice summary. 10:24
timotimo: the CSV is uploaded to the repo, anyway.
timotimo: wouldn't worry too much about empty answers, anyway.
lizmat jmerelo: so what should I link to in the Weekly ? 10:25
10:26 rindolf joined
jmerelo lizmat: the PDF summary is probably your best bet, until we drill down on the results. 10:27
lizmat: you can also link the CSV with all the answers.
10:28 stoned75 left 10:29 stoned75 joined
jmerelo lizmat: the PDF 10:45
lizmat: (short answer)
MasterDuke jmerelo: fyi, just tried the script and it does work. `moar --dump ../rakudo/blib/CORE.c.setting.moarvm >annotations; mkdir coverage; MVM_COVERAGE_LOG=a.log raku -e 'say "hi"'; raku ../MoarVM/tools/parse_coverage_report.p6 --annotations=annotations a.log ../rakudo/gen/moar/CORE.c.setting` 10:48
11:02 Altai-man joined 11:05 sena_kun left
jmerelo MasterDuke: so good candidate for spinning it off... somehow. It would need multi-file support, for starters, and some tests too. Thanks! 11:14
timotimo can you give an iframe a src with a data url? 11:15
MasterDuke from github's share button: <iframe src="gist.github.com/MasterDuke17/f7333...iframe> 11:17
if that's what you mean? 11:18
timotimo ah, no it isn't 11:30
data urls are where you put the actual contents in the href or src as base64
with a special prefix to mark what it actually is
MasterDuke ah 11:32
timotimo a way to have one file with the contents of many files in it, supported by browsers 11:37
11:46 Tirifto joined 11:47 aborazmeh joined, aborazmeh left, aborazmeh joined
AlexDaniel` one file with contents of many files… 11:53
did you use Junctions a bit too much? xD
timotimo ha
11:58 Manifest0 left 11:59 Manifest0 joined 12:20 Manifest0 left
jmerelo timotimo: you need to use <script> for that. Some frameworks also allow data-* attributes in tags. 12:29
timotimo hm. 12:30
12:52 Kaiepi left, stoned75 left 12:59 jmerelo left 13:00 BenGoldberg joined 13:04 stoned75 joined 13:07 Manifest0 joined 13:21 Manifest0 left, Manifest0 joined 13:26 Manifest0 left 13:27 Manifest0 joined 13:32 Manifest0 left 13:33 Manifest0 joined, BenGoldberg left 13:56 skids joined 14:01 aborazmeh left 14:06 Manifest0 left 14:07 Manifest0 joined 14:13 aluaces left 14:15 Manifest0 left 14:16 Manifest0 joined 14:25 epony joined 14:29 vrurg_ is now known as vrurg 14:32 ggoebel joined 14:33 Sgeo joined 14:38 Kaiepi joined
Geth Raku-Steering-Council/main: 64991d01e3 | Coke++ | nominations/2020/results.txt
Submit voting results of September 2020 election
 14:52
[Coke] Voting tabulated and submitted. Congratulations to the top 7: jnthn, lizmat, nine, JJ, codesections, AlexDaniel`, ugexe 14:54
moritz [Coke]++ # one-man voting council 15:02
15:03 sena_kun joined 15:04 molaf joined
[Coke] eh. with < 50 votes, it was not a terrible amount of work. Also, it stopped me from self-nominating. :) 15:04
happy to help
15:04 molaf left 15:05 molaf joined, Altai-man left
Altreus Is there a succinct way of flattening a hash into a list of (key, value, key, value) rather than pairs? 15:07
[Coke] m: my %a = <a b c d e f g h> ; dd %a.kv 15:08
camelia ("a", "b", "c", "d", "g", "h", "e", "f").Seq
[Coke] ^^ 15:09
Altreus ah! thank
MasterDuke i don't remember seeing anybody mention a blockchain, can we *really* be sure the voting was done correctly? 15:11
AlexDaniel [Coke]: it says =?UTF-8?Q? in the file?
[Coke] AlexDaniel: yes.
because unicode is hard and I'm lazy 15:12
codesections MasterDuke: xkcd.com/2030/
[Coke] MasterDuke: no. you have to trust me, basically.
I don't think any of the votes were signed, either.
AlexDaniel: I'll add a note about the malformed utf in the file.
moritz [Coke]: for increased trust, I recommend you take a few lego blocks, add a chain around them, take a picture, and add the caption "secured by block chain" :D 15:13
AlexDaniel Congrats to everyone! 15:14
codesections moritz: s/trust/buzzword compliance/ :D
Geth Raku-Steering-Council/main: 2a4d15bb8f | Coke++ | nominations/2020/results.txt
Fixup voter names that require unicode
[Coke] AlexDaniel: fixed the malformed unicode
AlexDaniel [Coke]: thank you for the work! 15:15
[Coke] future vote tally people need to do a better job on unicode handling in votes.
(I basically cheated and used utf-c8 rather than trying to follow the encoding in the file)
moritz: :P
15:15 rir joined 15:16 jmerelo joined
tobs (Pretty sure I signed my vote.) 15:19
[Coke] ah, so you did 15:21
looks like the only one, don't see a signature.asc on the others. 15:22
jmerelo So I guess that's the result github.com/Raku/Raku-Steering-Coun...38254907ce 15:26
Many thanks. Hope I'll be up to the task 15:27
And congrats to jnthn, lizmat, nine, codesections, AlexDaniel and ugexe
[Coke] jmerelo: github.com/Raku/Raku-Steering-Coun...esults.txt
AlexDaniel tobs: nice! 15:28
codesections jmerelo: thanks, and congrats to you too :)
MasterDuke congrats all 15:29
jmerelo [Coke]: and thanks for all the work in carrying this out. 15:33
[Coke] jmerelo: would have pinged you in my send earlier, but was looking for "JJ", apologies. :) 15:40
jmerelo [Coke]: :-) No problem. I registered this nick here in this channel, so it's what I use. 15:42
15:45 jmerelo left, Chi1thangoo left 15:46 sjaveed joined
leont [Coke]++ 15:47
[Coke] .tell ugexe results are submitted, congratulations (RSC) 15:50
tellable6 [Coke], I'll pass your message to ugexe
[Coke] I think that's everyone. 15:51
15:51 sjaveed left
timotimo mildly surprised jonathan isn't at a lower number due to the "anybody's going to vote for him anyway, so i can sneak in a different fifth" effect 16:10
Xliff Congrats to the new Raku Steering Council: jnthn++ lizmat++ niner++ jmerelo++ AlexDaniel++ codesections++ 16:20
ugexe++
16:25 BenGoldberg joined
guifa timotimo: but that risks everyone doing that and him not getting on the committee at all 16:38
16:43 dakkar left 16:46 domidumont left
rypervenche If I've got a variable that I would like to use in a grammar's actions, but I set said variable in my MAIN sub below the grammar, what would be the best way to do that? 16:50
codesections rypervenche: if I understand what you'd like to do correctly, you can do that by passing the argument into the grammar with :args 16:52
docs.raku.org/language/grammars#Pa...o_grammars
is that what you're looking for?
16:53 stoned75 left, eseyman left
rypervenche It looks close. Will I be able to use those arguments in an Action class as well? Here it looks like it's for the grammar only. 16:54
16:54 BenGoldberg left, Ben_Goldberg joined 16:55 Ben_Goldberg is now known as BenGoldberg 16:57 eseyman joined, approaching236 joined 16:59 BenGoldberg left
rypervenche codesections: I'm trying to use $number (which is a CLI-passed argument) in my Actions. Would a dynamic variable be a good tool for this? gist.github.com/rypervenche/eb3925...031be7b111 17:01
17:04 domidumont joined, stoned75 joined
codesections rypervenche: Yep, a dynamic variable seems like a good fit there (i.e., changing $number to $*number) 17:17
rypervenche Oh, that's very easy. I've never used dynamic variables before. 17:23
codesections They're powerful in the right situations, but can also make code very hard to reason about when used more extensively 17:24
17:29 Black_Ribbon joined 17:42 rir left
lizmat [Coke]++ # being the votemaster 17:43
17:45 domidumont left, nige joined 17:46 aluaces joined, nige left 17:47 jmerelo joined 17:53 kent\n is now known as Mickey_Mouse 17:54 Mickey_Mouse is now known as kent\n, Xliff left 18:00 nige joined
nige congrats to all the elected RSC members 18:01
just wanted to reach out to the new council - it would be great if we can continue a good working relationship together with TPF 18:07
JJ has been helping on the marketing committee and more Raku community members are welcome to join and help out 18:08
codesections nige++
nige On Friday a new legal committee for TPF was constituted and would be great if someone from the RSC could also join that committee 18:09
the legal committee has a commercial focus too - and we will be looking at putting things in place to help Raku flourish 18:12
El_Che Personally, I would love to see less perl+raku marketing ans market them as completely different languages. I suppose that's what Perl will do once 7 is out 18:13
codesections nige: is there a page describing the legal committee/any other info anywhere? I don't see anything on the TPF website
nige it was only constituted on Friday - however I can share some of the objectives etc (will send a gist soon) 18:14
jmerelo nige: commercial as in trademarks and such, right?
codesections nige: thanks 18:15
nige codimd.opusvl.com/s/jD95bV_Ue 18:16
yes - we are going to need different marketing strategies for the different brands (Raku, Perl 7, Perl 5, TPF, meta::cpan etc) - they all have different markets and messages etc 18:18
the TPF itself is in the process of rebranding 18:19
El_Che nige: sure, I mean everything shared to the outside world: conferences, tables, leaflets etc
18:20 molaf left
JJAtria[m] m: class Foo { has method bar { 123 } }; say Foo.bar # has method?? 18:21
camelia 123
nige yes - Perl and Raku are at very different points in their lifecycle - and Raku marketing needs to focus on early adopters
El_Che exactly
codesections nige: I'd be happy to volunteer for the legal committee (though I would also be happy to defer to any of the other RSC members if they would rather participate in that committee). 18:27
Before transitioning to software development, I was a practicing attorney and, though I no longer practice, I have maintained my license, so I may have a useful perspective for the committee
jmerelo nige: early adopters... not coming from Perl.
El_Che jmerelo: those that wanted to use raku, already do
(from Perl)
jmerelo El_Che: not so sure. Data do not really support that. I'd say that there's been more people adopting it lately, since it's called Raku 18:28
nige codesections - that would be fab you're very welcome to join. Also open to other members joining
jmerelo But I'll have to check anyway.
El_Che looking fron anedoctical data I'd say perl people are moving to Go, Python and co 18:29
jmerelo El_Che: they're moving... elsewhere. Including Raku.
El_Che (at go related events and communities I meat a suprisingly high number of Perl people)
nige well I think the recent survey showed that many of the current adopters came from Perl (and probably will in future) 18:30
so I think the marketing message needs to talk to Perl people too - it's not a binary thing though
jmerelo nige: if that's where they're mainly coming from, it's a very small niche indeed.
nige you can program in Perl + Raku
jmerelo nige: and dwindling at that.
nige: we should try and fish in Ruby's waters, for instance.
nige yes - I hear you on that
jmerelo Or early adopters. 18:31
El_Che ruby's waters are pretty dry
is experiencing what Perl lived through before
nige Rails did help Ruby along
and I think Raku could offer something similar 18:32
El_Che too little too late
raku needs something like cro to stay competitive
but itĹ› no longer a selling point in 2020
nige the other market not to forget is the Perl diaspora - a lot of programmers have a lot of fondness for Perl - and would love Raku 18:34
El_Che *there* you need the unique selling point 18:35
because they are not coming back from reacher ecosystems
I think that jj has a great point: early adopters 18:37
nige: lots of brainstorming on that commitee :) 18:41
AlexDaniel` JJ Atria: yeah, interesting :) 18:46
AlexDaniel hmm, the marketing question 18:49
remember “We suck at marketing.”? I think the marketing is fantastic. There are so many things that Raku promises or implies but does not really deliver 18:50
El_Che AlexDaniel: it could also be a focus question
AlexDaniel: you could be an inspirational speaker :P
AlexDaniel the real question is what Raku currently is and what it should become, and then it'd be nice to make it actually happen 18:51
El_Che See: 20:50 < El_Che> AlexDaniel: it could also be a focus question
nige AlexDaniel - that's the nub of it really - what are the top three brand values for Raku now? what are the top three in future? ... the recent marketing survey shows something interesting there 18:53
this helps reveal the brand trajectory
knowing what the values are helps to uncover who the early adopters are and what messages they need to hear 18:54
one of the difficulties to be honest is 'performance/speed' has come up as a value in future 18:55
this is not really a brand value per se - but more like a feature
timotimo we have at least 5 performance / 1 speed
is that a good ratio? 18:56
nige and branding is about being brutally honest
the marketing message needs to match the reality - otherwise it all backfires
El_Che there is also the danger of trying to placate the present (or rather past) early adopters vs the future ones
jmerelo nige: survey results are ready, check them out... 18:57
El_Che ie is it a trajectory of where you want to be in the feature or were the present users see you
nige thanks JJ - I need to go back and review - I saw the results about a week or so ago 18:58
the trajectory is the difference between the values now and the values in future 18:59
El_Che nige: it depends on whose values
nige sometimes bridging that gap is actually about addressing the problem - with something practical
jmerelo goes AFK for today. See you tomorrow!
nige other times it's highlighting some truth that was not salient enough
El_Che the future may be a consequence of the vision of the project
18:59 jmerelo left
nige bye JJ 19:00
El_Che and that vision may not correspond with the vision of a specific user base
19:02 Altai-man joined
nige fortunately even though speed may not be Raku strong suit at the moment - there are still many other qualities that make it attractive to some early adopting audiences 19:03
19:05 sena_kun left
El_Che certainly 19:05
nige for example a CS100 lecturer - who wants to give students a taste of functional / oo / procedural programming / reactive styles of programming
The python marketing committee is actively lobbying to have python included in teaching curriculums etc 19:06
El_Che well, it's already the case in most places 19:07
nige yep - they have done well - here in the UK, python seems to be the language of choice for school teaching 19:08
El_Che I think almost every where if I can be somewhat eurocentri
c
nige another potential early adopting audience is sys admins/devops doing command-line scripting - where performance is less at issue - and whipupitude is important 19:10
codesections: could you please send me your email and I will invite you to the legal committee slack channel (nige (at) 123.do) 19:11
El_Che do you fancy getting involved with some of the marketing plans? You're welcome to join the marketing committee ... 19:16
El_Che nige: I am too short of time at the moment, but maybe in the future
nige OK - no worries 19:21
codesections nige: my email is daniel (at) codesections.com (I list it on my website, so I don't mind posting it here)
Geth Raku-Steering-Council/main: 702080ca68 | Coke++ | nominations/2020/acceptance.txt
add file to track "signatures" of winners.
 19:22
raku-bridge <DataKinds> hey, does Raku have a WSGI implementation yet? 19:23
Altai-man github.com/zostay/RakuWAPI 19:24
raku-bridge <DataKinds> ah, so nothing usable yet then? 19:25
Altai-man Not sure if it's actually usable or not. I think it was presented as blogs.perl.org/users/sterling_hanen...rface.html and then evolved into RakuWAPI project? 19:27
raku-bridge <DataKinds> I suppose I can use Inline::Perl and plack ahah
<DataKinds> mmm, yeah, it looks like the whole spec is there, it's just that nobody seems to have actually put in the work to write the middleware 19:28
Altai-man github.com/tokuhirom/p6-Crust <- this? 19:29
raku-bridge <DataKinds> perhaps this too modules.raku.org/dist/Smack:cpan:HANENKAMP
Altai-man Yup, something along the lines.
raku-bridge <DataKinds> doesn't look like Crust has been updated in quite a long time, I wonder if it still works 19:30
timotimo any particular reason not to just listen for HTTP and mod_proxy the app?
raku-bridge <DataKinds> just for fun 19:31
timotimo OK
we have like PSGI or something, right?
raku-bridge <DataKinds> I've done that before, I've never set up a raw CGI/WAPI/WSGI/Rack/.... server
<DataKinds> yeah, the links above
19:32 abraxxa left 19:33 abraxxa joined
raku-bridge <DataKinds> it seems like there was some attempt made in making RakuWAPI compatible with PSGI? 19:34
Geth Raku-Steering-Council: codesections++ created pull request #35:
Add my acceptance and signature		 19:39
Raku-Steering-Council/main: b29719af65 | (Daniel Sockwell)++ | nominations/2020/acceptance.txt
Add acceptance signature
 19:40
Raku-Steering-Council/main: 7293d7329c | (Elizabeth Mattijsen)++ (committed using GitHub Web editor) | nominations/2020/acceptance.txt
Merge pull request #35 from codesections-forks/codesections-signature

Add my acceptance and signature
Raku-Steering-Council/main: e69e4aca37 | (Elizabeth Mattijsen)++ | nominations/2020/acceptance.txt
Add my acceptance and signature
 19:41
[Coke] wonders why codesections made a PR. 19:47
19:51 BenGoldberg joined 19:56 mowcat joined
lizmat perhaps easier from the web interface ? 19:56
codesections [Coke]: I don't have a commit bit on the Raku-Steering-Council repo. Am I supposed to? 20:02
AlexDaniel nige: we also don't have whipupitude for sysadmins/devops 20:08
nige: can't help not being brutally honest
I believe once we accept reality it should be easy to move forward in a useful way 20:09
lizmat wonders who we is 20:10
AlexDaniel nowadays? The council probably 20:11
nige: just as a single example, see docs.julialang.org/en/v1/manual/ru...-programs/ 20:13
nige: in Raku you can either use shell with all the interpolation issues, or try to chain `run` calls 20:14
both are far from giving a proper whipupitude experience, especially when it comes to writing script-like things (which I think sysadmins do often) 20:15
so, really, we're just lacking a solution there, and that is fixable 20:16
lizmat Yet another Rakudo Weekly News hits the Net: rakudoweekly.blog/2020/09/21/2020-...l-results/ 20:17
AlexDaniel: so on the one hand you say: we also don't have whipupitude for sysadmins/devops 20:20
zostay Altai-man: I am basically not doing anything with Raku. Things are too busy right now for O(fun) projects. I'm not sure if or when I will return to any Raku work at this point.
lizmat but otoh you say: so, really, we're just lacking a solution there, and that is fixable
zostay RakuWAPI needs to have an implementation. I don't have time to do that. The only thing close that I know of is Crust, which (last I knew) is more like PSGI in Raku than WAPI. 20:21
lizmat AlexDaniel: what is stopping you from working on such a solution?
AlexDaniel time
lizmat you appear to have time enough to say what's wrong with Raku 20:22
AlexDaniel lack thereof, that is
zostay I have a sort of production project that was built on Smack + Cofra which is a application framework I started building, but both are only half-baked.
lizmat AlexDaniel: seeing the other things you're doing to Raku, it appears to me that you *do* have time
you even appear to have time to be on the RSC
AlexDaniel lizmat: not exactly. I almost didn't even vote in this election – I haven't really been online for the last three days or so 20:23
and a week before that was preparation for the competition
so, I do have some time, yes I use it to express my opinions sometimes :) 20:24
but then, for whateverable I currently unstaged & uncommitted changes
20:24 BenGoldberg left
AlexDaniel so that's the first thing I'll be doing… 20:25
not sure I see your point really. I can't say things unless I'm actually in the process of improving the issues I talk about? 20:26
Geth Raku-Steering-Council/main: a4a8e6e1a8 | niner++ (committed using GitHub Web editor) | nominations/2020/acceptance.txt
Update acceptance.txt
AlexDaniel as far as I know nobody is working on improved shell-like features
currently have* 20:27
lizmat we have a saying in Dutch: "the best helmsmen are ashore" 20:28
codesections At least two people are working on more ergonomic shell features: github.com/gfldex/raku-shell-piping github.com/kjkuan/Shell-DSL 20:32
AlexDaniel you're creating a situation in which it is very difficult to be here and have a realistic view of Raku. If people outside criticize it, then they're framed as trolls. If people who are invloved criticize it, then they should just fix things they don't like and not talk about it.
it can't always be just positives 20:34
codesections: that is fantastic
codesections: we should get one of these into core
20:35 molaf joined, cpan-raku left 20:40 cpan-raku joined, cpan-raku left, cpan-raku joined
Altai-man has coded enough and is satisfied 20:40
AlexDaniel lizmat: and you can see in github.com/gfldex/raku-shell-piping that it is far from being a trivial thing to implement 20:42
20:43 aluaces left
lizmat Did anybody say it was trivial ? 20:43
20:44 stoned75 left, stoned75 joined 20:48 rindolf left
AlexDaniel I don't know what you're trying to say. But “you appear to have time enough to say what's wrong with Raku” implies that the efforts are comparable, which is obviously not the case. 20:49
21:27 Altai-man left 21:48 Xliff joined
Xliff What does this error mean? 21:49
"===> Failed to find dependencies: perl:from<native>"
Oh. Maybe "Missing native libperl!" 21:50
Yep. 21:51
OK...
21:51 Tirifto left
Xliff And how is this one fixed: 21:51
Enabled fetching backends [pswebrequest path curl wget] don't understand www.cpan.org/authors/id/J/JS/JSTOWE....16.tar.gz
You may need to configure one of the following backends, or install its underlying software - [git]
22:04 aborazmeh joined, aborazmeh left, aborazmeh joined 22:19 molaf left 22:32 pecastro left 22:35 zacts joined 22:37 stoned75 left 22:40 zacts left 22:41 zacts joined 22:59 stoned75 joined 23:16 BenGoldberg joined 23:23 guifa left 23:32 zacts left 23:51 aborazmeh left, BenGoldberg 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!