🦋 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.
jnthn We need to actually get RakuAST integrated into the compiler and produced by it before we can play with this really 00:00
Probably best is to ask me again when we get to that point.
SmokeMachine I’ll do that! 00:01
jnthn Main thing we need to worry about is any composability concerns, 'cus I've got a feeling that, like with the MOP, certain folks will have a lot of fun with custom compiler passes. 00:02
Like "ooh, now my extra type system stuff can get compile-time checked now" at the like
SmokeMachine That will be the first think I want to play with… :P 00:04
Geth ¦ problem-solving: FCO assigned to jnthn Issue Making a signature accept a list of Pairs github.com/Raku/problem-solving/issues/177 00:06
SmokeMachine that wasn’t a good description… sorry... 00:08
jnthn: sorry to insist… but I’m curious: how do you think to store the AST from the custom pass and get access to it on run time? would it be dome with something like: `&block but StoredAST(&block)` ? 00:11
jnthn Under current design, it'd probably be something like `$the-block-ast-node.meta-object but StoredAST($the-block-ast-node)` 00:12
SmokeMachine Interesting... 00:13
ang it would be gotten with something like this? `&block.^get-ast`? 00:15
os `.meta-object` isn’t .HOW?
jnthn Well, a custom compiler pass would be traversing RakuAST
00:16 vrurg_ joined
jnthn I'm using `meta-object` in the very general meaning of "an object that describes a program element", so in the case of a RakuAST::Block then it would be a Block object 00:16
00:17 vrurg left
jnthn In principle, though, during compile time, you will be able to call .ast on a Code object, thus how a trait handler would get at it 00:17
Calling that after CHECK time would, by default, be like "sorry, too late" 00:18
SmokeMachine make sense 00:19
thank you very much!
jnthn Welcome :) 00:20
Probably sleep time for me; 'night o/
SmokeMachine for me too! good night! 00:21
00:28 mensvaga1 joined, mensvaga1 left 00:29 mensvaga left 00:31 aborazmeh left 00:36 softmoth joined
Geth whateverable: 789d525aad | (Aleks-Daniel Jakimenko-Aleksejev)++ | 4 files
Unbitrot some tests
01:01 melezhik joined
melezhik chloekek: what is main use case for Raku modules delivered as Nix packages? Are you developing Raku applications for Nix OS ? 01:03
tellable6 melezhik, I'll pass your message to chloekek
01:03 sena_kun joined 01:05 Altai-man_ left 01:10 aborazmeh joined, aborazmeh left, aborazmeh joined, rbt left 01:11 rbt joined
Geth whateverable: 60de9e3bd8 | (Aleks-Daniel Jakimenko-Aleksejev)++ | 2 files
Unbitrot more tests
01:29 molaf_ left 01:42 molaf_ joined, melezhik left 01:47 aborazmeh left 01:49 softmoth left 02:05 Manifest0 left 02:06 Manifest0 joined 02:11 aborazmeh joined, aborazmeh left, aborazmeh joined 03:02 Altai-man_ joined 03:05 sena_kun left 03:11 aborazmeh left 03:36 pilne joined 03:48 pilne left 03:49 k-man__ joined 03:50 k-man_ left
cpan-raku New module released to CPAN! Term::Choose (1.6.4) by 03KUERBIS 03:53
04:50 evalable6 left, linkable6 left 04:51 evalable6 joined 04:53 linkable6 joined 05:01 Doc_Holliwood joined 05:03 sena_kun joined 05:05 Altai-man_ left 05:09 [Sno] left
Voldenet m: my $i = [*] 1000 … 1; say now - BEGIN { now } 05:25
camelia 0.087364
Voldenet m: my $i = [*] 1 .. 1000; say now - BEGIN { now }
camelia 0.0261417
Voldenet huh
Why is there such major performance difference? 05:26
guifa2 sequence operator isn't as optimized, lizmat is working on making some improvements there
m : my $i = [*] 1 ... 1000; say now - BEGIN { now } 05:27
m: my $i = [*] 1 ... 1000; say now - BEGIN { now }
camelia 0.08144355
05:28 cpup left 05:29 cpup joined
Voldenet ah, sequence op can do more - it makes sense 05:30
guifa2 yeah, the sequence operator internally IIRC runs a full code block 05:32
And then smart matches against the final term
So that's two separate code blocks that have to get run 05:33
05:46 sauvin joined, sjm_uk joined 05:51 [Sno] joined 06:02 ufobat_ left, reportable6 left 06:04 rindolf joined 06:05 reportable6 joined 06:22 andrzejku joined 06:44 Doc_Holliwood left, sjm_uk left 06:48 lichtkind joined 06:49 stoned75 joined 06:52 abraxxa joined 06:57 abraxxa left 06:58 abraxxa joined 07:02 ilogger2 joined, Altai-man_ joined 07:31 sjm_uk joined 07:43 dakkar joined 07:57 pecastro joined 08:19 rindolf joined 08:38 aborazmeh joined, aborazmeh left, aborazmeh joined 08:43 eseyman joined 08:49 kensanata joined 09:03 sena_kun joined 09:05 Altai-man_ left 09:07 Ven_de_Thiel joined 09:22 kensanata left
lizmat Files=1306, Tests=111236, 213 wallclock secs (28.64 usr 7.88 sys + 3001.87 cusr 269.23 csys = 3307.62 CPU) 09:54
10:11 squashable6 joined, rindolf left
El_Che does someone use the rakudo-pkg i386 ubuntu debs? 10:13
I think the next ubuntu will remove all docker images and the base tar.gz from where I build i386 images myself
10:14 domm left
El_Che if there is a use for i386 rakudo packages I can see to move the i386 builds to debian... or drop i386 altogether 10:15
(on other news, when ubuntu 20.04 is out in a few days, a package will be ready to be built: rakudo builds just fine in the beta 20.04) 10:16
10:18 lichtkind_ joined
El_Che I have been looking to expand to arm, but I have been out of time and I have no arm hardware :) 10:18
without JIT, do people want to run raku on arm?
MasterDuke several people have asked about arm, there was a question about it lizmat linked in yesterday's weekly 10:19
jnthn "without JIT" is true in the "we don't produce machine code" sense, but you still get type specialization and inlining and so on, which in many cases are the biggest wins anyway 10:21
El_Che jnthn: so raku on arm is an option? 10:22
I read some docs on how to do it, but running the builds on travis complicates things (I like creating the packages on travis so the build proces is transparent from a security point of view) 10:23
it requires qemu + docker
the point of my i386 images was originally for cloud providers still defaulting to i386 or giving it as an option for low RAM VMs, but that time is over I think 10:24
10:24 Ven_de_Thiel left
El_Che the usefulness of i386 packages (and builds) were more useful as a rakudo canary in my eyes 10:25
(if we support the arch, it's a good idea to build it regularly)
about arm: I've always wanted a rpi, but since I run all my home stuff from a NUC with docker I kind of lost the use case :) 10:27
10:34 Doc_Holliwood joined 10:35 _jrjsmrtn joined
lizmat clickbaits rakudoweekly.blog/2020/04/20/2020-...-progress/ 10:46
11:01 mowcat joined 11:02 Altai-man_ joined 11:05 sena_kun left 11:06 rindolf joined 11:08 stoned75 joined 11:26 aborazmeh left 11:35 Doc_Holliwood left
veesh i have a pri, but goshes, a NUC sounds sick 11:35
how much do they cost?
s/pri/pi/ 11:36
El_Che a dual thread celeron cpu one around 120€ 11:37
it's fast enough to run many containers: an ssh hop server, logitech media server, torrent daemon, ubiqui controller, pi hole, nfs and samba shares, etc etc 11:38
abraxxa I do all that with a Synology DS918+ NAS
El_Che 8gig ram + 2 ext usb disk with ZFS == cheapo nas
abraxxa sadly that's also crap Intel based so low-level security vulnerabilites all the way 11:39
El_Che I have a second nuc as a media device connected to the tv, also celeron
kodi flies, also retropie works nice on that machine 11:40
only downsize of the celeron is that hw accelation on browser in non straightfwd and cpu is too weak for 1080p youtube (while on kodi no problem because of gpu accelation) 11:41
I run a hacked chromium version for youtube
(which handles 1080p fine) 11:42
abraxxa OSMC 4K FTW
El_Che no 4K screen here yet
OSMC is just kodi, right? 11:43
because like I said, kodi flies on the low powered nuc
just chrome and firefox are sluggisch with stuff > 720p
I like running regular ubuntu so I can run other stuff, like retropie. My kids loved the old nintendo games 11:44
abraxxa yes 11:51
but hardware accelerated with a fantastic support and price
basically 100euros for a Kodi box with fast updates that always work and use fully utilize the hardware
11:53 dakkar left, dakkar joined 12:01 Doc_Holliwood joined
El_Che sure, I agree 12:02
but for a home server, a nuc matched my case better
for a media server, I would need to think hard with the newest rpis
when I bought the pair of nucs, the rpi where a lot slower 12:03
I wonder is de rpi is fast enough to play Nintendo64 games 12:04
12:08 domm joined 12:15 dakkar left 12:17 dakkar joined 12:33 tejr joined
Kaiepi releasable6, status 12:38
releasable6 Kaiepi, Next release will happen when it's ready. 5 blockers. 166 out of 338 commits logged (âš  3 warnings)
Kaiepi, Details: gist.github.com/06f45e2699c56a8a94...c8a5cca912
12:41 Doc_Holliwood left 12:43 stoned75 left 12:51 lichtkind_ left 12:52 aborazmeh joined, aborazmeh left, aborazmeh joined 13:03 sena_kun joined 13:05 Altai-man_ left 13:22 mensvaga joined 13:32 Xliff joined
Xliff [09:34:02] <Xliff> How would this command like look: sub MAIN (:@args) { } 13:39
[09:34:32] <Xliff> I don't see this references in docs.
13:41 melezhik joined 13:42 melezhik left, Doc_Holliwood joined
dakkar Xliff: `sub MAIN(:@args) { .say for @args }` when called `foo --args=a --args=b` prints `a` then `b` 13:45
is that what you were asking? 13:46
Xliff Yep. Thanks.
dakkar it's a named array, so it maps to a named repeatable option
[Coke] you can get a hint by putting that into a small file, and invoking it, you get:
Usage: foo.raku [--args=<Any> ...] 13:47
Xliff OK. My code doesn't work that way. Prints usage
13:49 mowcat left
Xliff Using :@args completely breaks the script. 13:50
13:50 melezhik joined
Xliff As in, the code cannot match :@args and won't work, even if its named. 13:50
Constantly shows usage and then exits.,
Acting as if :@args is required. 13:51
nine Xliff: how did you call it?
Xliff main --args=a --args=b
main --args=a,b
returns usage
m: sub MAIN (:@args) { } 13:52
camelia ( no output )
Xliff m: sub MAIN (:@args) { } MAIN()
camelia 5===SORRY!5=== Error while compiling <tmp>
Strange text after block (missing semicolon or comma?)
at <tmp>:1
------> 3sub MAIN (:@args) { }7⏏5 MAIN()
expecting any of:
infix stopper
statement e…
Xliff m: sub MAIN (:@args) { }; MAIN()
camelia ( no output )
13:54 rindolf left
nine Can't reproduce: nine@sphinx:~/test/main> cat main 13:54
sub MAIN(:@args) { }
nine@sphinx:~/test/main> ./main 13:55
nine@sphinx:~/test/main> ./main --args=foo
nine@sphinx:~/test/main> ./main --arg=foo
Usage: ./main [--args=<Any> ...]
Xliff Hrm.
nine: I'll have to golf on my end.
I'm on:
Try MAIN($filename, :@args) 13:56
[Coke] then I get this usage suggestion:
foo.raku [--args=<Any> ...] <filename>
13:57 rindolf joined
[Coke] and I can invoke from CLI with: raku foo.raku --args=3 --args=4 TEST 13:57
13:57 andrzejku joined
Xliff Hrm. OK. 14:03
I've had to work around. I'll circle back and try to confirm this, later.
14:33 aborazmeh left
[Coke] CLI interaction may be different from in-program? 14:37
14:44 lichtkind joined 14:58 melezhik left 15:02 Altai-man_ joined 15:05 sena_kun left 15:10 MasterDuke joined 15:12 caterfxo joined 15:14 stoned75 joined 15:19 wamba joined, wamba left 15:24 natrys joined 15:26 Xliff left 15:39 [Sno] joined 15:48 melezhik joined 15:50 chloekek joined 15:53 benlittle joined 16:10 rbt joined 16:21 melezhik10 joined, melezhik10 is now known as melezhik2
melezhik2 Hi chloekek: could you please follow up on my last question about Raku/nix? 16:22
16:29 cpan-raku joined, cpan-raku left, cpan-raku joined 16:30 wamba joined, wamba left 16:33 xelxebar joined 16:36 Doc_Holliwood left 16:38 caasih left, BuildTheRobots left 16:40 dylanwh left 16:42 cooper joined, freezing_ joined, tbrowder joined 16:43 peteretep joined, MasterDuke left, Doc_Holliwood joined 16:44 caasih joined, BuildTheRobots joined, peteretep left, dylanwh joined
El_Che about the end of i386 ubuntu releases: I was able to create a i386 debian container, the in the future the ubuntu i386 packages will be replace by debian 16:45
food 16:46
16:46 dakkar left, peteretep joined, wamba joined, ukine joined, wamba left 16:47 wamba joined, wamba left
chloekek . 16:52
tellable6 2020-04-21T01:03:10Z #raku <melezhik> chloekek: what is main use case for Raku modules delivered as Nix packages? Are you developing Raku applications for Nix OS ?
chloekek melezhik2: I use Nix for dependency management regardless of programming language, so when I wanted to use Raku libraries, I first had to build Nix infrastructur.
melezhik2 yeah, so you run your install/run you Raku applications using nix package manager, right? 16:53
Doc_Holliwood sup 17:02
17:03 sena_kun joined 17:05 Altai-man_ left
chloekek melezhik2: I build them with Nix and then I run them or deploy them to a server. 17:31
melezhik2 yeah. chloekek: I wonder how you _run_ them? 17:32
chloekek nix-build && result/bin/...
Or if they’re dæmons, I generate systemd service files that have Nix store paths in ExecStart= 17:33
melezhik2 yeah. so what deploy does? does it just copy nix descriptors to target systems? 17:35
chloekek I build locally with Nix, then I invoke nix-store -qR which prints a list of all Nix stores that the artifact depends on, then I rsync those over SSH. :) 17:36
s/Nix stores/Nix store paths/ 17:37
melezhik2 sorry, now I am starting to get confused, do you _build_ packages on target system?
chloekek I don’t build them on the server I deploy to. 17:38
I build them on my workstation. Then I rsync the build output and all the runtime dependencies.
melezhik2 ok. so you build them _once_ somewhere, then deploy them, basically over rsync? 17:39
chloekek Yes.
The way Nix works is that all the build output as well as the dependencies go into /nix/store.
And with the nix-store command-line program you can query that, e.g. to list the dependencies of something.
Then I know what all the things are that I have to upload with rsync.
There is nix-copy-closure which does essentially the same thing, but it requires that Nix is also installed on the server, which I find annoying to set up. 17:40
melezhik2 ok, i see. I am more interested in target server. So on target system, how do you _run_ ? Sorry if I ask you again, just want to be sure I understand
chloekek For instance, these are all the dependencies of one of my Raku programs: gist.github.com/chloekek/237b400d9...8ff8332c57
You see that among them is Rakudo, and dependencies of Rakudo such as glibc. 17:41
melezhik2 yeah, I see. so having all those files rsynced to a target server, how do you run an application?
chloekek One of the store paths in that list is called “crai-sysadmin”. 17:42
This is a Nix package I wrote that contains systemd service files for CRAI, my Raku app.
Here’s what such a systemd service file looks like:
gist.github.com/chloekek/44782b48d...39383e8680 17:43
After rsyncing, my deploy script creates a symlink in /etc/systemd/system to that file and then restart the service.
melezhik2 interesting ... 17:45
ExecStart=/nix/store/1bz26cjxx6gwi423dccsjqglyxfispkf-crai/bin/crai-fastcgi --database=/var/lib/crai/crai.sqlite3 --fastcgi-socket=/var/run/crai/crai.fastcgi
chloekek Yeah, that’s filled in by the crai-sysadmin Nix package during build, using a template .service file. 17:46
melezhik2 how does that ensure that all Raku dependencies will be resolved?
bearing in mind you have to ensure that PERL6LIB has to have all modules paths in it
chloekek /nix/store/1bz26cjxx6gwi423dccsjqglyxfispkf-crai/bin/crai-fastcgi is a Bash script generated by raku-nix, which sets up PERL6LIB and then invokes rakudo with my app. 17:47
Looks like this:
melezhik2 ahh , of-course , it's NIX way :-)
chloekek gist.github.com/chloekek/2f13a6c58...faa050320d 17:48
It’s absolute paths all the way down, which makes it so reliable (if it works on one machine, it probably will on another as well, regardless of environment)
melezhik2 so your deploy/run scenario in a nutshell is: 1) rsync nix files to server 2) run nix prepared script that ensures that all dependence get "linked" through adjusting PERL6LIB 17:49
chloekek Pretty much.
You can see the code in the deploy and sysadmin folders here: github.com/chloekek/crai 17:50
17:50 sjm_uk left
chloekek deploy/default.nix refers to sysadmin under the name “production”, because sysadmin/default.nix also outputs a development environment called “development” 17:50
deploy/deploy.raku is the entry point, that calls nix-store and rsync and ssh 17:51
17:51 Manifest0 joined
chloekek Note that there are ready-made tools for generating systemd services and uploading things, if you use NixOS, called NixOps. 17:53
I just like to handroll this myself so that I actually know how it works and don’t have to debug someone else’s code when things go wrong. :) 17:54
18:00 |Sno| joined
melezhik2 chloekek: thanks for breaking this down for me ... I need to think, maybe like a said I could how Sparrow could be useful in your workflow 18:01
it makes more sense for me now 18:02
jdv79 is Sparrow used for something in production?
18:02 [Sno] left
melezhik2 honestly I don't think anybody's used it, beside me. I used it on some my @work projects. Partly as automated tests , partly for daily tasks 18:05
but yeah I has some good experience in using Sparrow as automation test system for our production code 18:06
has -> have
AlexDaniel El_Che: debian food? :) 18:09
18:16 pmurias joined, benlittle left 18:20 benlittle joined 18:21 melezhik left
El_Che AlexDaniel: free and ethical? 18:32
18:33 pmurias left 18:44 m_athias joined
timotimo or just free and malloc 18:52
18:54 abraxxa joined 19:02 Altai-man_ joined 19:05 sena_kun left
chloekek Free as in free memory. 19:07
19:11 guifa2 joined
AlexDaniel chloekek: including memory used for disk cache or not? :) 19:12
chloekek Storage is cheap, just buy more disk.
Oh, disk cache. Buy more RAM.
19:14 benlittle left, benlittle joined
chloekek So I have a dark theme for CRAI but I don’t know when to enable it. Could put it in a cookie I suppose. 19:23
CSS has media queries for the light/dark preference but not all devices support it.
19:27 mensvaga left
AlexDaniel chloekek: just use media queries? 19:27
chloekek: I mean, really. Just default to something neutral for those devices that don't support it 19:28
19:28 MasterDuke joined
chloekek Thing is, my own use case already makes using media queries for this impractical. :) 19:29
AlexDaniel chloekek: it used to be that almost every website had just one theme, it wasn't too horrible
chloekek Twice a day I do two rounds of switching my terminal profile, editor, and various websites between light and dark theme. 19:30
And until the year of the Linux desktop, it will remain this way. :)
I could do it depending on the time of day, that would be neat. 19:31
AlexDaniel chloekek: interesting. Why? 19:32
chloekek I like the brightness to match my surroundings.
synthmeat simpler to just move to basement 19:33
timotimo i bet you can nick the calculation from redshift
of course that's only based on lat/lon and localtime
lizmat newer MBP's have automatic colour correction, I like it 19:35
as soon as I switch on a light, the screen turns less bright / blue 19:36
chloekek p6: say DateTime.now.hour
camelia 21
19:37 mowcat joined 19:40 andrzejku left 19:46 stoned75 left
Doc_Holliwood i just watchen an interesting kotlin talk. nice info but the speaker made me crazy 19:54
je live demoed and had the habit of writing his blocks { like this} 19:55
probably some kind of infant brain injury. only explanation =)
melezhik2 reverse keyboard? :-) 19:57
AlexDaniel Doc_Holliwood: there are many weird styles, has nothing to do with any kind of injuries 19:58
sometimes intentional, sometimes people think it's not as important (and it isn't, frankly) 19:59
I'm high on unicode, vertical alignment, and other stuff like this. Very pedantic, but you could as well argue that I have a brain injury because why else would I need the code to be formatted in this specific and consistent way. :) 20:02
Doc_Holliwood sometimes i think longer than i should about variable names in an attempt to make them sensible AND equally long so the =s in the assignments are vertically aligned ^^ 20:07
lizmat Doc_Holliwood: you can use spaces to align the =s :-) 20:08
20:10 mowcat left 20:11 rbt left, rbt joined 20:12 rbt left 20:13 rbt joined
Doc_Holliwood oh and i don't you if you know this. kotlin also has an anonymous variable "it" that you can use in lambdas and such 20:13
it's of course pronounced "it" like in "what is it". and the guy kept saying it like IT in IT department 20:14
guifa2 lizmat: TIL you can unspace nothing to fix off-by-one vertical alignment
m: my $a = 5; say $a\.Str
camelia 5
Doc_Holliwood everytime he said it my brain was like <"it", it's "it"!!> 20:15
guifa2 . o O ( i used to think I needed variable names to have length difference of ≥2 for unspace to be effective )
lizmat is reminded of a dancing called IT 20:16
guifa2 Doc_Holliwood: OTOH I've heard people call the show "the it crowd" rather than "the eye tee crowd" and it makes me cringe so bad 20:18
Doc_Holliwood gringe! that's the right word 20:24
yet nobody ever said paris hilton being an IT girl lol 20:25
20:25 melezhik2 left
Doc_Holliwood as it seems java and the jvm have caught up with a lot of new concepts. 20:25
lightweight new threads, higher order functions, type inference, parallel streams ... 20:26
i wonder, if it were more feasible to implement raku on the jvm than it was back then when the decision was made to develop Moar 20:27
the jvm today
El_Che not everyone loves java and having raku as a jvm-first lang would have closed many doors (and opened others) 20:28
Demos[m] Moar is easier to boot 20:29
very easy to boot raku on winodws
which is nice
El_Che I have the impressions that jvm languages are only popular with java people that are bored by java 20:32
20:35 pilne joined
AlexDaniel heh, latest graph: imgur.com/a/B6noOuG 20:45
20:45 melezhik joined
AlexDaniel it's nice to know that the script is still working… :) 20:45
20:46 MasterDuke left
melezhik chloekek: what is CRAI for? does it server as an index to build nix/raku packages, any other purpose? 20:47
chloekek I needed a cronjob that indexes metadata and hashes of distributions. 20:48
Might as well make a web interface for it.
melezhik yeah, I know. but I am asking who is the client of such an index?
so far it's only a raku-nix, right? 20:49
chloekek The tools/generate-nix.pl script in raku-nix, which is currently broken and I have to fix. :)
AlexDaniel melezhik: well, if it keeps improving at that pace then we might as well replace modules.raku.org/ with it 20:50
chloekek: btw is the latest version deployed? 20:53
melezhik yeah, so right now, it's raku-nix, in the future it's probably could be an index for raku modules eco system. correct?
AlexDaniel chloekek: I'm just wondering why links are not clickable: crai.foldr.nl/archive/https%3A%2F%...0.3.tar.gz
melezhik: maybe
melezhik are you guys thinking about changing underlying deploy mechanisms (zef) if would change index system for modules? 20:55
I'm just curious ...
chloekek AlexDaniel: yes. 20:56
AlexDaniel I don't think there are any plans yet. In fact, even the idea of CRAI replacing modules.raku.org is bit of a stretch at this point
chloekek AlexDaniel: links are not clickable because they don’t have an a element surrounding them. :)
AlexDaniel chloekek: github.com/chloekek/crai/issues/20 20:58
chloekek I didn’t bother making them clickable yet because sometimes you have git:// or ssh:// URLs or git@ “URLs” and they have to be converted to HTTP URLs.
I’m going to sleep. Goodbye 21:02
21:02 chloekek left 21:03 sena_kun joined 21:04 Black_Ribbon joined 21:05 Altai-man_ left 21:09 p6steve joined
p6steve p6: say set <a b c> 21:09
camelia Set(a b c)
21:10 daxim joined
guifa2 m: my $a = buf16.new(0x6162, 0x6364, 0x6566); say $a.decode("ISO-8859-1").encode("ISO-8859-1"); 21:12
camelia Blob[uint8]:0x<62 61 64 63 66 65>
guifa2 . o O ( endianness issue but how to fix it? )
21:12 melezhik left 21:14 p6steve_ joined
timotimo my kejoard Odidd 21:14
21:15 MasterDuke joined, molaf joined, p6steve left
p6steve_ p6: set <b c d> 21:16
camelia WARNINGS for <tmp>:
Useless use of "set <b c d>" in expression "set <b c d>" in sink context (line 1)
21:17 cognominal joined 21:22 p6steve_ left 21:23 p6steve joined 21:34 p6steve left 21:35 rindolf left 21:36 Doc_Holliwood left 21:40 xelxebar left 21:41 xelxebar joined 21:42 p6steve joined 21:46 p6steve left
tbrowder testing 21:47
just got conflicting confusing md 21:48
msgs from Nickserv. one said confirm, another said i was confirmed, the next said to confirm... 21:49
is there some wild spam thing going on? 21:50
21:56 daxim left 21:57 natrys left 22:01 daxim joined
lizmat tbrowder: I only know of issues on irc.perl.org today 22:06
so not here :-)
tbrowder thnx
22:11 aborazmeh joined, aborazmeh left, aborazmeh joined
guifa2 hrm, Comma no like $¢ =\ 22:13
22:19 p6steve joined 22:24 p6steve left, aborazmeh left 22:39 cpup joined 22:56 lichtkind left 22:57 p6steve joined
AlexDaniel jmerelo: I'm not kidding. >:) 22:58
tellable6 AlexDaniel, I'll pass your message to jmerelo
23:02 p6steve left, Altai-man_ joined 23:05 sena_kun left 23:08 veesh_ joined 23:10 veesh_ is now known as veesh 23:15 pecastro left 23:28 cpup left 23:29 cpup joined 23:49 k-man joined 23:56 melezhik joined 23:58 melezhik left