»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'p6: say 3;' or /msg camelia p6: ... | irclog: irc.perl6.org or colabti.org/irclogger/irclogger_log/perl6 | UTF-8 is our friend! 🦋
Set by Zoffix on 25 July 2018.
Xliff m: "ec2.amazonaws.com?Action=DescribeR...013-10-15" eq "ec2.amazonaws.com?Action=DescribeR...013-10-15" 00:01
camelia WARNINGS for <tmp>:
Useless use of "eq" in expression "\"ec2.amazonaws.com?Action=DescribeR...13-10-15\" eq \"ec2.amazonaws.com?Action=DescribeR...3-10-15\"" in sink context (line 1)
Xliff m: say "ec2.amazonaws.com?Action=DescribeR...013-10-15" eq "ec2.amazonaws.com?Action=DescribeR...013-10-15"
camelia True
Xliff .tell jonathanstowe I'd like to ask you some questions about XML::Class when you have some free time. PM me your email? Thanks. 01:31
yoleaux Xliff: I'll pass your message to jonathanstowe.
Xliff .seen jonathanstowe
yoleaux I haven't seen jonathanstowe around.
Xliff .seen jstoew
yoleaux I haven't seen jstoew around.
Xliff .seen jstowe
yoleaux I haven't seen jstowe around.
Xliff .seen jonathan
yoleaux I saw Jonathan 23 Dec 2015 01:14Z in ##hplusroadmap: <Jonathan> hah
Xliff Rut-roh 01:32
Xliff m: say ' 03:21
camelia 5===SORRY!5=== Error while compiling <tmp>
Unable to parse expression in single quotes; couldn't find final "'" (corresponding starter was at line 1)
at <tmp>:1
------> 3say '7⏏5<EOL>
expecting any of:
argument list
03:22
Xliff m: say 'Option ' «~» (1..11) 03:22
camelia (Option 1 Option 2 Option 3 Option 4 Option 5 Option 6 Option 7 Option 8 Option 9 Option 10 Option 11)
Xliff \o/ 04:55
\o\
/o/
\o\
Xliff is happy. 04:55
Xliff now has a EC2 testbed for p6-GTKPlus. 04:56
holyghost hi 06:48
moritz ho 07:24
El_Che moritz: you kiss your mother with that mouth? 07:49
Xliff lol
timotimo .seen rabidgravy 07:50
yoleaux I saw RabidGravy 17 Dec 2017 01:40Z in #perl6: <RabidGravy> that's perl6advent.wordpress.com/2017/12/...ier-title/ if you want to get in early
timotimo Xliff: ^
Xliff timotimo: ? 07:51
Xliff timotimo: Been up a while, so you may have to elaborate a lil. 07:51
jast possibly related to the person you wanted to reach earlier? I don't know any of the names unfortunately :) 07:52
a quick web search supports that hypothesis :) 07:53
timotimo aye 07:54
you probably won't reach 'em via irc
Xliff ooo 07:55
Xliff OK. Busy with $dayjob. That's what I got out of that after a quick (tired) skim. 07:57
I bugged about it in XML::Class so maybe there will be a response. timotimo++ 07:58
holyghost Xliff, are you going to take a look at PaganVisions2 ? 08:15
later on I mean
Xliff holyghost: When I have the time. Given my workload it probably will not be this week. 08:16
I am extremely busy, right now.
holyghost ok, take care 08:17
Xliff I will try to do something, but I can't make any promises. :(
holyghost ok, is your $dayjob Gtk based ?
just interested
Xliff Um. Yeah (?)
I'm trying to work on it so that it is.
holyghost ok, I need to fix Mathx::Stat and Bayes::Learn myself still, so take your time 08:18
Xliff Are you on a machine that can handle X11? I spun up another VM to demo my GTK projects.
holyghost The issues from JJ 08:18
Xliff Unfortunately I can't get Clutter to work from there since it relies on OGL and GLX.
At least not yet, anyways.
Is your last public key good for YOU? 08:19
holyghost yes
You're doing OpenGL ?
Xliff If so, I can set it up so you can take a look.
Well. Clutter is based on OpenGL because COGL is.
GNOME's latest desktop is Clutter-based
holyghost I read the OpenGL primer once, if you like I'll return the favor if it's on github for example 08:20
It's probably not OSS, I guess
Xliff It's on Githgub 08:21
Xliff *Github 08:21
holyghost can you give me a link ?
Xliff I'm spinning up the VM.
holyghost As long as I can hack ...
Xliff github.com/Xliff/p6-Clutter 08:22
I don't mind you hacking. Send PRs
Heh. I can start EC2 instances from the CLI using Perl6, now. 8-D
holyghost Looks like serious work 08:23
Xliff Yup
holyghost I am glad for the GL things, I think vanilla GL is too easy to program in 08:24
for the GL things in perl6 I mean
but I'm not too fond of the procedures in common OpenGL (Mesa for example)
The extensions I need to look into TBH 08:25
holyghost Anyway, send me a mail if you PR or issue anything, later on I'll take a look at the github Xliff dir 08:26
then
Xliff OKl. One sec.
Taking this to PM 08:28
holyghost the second I sent you 08:30
9Tas3Ni86QcUTKXpAyOUIn7x6u5+nG8V6b5Ommk00axbr3BUWUwQnS0vEdgyIeGMU4NHN0Sin1iu9+Z1 08:33
v5cO5uwQJBxgu3wc2jsOY7y/jb/dcu7+WxrwXd0YZUU45YSxVsUzO5oNslKQGaVIAc8Acy/UwlowNdFB
Y5sxx8r0hPlrIfOfNaoqEVOoqw1c157jNT7LA/rVViXjDy7IbdkCUU8OdAUKoeCoLM9fb7dANVpV8RJK
Ci+QHFMhddmALtjn31E+3bsAkNuEK64SiFutXqxMVYqGn7SLGv4ljVY1VZVd goon@daemon
sorry 08:34
BitchX is somewhat bad for PMs
holyghost ok 08:41
woolfy holyghost : Could you please take your discussion with Xliff to a private chat on irc or elsewhere. You discussion has little to do with the development of Perl 6. Thank you. 08:45
holyghost : Your discussion is a private one. If I want to privately discuss things with others, I will go to the private chat options. And I know, sometimes I can be chatty, but I will stop when other people are bothered. 08:46
holyghost woolfy : we're already doing that :-) 08:47
woolfy holyghost : Thank you.
holyghost woolfy, terminal irc isn't as easy as it sounds, then again that's everyone's option 08:47
Xliff woolfy: Actually... it does. 08:48
But that's beside the point.
Your point is well taken.
woolfy holyghost : That's not a good reason to not do it.
holyghost woolfy, I know
woolfy Xliff : your comments are actually interesting to read... 08:49
;-)
Xliff
.oO( *zing* )
woolfy: I've actually gotten Amazon to talk to a cro-based client for EC2 08:50
Early days, tho.
holyghost anyway, me and Xliff are going to write perl6 games in Cairo, HTH
Xliff holyghost: ALSO early days... :)
holyghost elder days :-)
woolfy Good luck! 08:52
holyghost thanks
woolfy Xliff : Nice, re: Amazon to talk to a cro-based client. I guess that will lead to a nice blog. Because that would be good Perl 6 promotion.
Question to #perl6 regarding the webserver on which perl6.org was located and that died. I am missing marketing.perl6.org/ and there was a lot of useful stuff on there. 08:53
woolfy Re marketing.perl6.org/ : I hope a backup was available and could at some point be restored. Some of my stuff was on there... 08:53
and I could try to gather all of it, but that will be many hours of work, and it would just be my stuff. While putting a full backup back, would probably be much less time. 08:54
At the moment, marketing.perl6.org/ says "Perl 6 Documentation", which is nice information, but not related to marketing. 08:55
holyghost *
Xliff Yeah. The XML deserialization needs some work though, hence my need to talk to RabidGravy.
I'm having an issue with XML::Class and class reuse. 08:56
woolfy: You might need to talk to AlexDaniel or timotimo about getting those backups. 08:57
woolfy Xliff : maybe. But I am placing it here, because they read this as well. I hope.
rba woolfy: Moritz was able to recover some stuff from the servers hard drive. I thing there was no backup.
rba think 08:57
holyghost woolfy, do download tarballs from time to time
woolfy rba That would make me very sad. And a bit angry. 08:58
holyghost ; that is not helpful.
holyghost why not ?
woolfy Sorry, I have to go. Checkout of hotel room in minutes...
holyghost ok
lizmat afk for the rest of the day& 08:59
Xliff o7 libmat 09:00
holyghost lizmat : /nick lizmat[afk] for example, it does not hurt on _some_ irc servers
rba is understanding woolfys feelings.
holyghost I would'nt know here
rba woolfy: we will add backup to further hostings 09:01
holyghost rba, can she backup with a downloadd of a tarball now ? 09:03
holyghost shell access then tar zcvf etc 09:03
Xliff OK. I've got to get some sleep. Good night! 09:04
holyghost Good night, Xliff 09:04
rba holyghost: I think the only way is to ask moritz to search the broken hard drives. 09:05
holyghost rba, I mean in the future
rba holyghost: sure, yet I like the files backuped from the hard drives too. 09:06
holyghost indeed
cron/crontab job then download with a shell, TMTOWTDI 09:07
rba As most websites have a github repo as source: Is this the one for marketing.perl6.org? github.com/perl6/marketing
timotimo yeah, the web/ folder has the web app 09:10
it does have the rakudo.org website readme in it 09:11
but the app itself is the right one, i'm sure
antoniogamiz if I have a method with the name that an attribute, it would be a getter? 10:47
jnthn Yes, it will suppress auto-generation of the default accessor 10:48
yoleaux 4 Jun 2019 21:45Z <vrurg> jnthn: when you get a couple of minutes for github.com/perl6/problem-solving/issues/3 I would like to ask you to share you opinion on backward compatibility issue.
antoniogamiz jnthn: ty :) 10:49
holyghost antoniogamiz : AFAIK if you use $.listptr, it has a get and set method, it's alike C#, then there's $!listptr 11:25
holyghost or python 11:27
antoniogamiz holyghost: ok. ty :) 11:31
timotimo oh, huh, JSON::Unmarshal doesn't compile, because it has "use v6.*" 11:40
m: use v6.*
camelia ( no output )
timotimo m: use v6.*; 11:41
camelia ( no output )
timotimo this doesn't error
m: use v6.*; use Test
camelia ( no output )
timotimo what the
is it because camelia is back in time a little?
c: HEAD use v6.*; use Test 11:42
committable6 timotimo, ¦HEAD(bd2ceef): «04===SORRY!04=== Error while compiling /tmp/nhGrVSVG_i␤No compiler available for Perl v6.*␤at /tmp/nhGrVSVG_i:1␤------> 03use v6.*08⏏04; use Test␤ «exit code = 1»»
timotimo c: 2019.03 use v6.*; use Test
committable6 timotimo, ¦2019.03: «»
timotimo c: releases use v6.*; use Test
committable6 timotimo, gist.github.com/2fddf45213626cfcd1...aff06e4660
timotimo bisectable6: use v6.*; use Test 11:43
bisectable6 timotimo, Bisecting by exit code (old=2015.12 new=bd2ceef). Old exit code: 0
timotimo, bisect log: gist.github.com/36fb7c9730e9573612...9bc00369e2
timotimo, (2019-05-20) github.com/rakudo/rakudo/commit/2d...a525e476a4
timotimo greppable6: use v6.* 11:44
maybe i actually wanted to put quotes around that 11:45
greppable6 timotimo, 9515 lines, 1060 modules: gist.github.com/8d787abcd72bf59875...6a8b6e60e3 11:46
timotimo vrurg: did you intend for "use v6.*" to no longer work? 11:49
greppable6: "use v6.*" 11:50
greppable6 timotimo, 14 lines, 11 modules: gist.github.com/70c727bca512bb9a21...af3e93f774
pmurias hi 11:51
timotimo damn it! :) :)
greppable6: "use v6\.\*"
greppable6 timotimo, Found nothing!
timotimo greppable6: use\s+v6\.\*
greppable6 timotimo, 6 lines, 3 modules: gist.github.com/b6fbcf52a60c17f361...2cdde8e6c5
timotimo vrurg: ^- these modules use "use v6.*" 11:52
timotimo hey tyil, is geth down? 13:16
tyil deployment says it should be up, but I can restart it
timotimo mhh, we're gonna need a watchdog or something 13:17
ideally figure out how it can disconnect without noticing
tyil kubernetes should be that watchdog
tyil might have been in a netsplit 13:17
timotimo well, kubernetes would have to check irc if geth is present :D
tyil one could add cro and expose a healthcheck over http
which could report it's current connection status 13:18
if it fails the healthcheck, k8s will kill it off and respawn it
timotimo would be possible, yeah
tyil I see a geth
timotimo yeah, geth is back \o/
tyil I still need to figure out why systemd-journald just dies and refuses to start if a server is using a lot of memory (which happens pretty often using perl 6, sadly) 13:19
that's the main cause of nodes dying and taking deployments with them
timotimo wow, journald dies? 13:20
tyil systemd is a curse in itself
timotimo could be that it's taking the core dump and that takes very long?
that's certainly a thing journald does by (i imagine) default
tyil idk, I still think any logging (and by extension, init system) that dies this easily and breaks itself is horrible
timotimo so a core size ulimit could help
tyil I can set one
if you think itll be beneficial 13:21
timotimo like, does it actually die? as in, the process disappears?!
tyil yes, that's the last thing I see in the server's consoles if they die
a couple OOMKiller messagse saying a Perl 6/moar process was using too much memory and gets killed
and then journald saying it's failing to start for a couple times 13:22
and then emptiness
b 28
I preferably would run with a sane init, but k8s doesn't really play nice with those
timotimo i wonder if dmesg gives any more interesting output after the oom happens 13:24
tyil by the time I notice all my perl stuff dying I can't ssh in anymore 13:24
I could try the web console from the provider next time, and check if that still works 13:25
maybe dmesg has something interesting
in case of low memory I'd prefer it to just kill a process and respawn it on another node
not completely kill the node, which causes all deployments to be scheduled to another node, which will then run out of memory, which then dies etc...
(in all fairness, I'm not a master of kubernetes either, which is why I'm looking for people in here that want to assist me for the perl deployments in my cluster) 13:26
vrurg timotimo: I didn't even know such construct exists. It's not documented. 13:30
timotimo can we make it work again, but give a deprecation warning? 13:31
bisectable6: if 1 { .perl.say } 13:34
bisectable6 timotimo, On both starting points (old=2015.12 new=bd2ceef) the exit code is 0 and the output is identical as well
timotimo, Output on both points: «Any␤»
timotimo i didn't realize that 13:35
vrurg timotimo: what's the meaning? `use v6;'? 13:36
BTW, Version can't parse it. I wonder how did it work before. 13:38
timotimo m: say v6
camelia v6
timotimo m: say (v6).perl
camelia v6
timotimo m: say (v6.*).perl
camelia v6.*
timotimo how do you mean Version can't parse it?
vrurg Sorry, it can. It was a mistype in my code. 13:42
timotimo OK :)
well, i'm glad to be a rubber duck
ugexe version can't parse what? 13:47
vrurg timotimo: anyway, what is the meaning of it? v6.* eq v6?
ugexe n/m
should have read 3 more lines
timotimo i guess
ugexe: is use v6.* equivalent to use v6?
ugexe v6.* contains v6 13:49
v6 is equivalent to v6 followed by any number of 0s
timotimo you don't mean that as in "string-contains", right?
ugexe v6.* is a range of v6 .. v6.9999999~ 13:50
well, not totally true, v6.a .. v6.999999~
timotimo v6 comes after that? 13:51
ugexe v6 is v6.0
timotimo ok
ugexe v6.a comes before v6.0
timotimo you said that up there already! sorry
ugexe hmmm 13:53
m: use v6.*; say $*PERL
camelia Perl 6 (6.c)
ugexe m: say $*PERL
camelia Perl 6 (6.d)
ugexe thats not right
vrurg What sense does it make in 'use v6.*;' then? 13:53
ugexe because its a version, so why wouldnt you be able to do so? 13:54
ugexe this is why i said `use Perl:ver<6.*>` syntax would have been better 13:54
its still an odd thing to do, but it should be more apparent what is going on 13:55
anyway v6.* should take the highest available version and its not
ugexe so anyone literally using `use v6.*;` is not going to be getting the behavior they think 13:56
timotimo should it also get a .PREVIEW? 13:57
in a regular version object sense it should
ugexe hard to say. ultimately the final .PREVIEW version would `emulates` the actual finalized spec version 13:58
er i misunderstood
you are saying v6.* if choosing the newest version it could be argued it should choose .PREVIEW in some cases 13:59
timotimo aye
but .PREVIEW is there to "opt in to features that may change before the final spec version appears" 14:00
vrurg My logic says that .PREVIEW must be explicitly stated.
timotimo and v6.* isn't really a clear "yeah i want that"
ugexe yeah no one wants .PREVIEW when they say 6.*, but module loading in general works that way
however
module loading does not always choose the highest version. it chooses the highest version in the first repository containing any match at all 14:01
timotimo huh, at first glance that sounds kind of odd
ugexe yes there are pros and cons for it
i kind of like the "include results from everything" approach but that isn't the performance path 14:02
timotimo can a repo object implement "include everything" semantics? i'd assume it could 14:02
ugexe yes now that .candidates exists 14:03
vrurg timotimo: as I have to do my primary work now, not always able to follow. May I ask you to create a ticket for this?
timotimo tyil: you're also running synopsebot? that seems to also be down ATM 14:05
tyil: if it's okay with you, maybe i'll take that offer of access to your k8s
tyil that's very much OK with me, I'd rather not be a single point of failure :>
timotimo aye, i'm a failure, so add me and it'll be better! :) 14:06
tyil :D
I'm restarting synopse, cpan-p6 and the discord bot, that should be all remaining bots :p 14:07
Geth doc: 6f17c5494b | (JJ Merelo)++ | doc/Type/Backtrace.pod6
Revises new definition, enhances description #2809
14:22
synopsebot Link: doc.perl6.org/type/Backtrace
vrurg m: my $v = Version.new("6.*"); say $v.ACCEPTS("6.d") 14:32
camelia False
vrurg ugexe: ^ is it correct?
ugexe well you are trying to accept a string, not a version 14:33
m: my $v = Version.new("6.*"); say $v.ACCEPTS(Version.new("6.d")) 14:34
camelia True
vrurg Ah, sure... Though I expected it to coerce.
vrurg m: my $v = Version.new("6.*"); say "6.d" ~~ $v 14:35
camelia False
ugexe m: say v6.d ~~ v6.c+
camelia True
ugexe m: say v6.d ~~ v6.* 14:36
camelia True
ugexe m: say v6.d ~~ v6.e+
camelia False
ugexe why would it coerce? maybe you are doing string comparison 14:37
vrurg I'm gonna need it in nqp, there it's a string used. Ok, anyway.
ugexe string ordering is different than number ordering 14:38
vrurg It's a Version on rhs, that's why. 14:38
ugexe does anything else magically coerce when using smart match? 14:39
vrurg m: "1.2" ~~ 1.2
camelia ( no output )
vrurg m: say "1.2" ~~ 1.2
camelia True
vrurg For example 14:40
ugexe how do you compare 1.a to 1.1
you have to choose some semantics
vrurg Version-wise I'd say they don't match – different versions. If that's what you're asking. 14:41
ugexe right. now how do you sort it
version is entirely a sort order mechanism 14:42
vrurg m: Version.new("1.a") < Version.new("1.1") 14:42
camelia WARNINGS for <tmp>:
Useless use of "<" in expression ".new(\"1.a\") < Version.new(\"1.1\")" in sink context (line 1)
vrurg m: say Version.new("1.a") < Version.new("1.1") 14:43
camelia True
vrurg m: say Version.new("1.a") > Version.new("1.1")
camelia False
ugexe missing the point
m: say Version.new("1.a") cmp Version.new("1.1")
camelia Less
ugexe m: say "1.a" cmp "1.1" 14:43
camelia More
ugexe by coercing you take away the choice 14:44
both are valid
vrurg Ok, no problem. The semantics of smarmatch is "rhs accepts lhs". So, Str ~~ Version -> Version ~~ Version; OTOH Version ~~ Str -> Str ~~ Str 14:45
ugexe what if i want to know if it accepts an actual string type and not a version? 14:46
i would even ask: is there a single instance of ACCEPTS coercing its argument? 14:47
vrurg Any practical use for it? A parameter with any value of any type? 14:48
ugexe m: Version.new("1.a") ~~ Version.new("1..a") # another example 14:49
camelia ( no output )
ugexe m: say Version.new("1.a") ~~ Version.new("1..a") # another example
camelia True
vrurg ugexe: Does this line in Numeric counts? '(try my \numeric = a.Numeric).defined'
ugexe i dunno, but just seeing that screams its not something to be emulated 14:50
vrurg Ok, it's not really important. I like that "1.2" ~~ 1.2 – it follows DWIM well enough. I would extend this to Str vs Version too. But won't insist. We have bigger problems to take care of. :) 14:51
Thanks, I must go now.
ugexe "1.2" ~~ 1.2 is the most simple possible example one could come up with though 14:52
ugexe m: say "1.a" ~~ "1..a" 14:52
camelia False
ugexe m: say Version.new("1.a") ~~ Version.new("1..a")
camelia True
ugexe so we can see Version is not just a string 14:53
not just a Str rather
ugexe looks like Str.ACCEPTS coerces too 14:59
so the problem is then this
m: say Version.new("v1.0")
camelia vv.1.0
ugexe that is not correct
so it cant just naively coerce
one cannot also assume they can just strip a leading v 15:00
vrurg ugexe: got back for a second. I would like to get back to this discussion. Not because I insist on coercion but because it's interesting from the perspective of smartmatch behaviors. bbl today! 15:08
Wahnberger Howdy :) I got this message "MoarVM panic: Memory allocation failed; could not allocate 1214058496 bytes". The code is just as simple: I have a list of domain names, which i am looping throug and do a http get with Cro (my $resp = await Cro::HTTP::Client.get("" ~ $dom );). i figured out that this call let moar hang 15:23
has anyone a hint where to look?
i have perl6 running on openbsd 6.4 in vmware 15:25
the byte count is diffrent every time 15:26
Wahnberger i found a similar report here github.com/croservices/cro-http/issues/74 15:29
that makes me wonder since i have other code using cro/client successfully :/ 15:30
ouhouh.. MoarVM version 2018.11-39 works fine 15:32
MoarVM version 2019.03 does not (in this particular case) 15:33
travis-ci Rakudo build failed. whitesource-bolt-for-github[bot] 'Initial WhiteSource configuration file' 16:07
travis-ci.com/rurban/rakudo/builds/114443660 github.com/rurban/rakudo/compare/0...616dd1c465
Kaiepi .tell wahnberger, did you try passing the html version as 1.1? 17:06
yoleaux Kaiepi: What kind of a name is "wahnberger,"?!
Kaiepi .tell wahnberger did you try passing the html version as 1.1?
yoleaux Kaiepi: I'll pass your message to wahnberger.
antoniogamiz o/ 19:15
timotimo \o
antoniogamiz someone knows how to specify mi6 to use more than one file to generate the README?
timotimo modules.perl6.org/dist/App::Mi6:cp...6.pm6#L131 - all i know 19:18
if you look closely, you'll find that it runs perl6 --doc=Markdown $the-file-you-put-in-the-config
and there's a phaser named DOC that runs only when documentation is being generated
antoniogamiz so... it cannot be done? 19:19
timotimo quite the opposite 19:20
anything you output in the DOC phaser will go in the file
antoniogamiz ah then I'm a bit lost here haha
what's the DOC phaser? :/
timotimo m: DOC { say "this only runs when the documentation of this file is being requested" }
camelia 5===SORRY!5=== Error while compiling <tmp>
Undeclared name:
DOC used at line 1
timotimo wait what? 19:21
m: DOC INIT { say "hi" }
camelia ( no output )
timotimo oh, i didn't realize you have to put it in front of BEGIN, CHECK, or INIT
antoniogamiz so I have to add a phaser INIT and DOC? 19:23
but what do I put in the DOC? 19:24
sorry for not get it
timotimo it'd be "DOC INIT", "DOC BEGIN" or "DOC CHECK" 19:25
antoniogamiz I've put that but it's not working :/ 19:28
I've opened an issue: github.com/skaji/mi6/issues/72
timotimo what exactly did you put, and what did it do or not do? 19:30
antoniogamiz I added exactly DOC INIT to the beginning of the pm6 file I want to be included in the README 19:31
and after that I've executed mi6 19:32
agh I must go, pls tell me what you think and I'll come back later 19:33
thanks a lot for the help ;)
timotimo gotta install mi6, hold on
timotimo oh, no, that's not how i meant it 19:33
zostay Anyone know of a way to get DBIish to find homebrew installed /usr/local/lib/libmysqlclient.dylib under macOS Mojave? I've tried symlinking, LD_LIBRARY_PATH, DYLD_LIBRARY_PATH, and just about everything else I can think of, but no joy. 19:43
timotimo i think we don't automatically look at these env vars in NativeCall 19:43
you can try dtruss to see where it looks, perhaps 19:44
zostay running dtruss on mojave requires arcane knowledge i don't have 19:49
that is, root is not normally permitted to run dtruss or dtrace
timotimo dtruss isn't the equivalent of strace on linux? 19:51
zostay i think it is, but you can't run it without getting real root access macos is rootless normally 19:52
root is not root
i try not to do system level debugging on macos if i can avoid it... if it's that involved, i'll just pin up a container and run my code in there :-p
timotimo can you attach gdb or lldb? 19:53
zostay probably
but that's like asking if i can fix the starter on my truck... i could eventually find it and probably figure out how to get it off and buy a replacement... but gee whiz, i don't want to fix my truck, just drive it 19:55
timotimo we'd probably want to set a breakpoint in MVM_nativecall_load_lib 19:56
if you're building with libffi, it'll be defined to just be dlopen, if you're building with dyncall it'll be dlLoadLibrary
lizmat weekly: blogs.perl.org/users/laurent_r/2019...atrix.html 20:55
notable6 lizmat, Noted!
jaffa4 hello 21:01
Null regex not allowed ------> '\'' ( <ESC> | ~ ( '\\' | '\n' | '\'' ) <HERE>)* '\'' )} 21:02
What is the problem here?
timotimo oh hey jaffa4 21:05
~ is an infix operator in regex
jaffa4 is it not?
timotimo there's nothing in front of the ~ in yours
timotimo m: say "hello" ~~ / ~ 1 / 21:05
camelia 5===SORRY!5=== Error while compiling <tmp>
Null regex not allowed
at <tmp>:1
------> 3say "hello" ~~ / ~ 17⏏5 /
timotimo m: say "hello" ~~ / a ~ 1 /
camelia 5===SORRY!5===
Unrecognized regex metacharacter ~ (must be quoted to match literally)
at <tmp>:1
------> 3say "hello" ~~ / a ~ 17⏏5 /
Unable to parse regex; couldn't find final '/'
at <tmp>:1
------> 3say "hello" ~~ / a ~ 17⏏…
timotimo m: say "hello" ~~ / a ~ 1 99 / 21:06
camelia Nil
timotimo it actually needs three pieces to parse correctly 21:06
but not having one in front gives that error
which seems LTA
jaffa4 LTA? 21:07
timotimo "less than awesome"
as in "not good enough for the standards of perl6 error messages" 21:08
could make a ticket for it if it isn't already there
jaffa4 ok
is there not operator in regex?
like - for classes
timotimo there's negative lookahead/lookbehind 21:10
jaffa4 so coming back to previous message 21:11
it is just a misleading error message. The real problem is ~ 21:12
ok ... I need to go bye 21:13
timotimo oh, sorry 21:13
yeah, you're probably not intending to use ~ thoro
there*
woolfy . 21:22
Xliff m: my @a = 'a'..'z'; .say for @a.batch(5) 21:50
camelia (a b c d e)
(f g h i j)
(k l m n o)
(p q r s t)
(u v w x y)
(z)
Xliff m: my @a = 'a'..'z'; my $b = 24/5; .say for @a.batch($b) 21:51
camelia Cannot resolve caller batch(Array:D: Rat:D); none of these signatures match:
(Any:D: Int:D :$elems!, *%_)
(Any:D: Int:D $batch, *%_)
in block <unit> at <tmp> line 1
Xliff m: my @a = 'a'..'z'; my $b = 24/5; .say for @a.batch($b.Int)
camelia (a b c d)
(e f g h)
(i j k l)
(m n o p)
(q r s t)
(u v w x)
(y z)
Xliff ^^ Coerce to int, maybe?
timotimo whoopsie-daisies. remote code execution in vim and neovim via modelines 22:24
jnthn timotimo: link? 22:37
timotimo github.com/numirias/security/blob/...-neovim.md
er, not remote; arbitrary
but it's trivial to set up a reverse shell 22:38
vrurg timotimo: there is no clear non-hacky way of excluding .PREVIEW from matching v6.* in the pragma. 22:44
mst timotimo: speaking as an ex-vi user: lol I win (sorry) 23:39
Xliff mst++ 23:46
\