00:08 Doc_Holliwood joined, Doc_Holliwould left 00:15 pecastro left 00:24 BenGoldberg joined 00:48 aborazmeh joined, aborazmeh left, aborazmeh joined 00:58 BenGoldberg left 01:13 grep0r joined 01:24 approaching236 left 02:00 mowcat left 02:02 aborazmeh left 02:12 molaf left 02:25 molaf joined 02:53 BenGoldberg joined 02:56 aborazmeh joined, aborazmeh left, aborazmeh joined 03:04 Manifest0 left 03:05 guifa left 03:06 Manifest0 joined 03:26 melezhik joined
melezhik .tell [Coke] - "why is there a sparrowdo reference in there?" Rakudist uses sparrowdo to run zef test on docker container, that is it 03:27
tellable6 melezhik, I'll pass your message to [Coke]
03:29 melezhik46 joined, melezhik46 is now known as melezhik2
melezhik2 .tell [Coke] as for the error mentioned, I see this "t enough positional arguments; needed at least 2" type of errors on and off for some Rakudist builds, I could even grep all the reports having the one, if it helps 03:30
tellable6 melezhik2, I'll pass your message to [Coke]
03:32 melezhik left 03:38 aborazmeh left 03:49 orinthe left, orinthe joined 03:52 melezhik2 left 03:54 MasterDuke left 04:03 orinthe left, orinthe joined 04:22 Doc_Holliwood left 04:26 BenGoldberg left, Ben_Goldberg joined, Ben_Goldberg is now known as BenGoldberg 04:31 lucasb left 05:05 orinthe4 joined 05:07 orinthe left 05:08 orinthe joined 05:10 orinthe4 left 05:12 Doc_Holliwood joined 06:04 parabolize left 06:09 orinthe5 joined 06:11 orinthe left, orinthe5 is now known as orinthe 06:16 molaf left 06:27 aluaces left 06:28 jmerelo joined, BenGoldberg left, Ben_Goldberg joined 06:29 Ben_Goldberg is now known as BenGoldberg 06:50 aluaces joined 06:52 PavelB left 07:02 Sauvin left 07:07 BenGoldberg left 07:21 sena_kun joined 07:46 bocaneri joined 07:55 bocaneri left, bocaneri joined 07:56 wamba joined 07:59 domidumont joined 08:08 Altai-man joined 08:11 sena_kun left 08:14 Doc_Holliwood left 08:16 abraxxa joined 08:17 Sgeo left, skids left 08:21 abraxxa left 08:22 abraxxa joined
jmerelo There's some trouble in Travis lately. Namely, being asked to migrate to .com 08:33
tellable6 2020-11-02T20:31:58Z #raku <SmokeMachine> jmerelo I’ve just started reading Raku Recipes. Really liking it! Thank you for your words about Red
jmerelo .tell SmokeMachine it's a great tool. Hope you enjoy the rest, too :-)
tellable6 jmerelo, I'll pass your message to SmokeMachine
SmokeMachine I Amy! :) 08:34
I Am! :)
08:34 sjn left
jmerelo I see timotimo and El_Che have noticed the problems with Travis. We still have lots of repos building in travis-ci.org 08:34
Migrating them to travis-ci.com implies "signing for the beta" 08:35
08:35 plant_enjoyer joined
plant_enjoyer hi, docs.raku.org/type/Channel#sub_await says "Since 6.d, it no longer blocks a thread while waiting." - what does it mean exactly? 08:35
if it waits, it blocks - no?
moritz plant_enjoyer: it means it waits, but the thread pool scheduler can still use the thread for other things while it waits 08:36
plant_enjoyer moritz: oh. but as the end user, I don't care about this, do I?
I mean I won't see this unless I do some unsafe stuff 08:37
moritz plant_enjoyer: well, if this weren't the case, you could run into thread starvation if you do many awaits at once
plant_enjoyer moritz: oh right, I can see how. nice :) thanks for explaining 08:38
jmerelo There are not so many repos (at least that I can see) in travis-ci.org, but still we need to organize migration travis-ci.com/search/Raku 08:39
plant_enjoyer do awaits on promises block though? the docs don't say
08:42 dakkar joined
moritz they do block control flow 08:43
plant_enjoyer do they block the thread though? (I'm not sure if these are equivalent) 08:50
09:06 pecastro joined 09:08 sjn joined 09:18 MasterDuke joined 09:24 orinthe4 joined 09:26 orinthe left, orinthe4 is now known as orinthe
moritz I'm not sure, but I'd say no 09:39
lizmat in 6.c await blocks the thread, in 6.d it does not
plant_enjoyer lizmat: might want to add it here too, then docs.raku.org/type/Promise#sub_await 09:40
lizmat doc PR's are very welcome :) 09:41
plant_enjoyer alright, if I remember about it on the weekend 09:44
lizmat ++plant_enjoyer 09:46
plant_enjoyer gets incremented 09:47
09:47 wamba left 09:58 BenGoldberg joined 09:59 wamba joined 10:03 Doc_Holliwood joined
lizmat clickbaits rakudoweekly.blog/2020/11/02/2020-...mma-comma/ 10:17
10:32 BenGoldberg left 10:51 ufobat joined 11:03 _jrjsmrtn left 11:05 __jrjsmrtn__ joined 11:15 Doc_Holliwood left 11:25 orinthe5 joined 11:27 orinthe left 11:28 orinthe joined 11:30 orinthe5 left 11:44 aluaces left 11:59 vZS1 joined 12:00 rindolf joined
vZS1 Hi, I'm missing some fonts on my Debian install for Raku. Is there a set of recommended fonts for Raku? 12:06
tadzik I don't think Raku needs any fonts by itself: where are you missing them? 12:07
oh, you may mean the unicode symbols in error messages and such, or for operators like «»? 12:08
12:09 sena_kun joined
vZS1 In "docs.raku.org/languages/variables#Twigils" for the "The . twigil", I don't have the font for the symbol before the closing bracket in the line starting with "SaySomething.b;" 12:10
language*
12:11 Altai-man left
vZS1 I'm missing the font for pretty much every character that is before "»' 12:12
In those output examples
Viewing page on Firefox browser, if that helps 12:13
jmerelo vZS1: I use Fira Code. Very good coverage. 12:14
vZS1: that should go to the console; every other application will have its own set of fonts, I guess. For instance, comma
MasterDuke the Noto fonts en.wikipedia.org/wiki/Noto_fonts have a lot of the unicode chars 12:15
12:15 Kaiepi left
vZS1 I'm give both a try. Ty 12:16
I'll*
12:17 Kaiepi joined 12:19 Kaeipi joined 12:20 Kaiepi left 12:25 vZS1 left 12:36 vZS1 joined 12:40 Kaeipi left 12:47 schlaftier left 12:48 schlaftier joined 12:50 Kaiepi joined 13:04 aborazmeh joined, aborazmeh left, aborazmeh joined 13:11 jmerelo left 13:18 lucasb joined 13:23 BenGoldberg joined 13:37 molaf joined 13:40 mowcat joined 13:51 mowcat left 13:57 BenGoldberg left 14:02 aborazmeh left 14:07 bocaneri left 14:16 vZS1 left 14:24 jmerelo joined 14:26 wamba left 14:27 aluaces joined 14:44 plant_enjoyer left 14:51 wamba joined 14:54 Doc_Holliwood joined, matiaslina left, CIAvash left, cj left 14:55 matiaslina joined, CIAvash joined, cj joined 14:57 unclechu left, matiaslina left, CIAvash left 14:59 ThaEwat left 15:00 MitarashiDango[m left, kiti_nomad[m] left, JJAtria[m] left, AlexDaniel` left 15:09 vZS1 joined, MilkmanDan left
vZS1 jmerelo: the spacing on Fira is neat. 15:09
15:09 MilkmanDan joined
timotimo github.com/microsoft/cascadia-code - this is kind of a "successor" to fira code 15:11
perryprog There's also JetBrains mono: www.jetbrains.com/lp/mono/ 15:14
15:23 Sgeo joined
vZS1 Noted down. Ty again. 😁👍 15:26
15:26 unclechu joined
timotimo haha, for a moment i read that as "voted down. try again" and was very confused 15:27
vZS1 Woops
15:30 Doc_Holliwood left
perryprog :( 15:32
15:33 matiaslina joined 15:34 CIAvash joined, abraxxa left 15:35 natrys joined
vZS1 Might be worth putting a note in the docs that to get fonts to work with browsers, putting ".font_extension" (.otf, .tff, .woff, etc) in the appropriate "/fonts" directory might be necessary (and to pick them from browser "Settings/Preferences". I had to do that for Tor browser. 15:36
Don't want noobs to miss out on all the fancy stuff. Some of them might not know how to fix it. 15:38
15:38 wamba left, wamba joined
vZS1 Especially when browsing online documentation 15:39
15:41 gdonald joined, MitarashiDango[m joined, JJAtria[m] joined, MilkmanDan left, AlexDaniel` joined 15:43 MilkmanDan joined, parabolize joined 15:47 unclechu left, matiaslina left, CIAvash left, JJAtria[m] left, AlexDaniel` left, MitarashiDango[m left
Zero_Dogg m: my @array = ( { "a" => 1, "b" => 2 } );@array.raku 15:50
camelia ( no output )
Zero_Dogg m: my @array = ( { "a" => 1, "b" => 2 } );say @array.raku 15:51
camelia [:a(1), :b(2)]
MasterDuke m: my @array = ( { "a" => 1, "b" => 2 }, );say @array.raku 15:52
camelia [{:a(1), :b(2)},]
Zero_Dogg Okay, so. I don't quite get why that does what it does. I would expect the array to contain a signle element, a hash with the keys a and b, instead it contains two elements. I know something is just clashing with perl knowledge I'm trying to bring over, but I don't quite get what
interesting. So without the comma, it's flattening it?
timotimo it's the same mechanism that makes [<a b c d e f>] give you an array of 6 elements
assigning to an array will iterate the outermost thing on the RHS to get elements 15:53
Zero_Dogg aha
timotimo m: my @array_one = <hello how are you>; my @array_two = @array_one; dd @array_one, @array_two;
camelia Array @array_one = ["hello", "how", "are", "you"]
Array @array_two = ["hello", "how", "are", "you"]
timotimo same reason the array_two has the same number of elements as array_one, rather than being an array with a single array as its first element
Zero_Dogg right. What was confusing me was that @array.push did the right thing. I think I get it now though, the trailing comma makes it a list of hashes, while leaving it out iterates over a single hash to build array elements 15:54
MasterDuke single arg rule 15:55
15:55 JJAtria[m] joined 16:02 AlexDaniel` joined, unclechu joined, ThaEwat joined, kiti_nomad[m] joined, CIAvash joined, MitarashiDango[m joined, matiaslina joined
timotimo right, push is for "pushing a single thing", whereas you use "append" to push multiple things, in which case the one-argument-rule would kick in again 16:02
m: my @a = 1, 2; @a.push([4, 5]); @a.append(9, 8, 7); @a.append(["a", "b", "c"]); dd @a 16:03
camelia Array @a = [1, 2, [4, 5], 9, 8, 7, "a", "b", "c"]
timotimo here we have append be called with three individual arguments, and called with an array of three elements, those are handled the same thanks to the single argument rule
Zero_Dogg Okay, it makes sense now, I just wasn't aware of it. That's quite handy. 16:05
timotimo aye, it's a thing that tends to Get People
m: my @a; @a[0][0][0][0][0] = "yo"; dd @a
camelia Array @a = [[[[["yo"],],],],]
timotimo ^- thankfully when you have a nested array with single elements, outputting it with the .raku method like dd does it will show you that 16:06
16:08 Altai-man joined 16:10 sena_kun left, sxmx left 16:11 sxmx joined 16:12 skids joined, MilkmanDan left 16:13 MilkmanDan joined 16:15 natrys left 16:16 sxmx left 16:17 Altai-man left, sxmx joined, sena_kun joined 16:19 e left, bocaneri joined 16:31 natrys joined 16:32 orinthe9 joined 16:34 orinthe left, orinthe9 is now known as orinthe 16:45 e joined 16:49 BenGoldberg joined 17:05 Geth_ joined, Geth left 17:06 MilkmanDan left 17:07 MilkmanDan joined 17:13 wamba left 17:20 mowcat joined 17:23 BenGoldberg left 17:30 wamba joined 17:51 Doc_Holliwood joined 17:55 cpan-raku left 17:57 cpan-raku joined, cpan-raku left, cpan-raku joined, dakkar left 17:59 approaching236 joined
Grinnz jmerelo: also note blog.travis-ci.com/2020-11-02-trav...ew-billing 18:07
you may want to start looking into github actions 18:08
jmerelo Grinnz: not a big fan of Github actions. But we're covered in that area, fortunately. 18:10
Grinnz unfortunately, travis seems to be giving up the ghost
jmerelo Grinnz: it says we will have 1000 minutes in a Linux environment. But I'm not totally sure whose plan will that be. If it's per organization, well, we're f*cked. 18:11
If it's per submitter, we're also f*cked, maybe a bit less. 18:12
Grinnz it does say OSS projects can request more, which is not very hopeful
jmerelo Grinnz: got to try anyway
Grinnz: do you understand 10k credits is per month? Or that's all you've got? 18:16
Grinnz yes i believe it's per month
jmerelo Grinnz: these are all the repos in Raku travis-ci.com/github/Raku Just a few of them are really active. Every build in Raku/doc is 7 minutes. 18:19
Grinnz: if that's the case, I think we're safe for now.
Grinnz: Uf, no, we're f*cked travis-ci.com/github/Raku?tab=insights it's 2k minutes per month. 18:20
Grinnz :/
jmerelo Grinnz: requesting 18:21
18:29 masak left, masak joined, masak is now known as Guest86785
vZS1 I wouldn't tie in too much with GitHub if I were you. Look what they did to youtube-dl. 18:30
jmerelo Someone should do the same for the Rakudo and MoarVM organizations, although they're probably far away from the limit and they're using alternatives anyway.
18:30 BinGOs left
jmerelo vZS1: that's also an issue. 18:30
Grinnz if you don't want to be dependent on an org for CI you have to host your own
jmerelo We're good with raku-community-modules. We've spen... 5 mins in the last month.
vZS1 GitHub is no longer what it used to be after the Microsoft takeover
18:30 domidumont left
Grinnz such as circle ci 18:31
jmerelo vZS1: Github actions is certainly something with great potential. But their support for it is totally dismal.
Grinnz: we do use Circle CI for documentation. We regularly run out of minutes there. 18:32
vZS1 I use Got hooks on my own hosts for everything. Relying on these companies for critical infrastructure is a bad idea. 18:34
Self-hoated GitLab is also a good solution that requires less work.
Git hooks* 18:35
Self-hosted*
GitLab has declarative workflows for build and test automation. 18:37
You'll save money in the long run as well. 18:38
jmerelo But we would need to host everything at GitLab. Move issues. PR history. Not a good idea. 18:40
self-hosting is not sustainable. At the end of the day, somebody's got to pay the electric bill.
18:41 ufobat left
jmerelo Raku is not a company or a thing. We need to rely on those free resources, those that need the least amount of support. 18:41
18:41 guifa joined
Grinnz right 18:42
jmerelo Travis a pretty good deal anyway. I've requested some more credits; Raku is a good case for getting more of them. Don't think it will be a problem.
18:43 bingos joined
jmerelo We need to spin off some tests, anyway, to other services. The less we spend in a single place, the better. 18:43
vZS1 Do what you think works for Raku but you already see the issues with migration from these solutions. They coax users towards vendor lock-in.
PR being a great example 18:44
And other "integrations"
Then something like the youtube-dl disaster happens and a project is caught with its pants down. 18:46
18:46 bingos is now known as BinGOs
Grinnz you either host things yourself (not possible in cases like this as mentioned) or you lock into *something* - all you can do is account for the possibility that the service will be lost at some point 18:47
fortunately for the git repo itself the backup plan is very easy
jmerelo Grinnz: well, issues and PRs can be backed up if needed too, via API. We already retrieved the tickets in RT and uploaded them to GitHub 18:50
vZS1 It is perfectly possible to self-host a project of Raku's scale. And it's not as costly as some would lead to believe. New CPUs like Ryzen consume very little wattage
jmerelo But I don't see self-hosting as a viable option. Even if, let's say, Primark came and said "Here's your hosting, or unlimited cloud credit", you would still be have problems. 18:51
Person needed to maintain them, upgrades, security... And you would need to work with Jenkins or GitLab or whatever, which would you lock in in yet another way.
vZS1 Depends on how much of a priority self-sufficiency is. 18:52
jmerelo vZS1: priority is using human resources wisely. Migrating to a new infrastructure (self-hosted or whatever) and then needing human resources to tend to it, well, it's exactly the opposite of self-sufficient. 18:53
There's no such thing as self-hosted with 0 maintenance. Even hosted elsewhere needs a certain amount of maintenance.
vZS1 Self-sufficiency is not efficiency 18:54
Two different things
jmerelo I mean, push comes to shove, we might need to move to GitLab. But someone would have to make a pretty big case (and put a good amount of resources on the table) for that to happen.
And I'm not even talking about self-hosted gitlab... 18:55
vZS1 Just remember you get all these things free now as long as it's profitable for these companies to provide you these things. 18:56
Having a migration solution ready is important. I love the language you provide. That's why I don't want to see it get a wrench thrown in its supporting infrastructure. 18:58
jmerelo Don't think it would be good optics for them to shut down an open source outfit. Even if it's a small one. 18:59
vZS1: we already had a wrench thrown throughout all the 1 single computer offered by moritz that went bust
vZS1: we survived through it, and not there's one person in charge of infra. There's no way out of that; we need infra to host websites. You can't run a language from GitHub pages. 19:00
vZS1 Don't you have a mirror host? 19:01
jmerelo Now infra is pretty stable, but rba++ needs to devote some time and money to that. But we needed the wrench. We would have happily continued there.
vZS1: there's nothing in production that's not in a repo. Precisely that's one of the differences with current infra, before there were lots of hacks in production that were impossible to reproduce. 19:02
vZS1: they were nice hacks, and used a lot of ingenuity. But you can't reproduce ingenuity. That's why today's infra is a bit more robust, and (mostly) reproducible.
vZS1 But that's not a shortcoming of running your own infrastructure. Reproducibility can be done on premise or cloud 19:03
Look at things like Ansible, Docker, etc.
jmerelo No, it's not. My point is that we did it because we had no other way out. CI is a different thing altogether. We have CI scripts running all over the place.
Travis is the biggest bunch, probably. But there are all kinds of them. Moving all of them to... whatever: pretty big deal. Moving Travis to whatever... well, just a big deal. 19:04
While it's probably a good idea to phase it out in favor of... Anything else, I wouldn't ask for everyone in the 5? organizations to do it now. 19:05
Let me see: there's (legacy) perl6, but pretty much nothing is running there now. Raku, with docs being the biggest thing. raku-community-modules (but just 6 minutes spent last month, not a lot is happening there). 19:06
Then Rakudo and MoarVM
vZS1 CI is a bit tricky because unless you built it yourself to be easy to move right at the start it's always going to be painful to try something else later. No easy answer for that. 19:07
jmerelo So best thing is to channel our best puss-in-boots-big-eyes face and ask for more minutes from Travis. Concurrently, try and see what's done there that can be done elsewhere, so that we don't need to rely so much on them.
I mean, those are just my 2€cents. 19:08
vZS1 How many minutes of CI do you use per month? 19:11
19:13 lichtkind joined 19:15 mid_home left
Zero_Dogg There's also gitlab, their CI is quite good, and offers 50 000 minutes per month for FOSS-projects 19:17
19:17 camelCaser left
Zero_Dogg they've got some hooks for working with github repos too, I think 19:18
jmerelo vZS1: 2250 in the case of the Raku organization. Most are for docs, but there are other projects. Raku/documentable, for instance.
Zero_Dogg: really?
vZS1 The CI market is incredibly saturated right now. There's no shortage of YAML-based alternatives.
jmerelo Hum, didn't know that about.gitlab.com/solutions/github/ thanks, Zero_Dogg 19:19
Zero_Dogg :) 19:20
jmerelo Zero_Dogg: needs a silver-premium plan about.gitlab.com/pricing/ Do you know where we can apply for that?
Zero_Dogg jmerelo: FOSS projects get gold for free - about.gitlab.com/solutions/open-source/join/ 19:21
vZS1 I think you get some free credits for Google cloud as well if you try the kubernetes integration for GitLab. I don't know if that's still around or not. 19:22
Last time I checked was a few months ago 19:23
19:23 camelCaser joined
jmerelo Zero_Dogg: thanks! Applying 19:24
Zero_Dogg: OK, applied on behalf of the whole community. 19:31
vZS1 You could look into IPFS as well if you would like to make it simple for people to mirror the webpages Raku has.
jmerelo Still waiting for the answer from Travis 19:32
vZS1: all pages are in GitHub. For instance, documentation is at rakudocs.github.io. Most of them are static, too.
perryprog Worth noting that Gitlab reduced the free tier of CI/CD to have 400 minutes from 2000 minutes, not sure if this applies here
about.gitlab.com/releases/2020/09/...ree-users/
19:33 orinthe1 joined 19:34 orinthe left
vZS1 jmerelo: thanks for the link. 19:35
19:36 orinthe joined
vZS1 Be a bit careful with GitLab. Last time I used it I couldn't "nest" variables properly. Would throw some random behaviour at me. Don't know if they fixed that yet or not. 19:37
Let me try to find the issue 19:38
Sec
19:38 orinthe1 left
jmerelo perryprog: at the end of the day, every one of them are not charities. They're private companies. Sticking to just one of them is not a good long-term strategy 19:39
vZS1 gitlab.com/gitlab-org/gitlab-runne...ssues/1809
jmerelo perryprog: but for the time being we are not using GitLab, except for Rakudo Star, I think. 19:40
vZS1 It's been open for 4 years now 😂
And I like how they call it a "feature request" instead of a bug. Lol
jmerelo Well, it been a nice while, but it's dinner time around here. Don't stay up late while they count the ballots. See you tomorrow! 19:41
vZS1 Have a good night.
19:42 jmerelo left 19:59 lichtkind left
Zero_Dogg perryprog: doesn't apply here, it's for private projects/repos. Public ones get 50 000 minutes. 20:06
20:08 MasterDuke left, Altai-man joined 20:11 sena_kun left 20:14 BenGoldberg joined 20:18 MasterDuke joined 20:22 mid_home joined
Geth_ ¦ doc: Altai-man self-assigned "Test" module is documented as a type github.com/Raku/doc/issues/3686 20:23
20:34 japhb left 20:39 approaching236 left 20:48 BenGoldberg left 21:19 natrys left 21:20 Black_Ribbon joined 21:43 MasterDuke left 21:47 vrurg left 21:48 vrurg joined 21:52 rir_ left 22:00 raku-bridge joined 22:04 chloekek joined 22:20 mowcat left 22:23 sena_kun joined 22:24 Altai-man left 22:38 rindolf left 22:47 aluaces left 22:52 Altai-man joined 22:55 sena_kun left 23:06 japhb joined 23:09 sena_kun joined 23:11 Altai-man left 23:16 approaching236 joined 23:23 PavelB joined 23:26 Xliff joined 23:28 BenGoldberg joined 23:33 wamba left 23:43 AlexDaniel joined, AlexDaniel left, AlexDaniel joined
AlexDaniel tony-o: there's a problem-solving PR waiting for your review, can you please check? 23:44
23:45 sena_kun left 23:52 chloekek left