🦋 Welcome to the MAIN() IRC channel of the Raku Programming Language (raku.org). *** CONFERENCE TIME conf.raku.org *** This channel is logged for the purpose of keeping a history about its development | evalbot usage: 'm: say 3;' or /msg camelia m: ... | Log inspection is still being worked out
Set by lizmat on 6 August 2021.
00:02 reportable6 left 00:03 reportable6 joined 00:04 b2gills left
Xliff request-body -> $i is copy where *.defined { 00:06
$i<public-ip> //= qqx«dig +short { $i<public-hostname> } @»; ... }
Why is this giving me "Cannot modify immutable Any type object"? 00:07
00:18 perlbot left, simcop2387 left 00:25 simcop2387 joined 00:27 perlbot joined 00:31 Doc_Holliwood left
raydiak hard to say for sure without knowing what is in that array, but I'd guess you're getting a shallow copy of a container with immutable contents 00:33
m: my @a = Map.new("foo" => Any),; for @a -> $i is copy { $i<foo> //= 1 };
camelia Cannot modify an immutable 'Any' type object
in block <unit> at <tmp> line 1
raydiak one way to fix it would be to just change $i to %i, so whatever it is gets contents copied into a mutable Hash 00:35
m: my @a = Map.new("foo" => Any),; for @a -> %i is copy { %i<foo> //= 1 };
camelia ( no output )
01:13 b2gills joined 01:19 tejr left 02:08 Xliff left 02:09 neshpion joined 02:17 ameameFan joined 02:29 jess joined 03:29 bloatable6 left, sourceable6 left, evalable6 left, releasable6 left, shareable6 left, reportable6 left, linkable6 left, squashable6 left, bisectable6 left, greppable6 left, statisfiable6 left, unicodable6 left, nativecallable6 left, notable6 left, benchable6 left, coverable6 left, quotable6 left, committable6 left, tellable6 left 03:30 benchable6 joined, statisfiable6 joined, quotable6 joined, notable6 joined 03:31 tellable6 joined, greppable6 joined, unicodable6 joined, reportable6 joined, sourceable6 joined, coverable6 joined, committable6 joined, nativecallable6 joined 04:29 bisectable6 joined 04:30 linkable6 joined 04:31 bloatable6 joined, shareable6 joined 04:32 evalable6 joined 05:07 neshpion left 05:30 squashable6 joined 05:42 jbarton joined 05:46 frost joined 06:02 reportable6 left 06:04 reportable6 joined 06:24 Doc_Holliwood joined 06:31 releasable6 joined 07:15 jrjsmrtn left 07:19 jrjsmrtn joined 07:30 ufobat_ joined 07:31 Sgeo left 07:38 dakkar joined 08:07 DarthGandalf left 08:10 DarthGandalf joined 08:26 Doc_Holliwood left 08:27 Doc_Holliwood joined 08:36 Doc_Holliwood left 09:19 ameameFan left 09:31 ameameFan joined 09:40 jbarton left
frost m: my @a = hyper for ^100_000 { .is-prime }; say now -INIT now; 09:52
camelia 0.184259196
frost m: my @a = do for ^100_000 { .is-prime }; say now -INIT now;
camelia 0.039158687
lizmat frost: there's a number of issues at work here 09:58
1. is-prime internally keeps a list of the first N primes, where N is I believe at least 100_000 10:01
so for the cases you're hypering over, the actual amount of work is minimal (as you can see in the time of the unhypered version) 10:02
2. hypering introduces overhead: threads need to be started, results must be cached until they can be produced in the right order, etc. etc.
if the workload is very small, the overhead will drown the actual work being done 10:03
frost lizmat Thanks, I'm just a little strange about the explanation in the document for introducing hyper.
Yes, it does.
lizmat 3. batch size is also important for getting a good balance between the overhead and amount of work 10:04
compare: 10:05
m: my @a = ^5_000_000 .map(*.is-prime); say now -INIT now;
camelia 1.673786247
lizmat m: my @a = ^5_000_000 .hyper.map(*.is-prime); say now -INIT now; 10:06
camelia 6.636019575
lizmat m: my @a = ^5_000_000 .hyper(batch=> 20000).map(*.is-prime); say now -INIT now;
camelia 1.408956095
frost m: my @a = hyper for ^100_000_0 { $_**2 }; say now -INIT now;
camelia 1.698961276
frost m: my @a = do for ^100_000_0 { $_**2 }; say now -INIT now; 10:07
camelia 0.715294501
lizmat $_ ** 2 is still not a lot of work :-)
frost But it is more fast !
Oh, on my machine 10:08
hyper is more faster
Maybe I have less cores :) 10:09
lizmat yeah, it's hardware dependent as well, depending on number of physical cores
virtual cores don't help much if they're doing the same kind of operations at machine level 10:10
10:12 holly_ left
El_Che hi lizmat 10:13
(and everyone else as well of course :) )
lizmat El_Che o/
tbrowder o/ 10:15
Altreus \o 10:16
tbrowder g'day all, my house is up for sale starting from yesterday, gonna mv a bit west where internet access is notoriously bad 10:18
but maybe i can get docs without using the workaround doc site 10:20
speaking of docs, has anyone considered sphinx-doc.org as a possibility for some of our docs woes? 10:23
lizmat tbrowder: I understand StarLink is starting operations soon, maybe that would be a solution for you? 10:24
tbrowder we
lizmat www.starlink.com 10:25
tbrowder StarLink, i'll look, thnx
10:26 Xliff joined
lizmat see also: www.starlink.com/faq 10:28
even in the little villages around here, they are now very quickly installing fiber
they realize that all those people will be using StarLink in a few years if they wouldn't have fiber 10:29
El_Che I hope the cables are water resistent
El_Che ducks
lizmat once they have fiber, they can be pretty sure they will remain customers, and not switch to something like StarLink
tbrowder hm, interesting. i will be in an apartment for next few months so not much flexibility, but it may help at new house site. but fiber would be cool. 10:30
gfldex lizmat: there is one tiny obstacle: arstechnica.com/information-techno...-shortage/
tbrowder att is the "best" service in area we're moving to, and they have some fiber there, but for some reason they haven't expanded their area over the years. 10:33
10:35 ufobat_ left
tbrowder ref raku and jnthn's new work, should any core pod work wait for the big move? 10:37
lizmat gfldex: well, but it is clear that these companies are suddenly in a hurry now :-)
tbrowder: I would suggest any core pod work to be done on the RakuAST branch ? 10:38
El_Che Your Starlink will detect and melt snow that falls directly on it,
it's time for a Skynet joke
tbrowder lizmat, that sounds a little daunting, but a peek wouldn't hurt. 10:39
we don't worry about snow in florida ;-)
lizmat El_Che: I think it just gets warm from the power it uses :-) 10:40
El_Che "all snow faces must die"
lizmat *or* it gets its heat from the space laser on the StarLink satellites
Altreus that sounds related to the "but sometimes" episode of Technology Connections 10:41
tbrowder is there any way to lock one's branch to rakuast on the git web interface for rakudo? 10:42
i guess i could fork it and control it on my copy... 10:44
ok, that works 11:00
11:01 dgl left
tbrowder anyone know where the perl/raku conf being planned for next year in us will be? 11:02
lizmat I am not aware of any date / location yet 11:06
I just hope it will be possible to have an in-person conf again
11:06 dg joined 11:11 Doc_Holliwood joined
tbrowder roger! 11:38
erg, i'm trying to checkout rakuast branch but git can't find it on my end. do i have to do something fancy like "git update-ref"? 11:40
i can see the branch name on github:rakudo/rakudo 11:41
lizmat just doing a "git checkout rakuast" did it for me? 11:42
if not, you will need to find someone with more git foo 11:43
tbrowder i had an updated copy of the master branch so i must be missing a ref somehow
git foo is definitely missing on my end :-( 11:44
Altreus it would be nice to join a corridor track again
El_Che git pull first?
Altreus do a fetch and then check git branch --remote 11:45
in fact git branch --remote | grep rakuast
if it's not there then the obviousest answer is you've got the wrong upstream
El_Che git remote -v 11:52
Altreus dump all the commands and let them pick :D 11:53
11:54 ufobat_ joined 12:02 reportable6 left 12:03 reportable6 joined
El_Che that's my strategy vis-a-vis git :) 12:05
tbrowder Altreus: thanks, it looks like i have the "wrong" master! 12:15
who knew...
El_Che Altreus
tbrowder going back to drawing board with tail between legs... 12:16
El_Che from this day forward he will be known as "The Master of Git"
(of if a new repo "The Main of Git")
tbrowder all hail Altreus
El_Che a git subcmd: git whatwouldAltreusDo 12:17
MasterDuke58 lizmat: is-prime doesn't really keep any list at all. it's so fast to check that it doesn't make sense
lizmat MasterDuke58: I thought the new implementation we use, had a bitmap or somesuch for the first N numbers being prime? 12:19
guess I'm mistaken then :-)
MasterDuke58 well, there is an if statement for the first 4 primes 12:20
there is a table kept github.com/MoarVM/MoarVM/blob/mast...1520-L1533 but it's for doing the calculation, not of the primes themselves 12:21
MasterDuke58 doesn't really understand the math, just copy-pasted the code and it worked and called it a day
lizmat MasterDuke58++ 12:22
Altreus Sadly I have had to accept this role in every company I've worked at 12:23
I didn't even set out to learn it. I just remembered how I solved problems :D
12:29 Doc_Holliwood left 12:30 Doc_Holliwood joined
El_Che Altreus: like this? i.imgflip.com/5jd4hq.jpg 12:32
Altreus yes, my role is to watch everyone else screw up git :D 12:34
El_Che someone has to watch the world burn 12:35
Xliff radiak++ # Thanks for that tip! 12:59
12:59 Xliff left 13:02 Doc_Holliwould joined
tbrowder ok, got the "correct" master again WITH the rakuast branch, whew, thanks again Altreus. 13:03
Altreus yay :) 13:04
distribution is a blessing and a curse
13:06 Doc_Holliwood left
tbrowder i remember same problem before, "git help branch" does not show the "--remote". i think Altreus helped before and i filed a git bug. am i blind? does anyone see "--remote" in "git help branch"? 13:08
Altreus looks like it is actually 'remotes' 13:11
-r --remotes
I'm guessing it either does unique starting strings, or has an undocumented alias 13:12
tbrowder ok, now i see it, way down the page. the online git manual page shows it at the top in the summary, that's what keeps fooling me. 13:15
Altreus tells you what it does when you know it's there, but doesn't tell you that it's there when you want something that does that
tbrowder yepper 13:17
so, how about sphinx for raku docs?
.ask jmerelo how about sphinx-doc.org for raku docs? 13:20
tellable6 tbrowder, I'll pass your message to jmerelo
13:39 ameameFan left 13:52 frost left, Doc_Holliwould left 14:05 Doc_Holliwould joined 14:36 Sgeo joined
raydiak speaking of pod, is someone willing to merge github.com/rakudo/rakudo/pull/4433 ? 14:45
idk what that one ci failure is about, but it's clearly nothing to do with my changes
heh it's so old now you can't even see the log. it was something early on in the build tooling, nothing to do with pod2text 14:48
vrurg raydiak: I tried to initiate re-run for the failed check, but it resist me. Unfortunately, I'd rather not merge it because I barely familiar with Pod internals. 15:08
raydiak: I think the checks are too old for re-runs. You can do a minor commit to initiate new ones. 15:10
MasterDuke58 just force push a no-change commit 15:11
e.g., something like `git commit --reset-author && git push --force` 15:13
Altreus rebase it :P 15:18
MasterDuke58 or yeah, just rebase up to master and then force push 15:21
[Coke] (moving, fiber). I remember when this house got *cable* (not just internet over cable). any moves to a new house, we need fiber or a really good reason why not. :) 15:22
Altreus If you're turning the branch head into a new commit anyway you might as well make the most of it :)
on the plus side there's a house on the market with better internet than ... somewhere else ... 15:23
15:36 evalable6 left, linkable6 left 15:37 Xliff joined, evalable6 joined
raydiak thanks for the tips, I rebased it, ci is running. I won't be around for several hours. I'll ping lizmat on it if nobody merges by the time I'm back, I just avoid hassling her specifically since she already does so much... 15:51
Altreus It's possible you just pinged her anyway ;) 15:54
raydiak oh...right :)
anyway, I'm off...y'all have a good * o/ 15:57
15:57 ufobat_ left 16:19 Xliff left 16:22 Xliff joined 16:35 dakkar left, squashable6 left 16:41 squashable6 joined
Geth setup-raku/dependabot/npm_and_yarn/path-parse-1.0.7: 471d74b3b2 | dependabot[bot]++ (committed using GitHub Web editor) | package-lock.json
Bump path-parse from 1.0.6 to 1.0.7

Bumps [path-parse](github.com/jbgutierrez/path-parse) from 1.0.6 to 1.0.7.
  - [Release notes](github.com/jbgutierrez/path-parse/releases)
  - [Commits](github.com/jbgutierrez/path-parse/...ts/v1.0.7)
... (6 more lines)
setup-raku: dependabot[bot]++ created pull request #14:
Bump path-parse from 1.0.6 to 1.0.7
17:30 ufobat_ joined 17:34 ufobat_ left 17:39 linkable6 joined 18:02 reportable6 left 18:33 xinming joined 19:33 evalable6 left, linkable6 left 19:34 linkable6 joined 19:36 evalable6 joined 20:03 reportable6 joined 20:50 jcallen left 20:53 jcallen joined
Geth setup-raku: 471d74b3b2 | dependabot[bot]++ (committed using GitHub Web editor) | package-lock.json
Bump path-parse from 1.0.6 to 1.0.7

Bumps [path-parse](github.com/jbgutierrez/path-parse) from 1.0.6 to 1.0.7.
  - [Release notes](github.com/jbgutierrez/path-parse/releases)
  - [Commits](github.com/jbgutierrez/path-parse/...ts/v1.0.7)
... (6 more lines)
setup-raku: a52850b116 | (Shoichi Kaji)++ (committed using GitHub Web editor) | package-lock.json
Merge pull request #14 from Raku/dependabot/npm_and_yarn/path-parse-1.0.7

Bump path-parse from 1.0.6 to 1.0.7
setup-raku/dependabot/npm_and_yarn/ws-5.2.3: c6ec7be8e7 | dependabot[bot]++ (committed using GitHub Web editor) | package-lock.json
Bump ws from 5.2.2 to 5.2.3

Bumps [ws](github.com/websockets/ws) from 5.2.2 to 5.2.3.
  - [Release notes](github.com/websockets/ws/releases)
  - [Commits](github.com/websockets/ws/compare/5.2.2...5.2.3)
... (6 more lines)
setup-raku: dependabot[bot]++ created pull request #15:
Bump ws from 5.2.2 to 5.2.3
[Coke] why is the license on setup-raku MIT and not Artistic 2?\ 21:21
ls 21:23
It would be nice if, when running "perl Configure.pl" in rakudo, and generating moar/nqp, if it dropped a config.status in all 3 source dirs, not just rakudo 21:29
vrurg [Coke]: no idea why MoarVM doesn't, but nqp does it. 21:30
[Coke] vrurg++ 21:31
vrurg raku.org is down. 21:32
Back up again. That was quick. :)
gfldex doesnt work for me 21:33
[Coke] I don't see a config.status in my nqp directory
(of course, I'm getting a failure atm) 21:34
gfldex Looks like raku.org drops requests.
rba server is running. seems to be that the connection to the webhosting server is getting down/up a few times. 21:38
21:40 melezhik joined
melezhik Hi all. 21:40
I've added k8s checks to the Bird - github.com/melezhik/bird/blob/mast...ecntconfig
if someone interested
vrurg [Coke]: what kind of failures? Configure.pl only creates config.status on success, if memory serves me right. 21:44
[Coke] doesn't matter, deleted my checkout and started over, no more failures. 21:45
(will report out on if there is a config.status or not.)
vrurg rba: right, looks like it is flapping.
[Coke] (looks like, yes)
is there a way to run a file as if it were a bat file without renaming it? 22:09
Geth Raku-Steering-Council/RSC_code: 524c073303 | (Daniel Sockwell)++ | 3 files
Clarify Raku Steering Council Code

The current Raku Steering Council Code combines details of the initial proposal for adopting the code with actual implementation details. This commit revises the Code to remove the portions that are not relevant post-adoption (but preserves the initial proposal for future reference). It also provides definitions for a few terms that were ... (14 more lines)
22:17 leedo left
Geth Raku-Steering-Council: codesections++ created pull request #44:
Clarify Raku Steering Council Code
[Coke] I'm not an active member? 22:19
22:20 leedo joined
[Coke] ah. I did not vote in the initial election. 22:25
moon-child topic says 'CONFERENCE TIME', but it is no longer conference time! 22:28
22:34 melezhik left, ChanServ sets mode: +o lizmat
lizmat 🦋 Welcome to the MAIN() IRC channel of the Raku Programming Language (raku.org). This channel is logged for the purpose of keeping a history about its development | evalbot usage: 'm: say 3;' or /msg camelia m: ... | Log inspection is still being worked out 22:35
lizmat moon-child++ # tah 22:35
# thanks for the nudge
moon-child :)
22:35 ChanServ sets mode: -o lizmat
moon-child perhaps there should be a 'topic bot' that can schedule changes 22:36
'topicable' is more consistent but sounds more awkward 22:37
22:38 silug left
El_Che weekly: www.debian.org/News/2021/20210812?...um=twitter (death raku community member, used to pkg raku in debian) 23:11
notable6 El_Che, Noted! (weekly)
23:13 Xliff left
Geth setup-raku: c6ec7be8e7 | dependabot[bot]++ (committed using GitHub Web editor) | package-lock.json
Bump ws from 5.2.2 to 5.2.3

Bumps [ws](github.com/websockets/ws) from 5.2.2 to 5.2.3.
  - [Release notes](github.com/websockets/ws/releases)
  - [Commits](github.com/websockets/ws/compare/5.2.2...5.2.3)
... (6 more lines)
setup-raku: 960633c96a | (Shoichi Kaji)++ (committed using GitHub Web editor) | package-lock.json
Merge pull request #15 from Raku/dependabot/npm_and_yarn/ws-5.2.3

Bump ws from 5.2.2 to 5.2.3
23:54 CIAvash left, rba[m] left 23:55 juanfra left, happy-dude left, cnx left, aolko[m] left, AlexDaniel left 23:56 japhb left 23:59 japhb joined