00:28 ggoebel_ left
Geth Blin: hythm7++ created pull request #26:
Add Pakku to Blin Testers
01:50
02:29 HarmtH_ joined 02:30 HarmtH left 04:32 frost-lab joined 04:37 frost-lab left 04:48 frost-lab joined 06:53 patrickb joined 07:16 patrickb left 08:19 domidumont joined 09:12 sena_kun joined 09:24 AlexDaniel` left, unclechu left 09:31 unclechu joined 09:42 AlexDaniel` joined
MasterDuke it would be awesome if a coverage report had filenames, line numbers, etc of the code that caused the rakudo code to run 10:45
e.g., it says that Enumeration.Numeric is covered, but i can't easily find where/why in the spectests 10:46
oh, i can at least easily get the filename 10:49
oh, and i should have at least a guess of the linenumber. can probably hack something into the coverage parser script 10:53
10:54 [Tux] left 10:57 [Tux] joined
lizmat MasterDuke++ 10:57
11:18 [Tux] left 11:22 [Tux] joined 12:05 frost-lab left 13:14 squashable6 left 13:15 squashable6 joined 13:30 domidumont left 14:01 domidumont joined 14:07 domidumont left 14:11 domidumont joined 14:28 domidumont left 14:29 domidumont joined 14:40 domidumont left
releasable6 Next release in ≈4 days and ≈3 hours. 2 blockers. Please log your changes in the ChangeLog: github.com/rakudo/rakudo/wiki/ChangeLog-Draft 15:00
15:09 domidumont joined
[Tux] Rakudo version 2020.09-74-g409737062 - MoarVM version 2020.09-12-g8c41f6692
csv-ip5xs0.805 - 0.832
csv-ip5xs-207.893 - 8.168
csv-parser27.647 - 29.248
csv-test-xs-200.380 - 0.395
test7.567 - 7.715
test-t1.983 - 2.085
test-t --race0.837 - 0.866
test-t-2031.626 - 36.053
test-t-20 --race9.028 - 9.612
15:14
15:34 domidumont left
codesections Ok, after more back-and-forth with the lawyer (sigh) we've settled on adding the following message to the Rakudo welcome message (we can probably revert the capitalization in a few months, but the lawyer wants it for now) 16:15
Welcome to RAKUDO™ v2020.10 implementing
RAKU™ Programming Language v6.d 16:16
(sorry for the duplicate first line) 16:17
(I meant to paste the second)
16:19 sena_kun left 16:20 sena_kun joined
codesections .ask sena_kun do you still think we can add that pretty quickly? We'd like to get a screenshot to the lawyer (for our trademark application) as soon as we can 16:24
tellable6 codesections, I'll pass your message to sena_kun
timotimo why doesn't python have any tm in its startup message 16:55
16:56 ggoebel joined
ggoebel codesections: thank you for your work with the lawyers on trademarks (codesections++) 16:56
timotimo maybe since they didn't put tm in their stuff we can steal their name 17:01
codesections tin 17:03
timotimo: :D I don't envy python's situation for trademarks -- they have a lot of statements about being named after Monty Python, which seems like it might complicate things 17:07
[Coke] why does it have to be capitalized? 17:14
Did they give a reason? 17:15
codesections They want the trademarked term to be visually distinct from the surrounding text 17:16
at first, they wanted it in bold and on its own line 17:17
lizmat is working on it
codesections lizmat++ 17:18
[Coke] ... we can do bold. 17:20
lizmat [Coke] can we?
timotimo indeed
lizmat you mean, use the unicode bold equivalent?
[Coke] 𝐑𝐚𝐤𝐮𝐝𝐨™
timotimo won't be as easy on windows, because not everywhere there's ansi.sys loaded
lizmat wouldn't that hurt the searchability ?
timotimo in fact, probably almost nowhere
[Coke] He already has unicode TM in there. :
:)
fair.
torn between "work everywhere" and "hey, but we have unicode" 17:21
lizmat codesections: thoughts on using 𝐑𝐚𝐤𝐮𝐝𝐨™ vs Rakudo™
?
timotimo ρακUδο 17:22
lizmat m: dd "𝐑𝐚𝐤𝐮𝐝𝐨".contains("Rakudo")
camelia Bool::False
lizmat hmmm... I guess we could have a modifier for that :)
google handles 𝐑𝐚𝐤𝐮𝐝𝐨 fine 17:23
codesections I guess there's also the option of using terminal escape codes for bold…
timotimo you can never escape the terminal 17:24
codesections But I'd probably say to stick with the all-caps version for now. It's what the lawyer is expecting, and we can always change it later 17:25
[Coke] yup, just anticipating the question on the behalf of others. (maybe we can include a little disclaimer in a faq to point people at.) 17:26
again, thanks for dealing with this issue.
codesections happy to do so :D 17:27
18:21 patrickb joined
Geth rakudo: 80a4ada62e | (Elizabeth Mattijsen)++ | 3 files
Add 𝐑𝐚𝐤𝐮𝐝𝐨™ and 𝐑𝐚𝐤𝐮™ welcome message

  - to be shown when starting the REPL
  - to be shown when doing `raku -v`
As part of the effort of trademarking Rakudo and Raku. codesections++ for working with The Perl Foundation on this.
18:29
lizmat $ raku 18:39
Welcome to 𝐑𝐚𝐤𝐮𝐝𝐨™ v2020.09-72-gd713e451a.
Implementing the 𝐑𝐚𝐤𝐮™ programming Language v6.d.
Built on MoarVM version 2020.09-12-g8c41f6692.
argh... missed a L 18:41
Geth rakudo: 4e9178f849 | (Elizabeth Mattijsen)++ | src/Perl6/Compiler.nqp
Use consistent capitalization
18:42
codesections Thanks, lizmat! Should I build from master and grab a screenshot or should we do a point release? I kind of think the point release would be better since that's what is more public ally 18:47
*publicly available 18:48
(and since the version number will be different/shorter) 18:49
tbrowder .tell niner requested changes have been made 18:55
tellable6 tbrowder, I haven't seen niner around, did you mean nine?
lizmat codesections: I could shorten the versions for now, just for the screenshot 18:56
doing a point release for this feels... icky ?
tbrowder .tell nine requested changes have been made
tellable6 tbrowder, I'll pass your message to nine
tbrowder .tell vrurg requested changes have been made
tellable6 tbrowder, I'll pass your message to vrurg
codesections Well, we *are* representing to the trademark office that this message currently appears in a launch screen -- I'm not wild about doing that with a modified message that requires building from souce 18:59
[Coke] when is the next release expected?
(a point build would have to cherry pick that change) 19:00
lizmat codesections: how about the REPL only showing the main parts of the version
and the raku-v version the whole thing
MasterDuke releasable6: status
releasable6 MasterDuke, Next release in ≈3 days and ≈23 hours. 2 blockers. 0 out of 76 commits logged
MasterDuke, Details: gist.github.com/f9d6cd8871bf6012da...d2b7a0854a
codesections lizmat: that seems like a good compromise 19:01
especially if/assuming both messages are in the next release, and that release is pretty soon a 19:03
lizmat $ raku 19:10
Welcome to 𝐑𝐚𝐤𝐮𝐝𝐨™ v2020.09.
Implementing the 𝐑𝐚𝐤𝐮™ programming language v6.d.
Built on MoarVM version 2020.09.
$ raku -v
Welcome to 𝐑𝐚𝐤𝐮𝐝𝐨™ v2020.09-75-g80a4ada62.
Implementing the 𝐑𝐚𝐤𝐮™ programming language v6.d.
Built on MoarVM version 2020.09-12-g8c41f6692.
codesections looks perfect. Tn 19:11
*thanks
(wow, I can't type today!) 19:12
Geth rakudo: bae905ef6a | (Elizabeth Mattijsen)++ | 2 files
Add shorten-versions named and use that for the REPL

So that REPL users won't need to see "strange" version numbers, even if they're not running a released version.
19:12 sena_kun left
codesections I'll build from master and get the screenshot 19:13
lizmat codesections++ 19:14
Geth rakudo/nativecall-fix-reverted: 272d48ad03 | (Elizabeth Mattijsen)++ | 2 files
Revert "Fix NativeCall/C++ to allow double instantiation"

This reverts commit ecaf334c851de35c6e8fbca6a4012bcc4d21246c.
19:30
lizmat .tell sena_kun created a branch with the NativeCall fix reverted, for easier Blin testing: nativecall-fix-reverted 19:31
tellable6 lizmat, I'll pass your message to sena_kun
nine lizmat: thanks for taking care of that! 19:35
tellable6 2020-10-20T18:56:44Z #raku-dev <tbrowder> nine requested changes have been made
nine huh....requested changes to what? 19:36
MasterDuke nine: something with github.com/rakudo/rakudo/pull/3973 perhaps? 19:40
Geth rakudo: 41a76cadd1 | (Tom Browder)++ | src/core.c/IO/Path.pm6
Add fix for issue #3971

File modified:
   src/core.c/IO/Path.pm6
19:56
rakudo: 134adcf953 | (Elizabeth Mattijsen)++ (committed using GitHub Web editor) | src/core.c/IO/Path.pm6
Merge pull request #3973 from tbrowder/clobber

Fix issue #3971: attempt to copy a directory to a file destroys the file!
linkable6 RAKUDO#3971 [open]: github.com/rakudo/rakudo/issues/3971 [BLOCKER] Attempting to copy a directory to an existing file throws but destroys the file
nine Oh....I didn't realize that he only changed the PR description, not the commit message itself :( 19:57
Geth roast: ce40139279 | (Tom Browder)++ | S32-io/copy.t
Add test for issue #3971

This commit:
  + Adds a new subtest
File affected:
   S32-io/copy.t
roast: 7df3c21006 | (Elizabeth Mattijsen)++ (committed using GitHub Web editor) | S32-io/copy.t
Merge pull request #690 from tbrowder/clobber

Add test for issue #3971
lizmat nine: how do you change a commit message after it has been pushed ? 19:59
nine amend, then force push
lizmat well, that introduces a race condition, no ? 20:01
nine ?
MasterDuke you should only force push to branches, not to master
generally
nine yes 20:02
lizmat you could lose commits on master that way ?
nine Oh, no, I misunderstood the question then
on master there's just no good way
lizmat ah, and since it was a PR, ok 20:04
timotimo you can push --force-with-lease or something like that 20:21
it will only actually do the force if what you locally have as the previous version of the remote branch is still actually there on the remote
yes that's the right name
20:28 ggoebel left
nine still not a good idea with github as it might upset repos that are updated on push notifications 20:29
Geth nqp: patrickbkr++ created pull request #677:
Bump MoarVM
20:42
nqp: 4b531cb6b0 | (Patrick Böker)++ | tools/templates/MOAR_REVISION
Bump MoarVM

This pulls in functions to reset STD IO handles. They are required for an upcoming change in Rakudo.
20:46
nqp: 503f2a33ed | (Patrick Böker)++ (committed using GitHub Web editor) | tools/templates/MOAR_REVISION
Merge pull request #677 from patrickbkr/moar-bump

Bump MoarVM
20:57 patrickb left
Geth rakudo/mathematical-succ-pred: f4910fc019 | (Elizabeth Mattijsen)++ | 2 files
Add Mathematical char ranges to .succ / .pred handling
21:06
rakudo: lizmat++ created pull request #3974:
Add Mathematical char ranges to .succ / .pred handling
21:07
lizmat and that concludes my hacking for today& 21:08
tbrowder .tell nine i changed msg too 21:18
tellable6 tbrowder, I'll pass your message to nine
tbrowder .tell nine sorry, i'm mistaken again. i should have cancelled the PR and resubmitted with a better log as i've done in the past 21:23
tellable6 tbrowder, I'll pass your message to nine
tbrowder if someone can revert the two merges i could resubmit them with better commit msgs 21:34
lizmat tbrowder: would that change the code of the PR ? 21:35
tbrowder i would save the code changes and resubmit a new a 21:36
new PR with same changes for rakudo and roast.
the window on git fools me every time, so sorry i think it is the commit msg 21:37
going to save changes now... 21:38
lizmat if there are no code changes, I would be against reverting and resubmitting 21:39
so the commit message was not optimal: it wouldn't be the first :-) 21:40
tbrowder ok, whew, AlexDaniel` had some very pertinent comments, and added the most recent issue discussing the ramifications of the underlying issues. this was a band-aid for the issue found while working on File::Copy so we can try to fix those issues. 22:04
AlexDaniel` it's not a bad thing to have for a user-friendly error message
tbrowder *this pr is a band-aid on a surface scab 22:05
AlexDaniel` basically “hey, you're trying to use `copy` on a directory, you can't do that”
tbrowder exactly, and it should save my poor file
AlexDaniel` so if you change the check to just `self.d and fail X::IO::Copy.new …` then it starts to make sense 22:06
then you can probably add the same to `move` too
but in the end we have to figure out how to get libuv to bail if it opens a directory… and I have no idea how to approach that 22:07
tbrowder well, i first want to get dir with files to copy, but i thought this was too important not to put in core immediately even though other issues can be worked on in the module. 22:08
elf 22:09
AlexDaniel` it is not important, `copy` is inherently a destructive operation for the destination
tbrowder well my concern is for someone who makes a stupid mistake 22:11
as i did 22:12
anyhoo, glad it awakened interest :-D 22:13
AlexDaniel` but it has to be at least two mistakes :) One to assume that `copy` even works for dirs as sources, then also to run it with a destination that is a file
tbrowder true, but i was trying to be exhaustive writing cases when i discovered the clobber 22:15
AlexDaniel` I don't mind telling the user that whatever they're trying to do doesn't make sense, but the key here is source, not the destination
as for making `copy` work on directories… should it? Was there a decision?
tbrowder agree, but we're going toward allowing it, just not yet
no, no decision that i know of. i just said i wanted to develop a module to help investigate os issues. 22:16
AlexDaniel` if anything, I find `copy` to be ambiguous and confusing. Maybe having separate `copyfile` and `copydir` is better 22:18
and `copy` can just be made to call `copyfile` for now, some sort of soft deprecation will do I think 22:19
ah btw the relevant libuv code seems to be here: github.com/libuv/libuv/blob/ed5b42...1136-L1307 22:22
and yeah, that doesn't help. It just opens the source, whatever it is 22:23
tbrowder ok, that's for later. i agree with ugexe that there should be separate io::dir and io::file classes. 22:27
22:40 linkable6 left, evalable6 left 22:41 linkable6 joined 22:43 evalable6 joined 23:33 nebuchadnezzar left 23:34 nebuchadnezzar joined
tbrowder at the moment i am focusing on the copy subroutine and using it to do most if not all of the copying other than file-to-file 23:57