»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'perl6: say 3;' or rakudo:, niecza:, std:, or /msg camelia perl6: ... | irclog: irc.perl6.org | UTF-8 is our friend!
Set by sorear on 25 June 2013.
timotimo doesn't quite know why he thought he would be able to sleep just by lying down and closing his eyes 00:02
i don't really see a reason why Pod::To::HTML should be any slower on moar than on, say, parrot 00:04
or use up much more memory
:\
jnthn Things typically use less, but it may hit something odd or pathological. 00:05
lue timotimo: it's my generator, not the ::HTML; it dies during backlinks processing
timotimo oh, ok
jnthn lue: Does it run on any backend?
lue jnthn: still runs fine on parrot, trying perl6-j now 00:06
timotimo just as i laid in bed i wished for a way to get a dump of the partial parse tree when rakudo fails to parse something
is that kind of sort of feasible?
probably has to do some difficult interaction with the highwater stuff 00:07
[Coke] using the rule debugger?
jnthn Highwater only tracks how far we got
timotimo you can't put the rule debugger into Perl6/Grammar ;)
and rxtrace is practically useless
jnthn Fixing rxtrace is probably an easier way to make things better... 00:08
timotimo (especially since if you compile any perl6 program with it, all you get is everything up to compunit)
raydiak I don't have parrot installed, run the Pod::To::HTML stuff w/JVM...so the comparison I was making was moar vs jvm (perhaps unfair, granted) 00:11
lue jnthn: here's the failing grammar: gist.github.com/lue/b99b113e4c82873293d0 It leaks and dies by kernel before the DEBUG() in TOP can be called. 00:13
00:13 rurban left 00:14 rurban joined
lue oh, I think I might be backtracking too much in that grammar :) Can't remember how many of those C<regex>s can't be C<token>s 00:14
jnthn Yeah, backtracking in grammars is a sure way to eat memory 00:15
lue waitaminute, what the hell am I doing with '\/' and '\>' and '\»' ‽ 00:17
00:19 berekuk joined
jnthn 'night, #perl6 00:20
lue ♞ jnthn o/
Huh, changing that to be '/' and such didn't break anything. Somehow.
lue doesn't even know how stuff like '>' worked in the first place 00:21
00:22 virtualsue joined
lue .tell jnthn doing s:g/regex/token/ (as well as removing useless backslashes in single-quoted things) makes it not eat memory on perl6-m anymore, *but* perl6-m doesn't catch nearly as many hits as perl6-p and perl6-j from the parsed files (-p and -j still work as before with the changes) 00:25
yoleaux lue: I'll pass your message to jnthn.
timotimo Unexpected named parameter 'deserialize_past' passed 00:27
at gen/moar/stage2/NQPHLL.nqp:1938 (/home/timo/perl6/rakudo/../install/languages/nqp/lib/NQPHLL.moarvm:add_load_dependency_task:12)
.. huh?
lue .tell jnthn for example, the test string "# L<S82/With Frag/QQW LIST-LIKE>" works on perl6-p and perl6-j, but not perl6-m
yoleaux lue: I'll pass your message to jnthn.
00:28 Rotwang left
timotimo ah, that was renamed in nqp, but not in rakudo itself 00:31
00:33 bjz left 00:34 bjz joined, tgt joined
timotimo lets rakudo catch up on the naming 00:35
00:38 jnap left, tgt left 00:44 ajr_ left, benabik joined, sisar joined
timotimo head too derped to figure out what broke now :| 00:50
00:51 daniel-s joined 00:52 berekuk left
daniel-s Hi there. 00:59
timotimo greetings human 01:00
daniel-s Is there an xml library for perl6?
timotimo you could parse xml with regex! :D
daniel-s Something like lxml
timotimo well, modules.perl6.org has XML, XML::Parser::Tiny and XML::Writer 01:01
hopefully with some tests to them :)
lue timotimo: they all have tests :)
timotimo well, beyond the "try to 'use' the module" test of course ;) 01:02
lue timotimo: you do see the tests medal next to them, right? :) 01:04
timotimo oh
now i even understand what the icon is!
it's - quite fittingly - a test tube 01:05
01:08 sisar left 01:13 benabik left
timotimo i didn't even know someone made a markov chain module for perl6 01:21
01:23 virtualsue left 01:28 virtualsue joined 01:29 Ben_Goldberg joined, BenGoldberg left 01:35 tgt joined 01:39 tgt left 01:40 aindilis joined
grondilu One of the reason I don't use the REPL is that the command history is not persistent. Thus it's simpler for me to use bash's history. Any plan on creating a .perl6/history? 01:44
01:45 obra is now known as jesse, jesse is now known as obra
lue grondilu: not one that I know of. 01:46
01:54 pmurias left 02:04 Ben_Goldberg is now known as BenGoldberg 02:05 bjz left 02:08 eternaleye left
BenGoldberg TimDoady: Does this code look familiar to you? rosettacode.org/wiki/Parsing/Shunti...rithm#Perl ;) 02:10
geekosaur dowdy doady? :p
BenGoldberg oops
TimToady :) 02:11
BenGoldberg doesn't look at what he's typing
02:13 skids joined 02:21 molaf joined 02:22 tgt joined
TimToady BenGoldberg: yeah, that looks vaguely familiar for some reason... :) 02:25
02:25 benabik joined 02:26 tgt left 02:42 stevan__ joined 02:44 stevan_ left 02:45 ilbot3 left 02:46 bjz joined, ilbot3 joined 02:52 FROGGS left
raydiak okay, I added perl6.org-ish boxes and colors, and made search wider and fluid...looking better? 02:52
p6doc.cyberuniverses.com/
s/fluid/liquid/ 02:53
(or whatever the kids call it these days)
03:05 FROGGS joined 03:13 molaf left
dalek c: 41a7a00 | raydiak++ | / (7 files):
Add colorful shininess
 03:15
lue raydiak: looks good, though my silly and perhaps untrustworthy brain thinks green & yellow are too close to each other on the spectrum to be the only significant colors for most of the page :P 03:17
(if/when a LHS/RHS navigation box happens, it would of course take up the blue Perl6 color) 03:18
raydiak lue: the blue I left on the links is meant to help with that some, but yes it will get easier to balance many colors when there are more boxes to color :) 03:19
lue Or maybe my brain's "yech!" alarm wants the background of the main box to be lighter yellow instead of white...
raydiak: I did notice that the blue links helped out a bit there :) 03:20
But definitely raydiak++ for a making a pleasing layout (or any kind of layout for that matter) :D 03:21
raydiak I was thinking about making the search box blue...
lue I can see that being a possibility. 03:22
03:23 tgt joined
raydiak I didn't think people would like having the whole content area background be yellow...perhaps if it was a *very* pale sunny color, but it's pushing the limits of some peoples tolerances for warm colors 03:23
raydiak would personally object to having to stare at urine-colored docs :) 03:24
lue yeah, I could see that too. I'm just trying to come up with little ideas to make my brain like the coloration more :) .
raydiak I certainly see what you mean...will meditate on it some more
lue I'm finding it difficult to imagine possible color-balancing solutions with green and yellow already on the page. Maybe it's just that I gravitate towards more red/blue colors, who knows.
raydiak tomorrow is "clean up the construction dust" time, so I'll probably make some more little tweaks still while I'm in there...so your feedback is still pertinent 03:25
03:25 virtualsue left
raydiak I guess I just shy away from red mostly, and I thought blue was just so over-done as most peoples favorite color 03:26
lue I'll probably play with the colors myself, merely "visualizing" it doesn't seem to help :P
03:27 tgt left
lue raydiak: my main issue is that I can't think of a good way to put either red or blue next to the existing colors without lacking color from the other end of the spectrum :) 03:27
raydiak I wouldn't think any of this could go together, but it mostly seems to work for perl6.org somehow 03:29
probably b/c that site has all the content split up into smaller boxes, no single one is most dominant
lue yeah :)
One thing I notice is that the headers in particular still feel very much like the plaintext style, but I'm sure that's not so hard to fix. 03:30
raydiak you mean the section headings, or the global header? 03:31
03:31 dayangkun left
lue section headings, still looks like the same blue aliased fonts. 03:32
raydiak yes, I didn't change anything in there yet 03:33
though if they are aliased, that's your system :) 03:34
lue what font are they even...? 03:35
raydiak I don't even try to specify a certain font, just font-family, which your browsers chooses something appropriate for 03:36
lue Wow, helvetica is ugly
raydiak so they are font-family: sans-serif in the CSS
lue I saw font-family: helvetica, sans-serif
raydiak ah, must be coming from one of the other parts I didn't get to merging into the main stylesheet yet 03:37
like the perlcabal one, or the inline one being added by Pod::To::HTML
lue my normal sans-serif font looks just fine by the way :)
"Inherited from body#___top" 03:38
raydiak: ^ that's where the helvetica, sans-serif declaration comes from, it seems
raydiak no, that's where it is inherited from (a parent element) 03:39
lue perl.css:6 actually (looking at the less useful part it seems)
raydiak ah, yes
that's the perlcabal one
one of those nasty things pod2html forces on us
will fix (probably tomorrow) 03:40
lue I get the sense Pod::To::HTML should perhaps not be giving default stylesheets :)
(not inline anyway)
raydiak I get the feeling it's something that got pasted in and probably was never meant for long-term reuse
I mean, it inherently breaks local docs if you aren't on the net, for one 03:41
lue yeah. If/when you get to hacking on that module, I wouldn't mind if it disappeared all of a sudden :) [others might, though...]
raydiak until I have a more clear idea of what other ramifications that might have, I'll probably just add a :!css flag or something to the pod2html sub 03:43
lue seems like the less destructive choice, yes :)
One more thing: If the page is shorter than your browser, the red box at the bottom just stops. Could probably be fixed with even a border running along the bottom, though I don't know how you'd prefer things to look ultimately. 03:44
raydiak good catch, I didn't happen across a page that short
lue substr is one such page :) 03:45
raydiak haha maybe on your fancy expensive monitor
raydiak does most of his hacking on a half-decade-old 13" laptop with no hard drive :P 03:46
lue Also, I should think a small Camelia picture needs to be in the header at some point :)
raydiak definitely agree
I also think replacing "search" with some looking glass icon would be a step up 03:47
as long as you have a bigger screen than I, does the search box get weird-looking long at full width? 03:48
lue Nah, I never really thought "wow, that's big" /me will take a picture of what he sees... 03:49
raydiak lue++ # lots of helpful feedback
suppose if I were smart I'd try zooming out...but you can't rely on that to be too terribly accurate in many cases 03:50
lue raydiak: this is what I see: i.imgur.com/9GuTc67.png 03:52
raydiak oh...need to set background-color on the input box :P 03:53
not supposed to have a grey rectangle there :)
it is a little long, but I guess if you have the width, it doesn't look too terrible overall... 03:54
lue raydiak: as long as it feels like it's on the right-hand side (as opposed of "everywhere but the title"), I don't think it's that bad. 03:55
s/of/to/
raydiak the whole thing including label is set to 40% of it's container, so should always be somewhat right of the horizontal center, unless you hit the min-width:18em 03:56
nevertheless, a max-width:40em or something might be good
lue Actually, I think a good metric would be "is the search box wider than the main content box?" :) 03:57
raydiak you code up the css for that w/out JS or non-compliant browser hacks, and I'll...be surprised :) 03:58
lue I like the search box being normal like that (unless of course you have a better color scheme for it ☺). Feels like search boxes should only be colored differently if white doesn't show up as a major color of the page (e.g. white-on-black coloring), but that's just me. 03:59
raydiak yeah blue may or may not look good like I talked about earlier, but the last thing I meant when I mentioned it was how you have a light grey rectangle in the middle of the otherwise-white box 04:00
lue Huh, that might've been my system, or perhaps even a monitor weirdness(!) 04:01
raydiak you took your screenshot with a camera? ;) 04:02
lue No, but sometimes my cable's signal gets interfered with and bleeding happens. I noticed that not realizing you were talking about that and it's gone now, after jiggling the cable. 04:03
raydiak I'm sure it's still there, just harder to see
it's defaulting to the default color of controls on your system, b/c I forgot to set it to white explicitly
didn't notice b/c input boxes are pure white on my end, but can't at all rely on that 04:04
lue I get #FFFFFF now, at least
raydiak huh
for the input itself?
lue It was #FCFCFC before
raydiak oh, if you're still looking at it on my server, I already changed it
it's live on doc.perl6.org now, too 04:05
lue yeah, #FFFFFF is from your server, #FCFCFC is from the picture I took.
raydiak (w/out the little tweaks I'm making as we talk)
I also added a (still generous) max-width to it 04:06
there, fixed the footer problem by just making it a normal box insetad of bleeding it off the edge 04:11
looks better that way anyway, imo
lue looks good :) 04:12
dalek c: d5fc7b6 | raydiak++ | / (2 files):
We're using fixed width; may as well center it
 04:22
doc: e9bfb43 | raydiak++ | / (7 files):
doc: Merge branch 'master' into shiny - catch up template changes
04:22 dalek left
raydiak uh...heh...I really borked something 04:22
raydiak still does not like squashing commits
04:22 dalek joined, ChanServ sets mode: +v dalek 04:23 tgt joined
raydiak all the stuff that just got dalek kicked was stuff I had previously squashed at some point, and something I did made it all appear anyway...maybe I didn't squash at the right step or something 04:24
methinks they were still unsquashed in my fork when I did another merge 04:25
04:28 tgt left
raydiak also noticing that an empty commit message on a merge does not actually abort as it claims 04:31
totally OT: anyone have experience using windex on their glasses? just got them a few weeks ago so idk, and was going to clean them, but someone told me the fumes from cleaner not meant for eyeglasses will harm your eyes or something 04:37
not sure if I buy it
geekosaur not that likely, but it can remove protective or anti-glare coatings 04:38
that said, ammonia is a good thing to keep far away from your eyes 04:39
raydiak didn't get any of the coatings...guess I'll give it a shot 04:40
what does ammonia do to your eyes? dry them out? 04:41
geekosaur not sure offhand. it's generally not very good for membranes and such though 04:42
at the very least it'll be strongly irritating
raydiak well, I'm not blind yet... 04:44
04:45 virtualsue joined
geekosaur just because your body's mostly made of things with a lot of amino groups does not mean ammonia is good for it :p 04:45
raydiak but they look so close, especially with smudged glasses... 04:46
geekosaur they are close. just not close enough :p
raydiak I think I'll take your word for that one...I don't have to learn *everything* the hard way :) 04:47
geekosaur (organic chemistry is fun! as long as it's not taking place on/in your body...) 04:51
raydiak I saw a book on that once...so mandifferent little symbols...worse than p6, iirc 04:52
*many different
lue DNA thinks arsenic is close enough to phosphorous, for instance :) 04:53
geekosaur most chemistry does, actually. at least as far as putting it together. taking it back apart, not so much, whch is the problem 04:54
raydiak ouch...keep arsenic away from my DNA, noted
geekosaur keep it away period. the dna part isn't what kills you, it's that adenosine triarsenate can't be disassembled 04:55
since adenosine triphosphate is what provides energy inside a cell, and arsenic is all too willing to displace the phosphorus, this is a bit of a problem 04:56
raydiak without knowing much about chemistry or biology, all I can say is that sounds like bad news 04:58
geekosaur (short version: cellular processes break a phosphate group off of atp, send the adp back to be recycled into atp, and use the phosphate to do stuff. this does not work if an arsenate snuck in place of the phosphate, which it does very readily) 04:59
lue geekosaur: I heard arsenic was bad because of DNA, good to hear it's more specifically issues with ATP. 05:00
geekosaur well, it's not good for dna, but you don't typically live long enough for it to get into the dna 05:01
but it has the same holds-on-too-tight issue in dna, which means dna doesn't unzip any more and the cell can't reproduce if somehow it manages to remain functional after the atp is corrupted
raydiak hm...bad news indeed...something along the lines of "suffocate your cells" if I understand
geekosaur exactly 05:02
and ithappens fast enough that you don't generally reach the dna version of it before your cells die
raydiak geez, wish I could stay now, it's getting interesting :) but gotta run atm...nice chattin! 05:03
geekosaur lue was probably hearing about the claim of a bacterium that had found a way to survive in arsenic, which was only partially true (and the claim that it had found a way to survive arsenic getting into the dna was also false)
(but it only reached that point because it had found a way to keep from choking on corrupted atp) 05:04
lue geekosaur: I think I may have heard it around that time, though I didn't learn about the "infects DNA" part because of it I don't think (at least, the explanation wasn't (directly) linked to the bacterium)
geekosaur that was one of the key parts of the initial claim, that it not only had found a way to function without losing all its atp but that it had also managed to survive the subsequent incorporation of arsenic into its dna 05:05
followups demonstrated that it had managed to mostly keep the arsenic out of its dna, but sure enough the places it did get in were incapable of reproducing
also while it was surviving, it was (a) not thriving (b) surviving despite the arsenic, not actively managing to use it in place of phosphorus (another part of the initial claim) 05:06
also the way I am presenting this is a bit unfair, NASA said "we may have found evidence of a bacterium that might have figured out how to use arsenic" and it was mostly the press that went "OOO WE FOUND ARSENIC BASED LIFE!!!eleventyone!" 05:12
DrEeevil typical reinterpretation 05:15
geekosaur: seen the "we can create energy with fusion" claim from the NIF ? 05:16
geekosaur I saw the actual release, which did not claim that although I imagine the same idiots have interpreted it that way
DrEeevil well, yes, if you ignore the factor 5000 difference between actual input and output, then ... theoretically, if the measurements were good ...
it's so dishonest what "journalists" do
geekosaur they were quite clear, they have reached a milestone on the path to creating energy but are a long way from that endpoint 05:17
ens geekosaur: NASA doing dangerous science experimenting with deadly arsenic bacteria - could prove fatal to humans.
These ivory tower intellectuals spend their budget, already far in excess, dreaming up new ways to exterminate our species. Massive cuts are in order! 05:18
ens spins it like a DJ
DrEeevil riiiiight
ens: did you read about DHMO ? ;)
ens hrm, should i google?
geekosaur ens: not likely, it's otherwise a perfectly normal and common soil+water bacterium, the normal form is non-infectious, the arsenic-adapted form is so borderline it's even less capable of infecting much of anything, the main danger is the arsenic-laced water it was found in
ens geekosaur: ah i don't believe it, i'm just emulating media spin. 05:19
geekosaur although yes it does sound a bit like something idiots would say :p
and yes, you should google that for the yuks 05:20
it's the classic example of mocking the idiots
ens dihydrogen monoxide hoax. lmao.
upload.wikimedia.org/wikipedia/comm...drogen.jpg 05:21
geekosaur ens: wait until some wag realizes there's only one letter difference between NASA and NSA
05:21 pecastro left
ens oh we have to put on the voice of those youtube conspiracy theorists and be like "coincidence? you decide!" 05:22
and then lace it will arbitrary references to biblical connections and free masons, chemtrails. etc.
that youtube series 'spirit science' is like an epic treatise on the art of complete lunacy. 05:23
05:24 tgt joined 05:27 tgt left 05:28 tgt joined 05:33 tgt left 05:34 Alina-malina left 05:39 virtualsue left 05:41 Alina-malina joined 05:49 grondilu left 05:53 rurban left
timotimo ohai 06:03
06:14 SamuraiJack joined 06:18 BenGoldberg left
timotimo someone could add patience diff to Algorithm::Diff 06:22
06:27 thou left 06:29 tgt joined 06:33 tgt left
raydiak fears fusion-inducing arsenic bacteria put in the chemtrails by masons as foretold in the bible 06:36
06:41 darutoko joined 06:54 rurban joined 06:59 rurban left, rurban joined
timotimo rebased his lex-to-local branch a second time on top of almost-newest nqp master 07:05
07:30 tgt joined 07:34 tgt left 07:43 ens_ joined, ens left
dalek osystem: 066380f | lue++ | META.list:
Add Term::ColorText
 07:57
timotimo lue: i like that module 08:01
gist.github.com/timo/4416af400d9c86fd0f3c ← the lighter shades mean that there were objects there that had a size of 0 08:02
i don't quite know what's causing this.
it's not in the free list at any rate.
though *maybe* i'm reading the free list wrong? 08:03
08:04 SamuraiJack left
timotimo ooooh, i think i know what's wrong 08:05
08:05 berekuk joined
lue timotimo: thanks :) I set up functions like that every so often, and this time I decided to actually make it a(n independent) module :) 08:06
timotimo: that looks pretty :) 08:08
♘ #perl6 o/
timotimo gnite lue 08:10
08:15 berekuk left 08:16 rurban left 08:30 tgt joined 08:32 virtualsue joined 08:35 tgt left 08:47 rurban joined
masak mornin', #perl6 09:00
09:05 hoverboard joined
moritz \o masak 09:06
09:07 darutoko left 09:13 darutoko joined
raydiak good morning to people across oceans from me :) 09:13
09:14 virtualsue left
raydiak moritz: started new look for docs, feedback welcome (that goes for anyone else too) 09:15
moritz raydiak: I like it, though IMHO the coloring give the footer too much wheight 09:19
09:19 rurban left
moritz and I'm not sure I like that inline <code> doesn't have a different background color anymore 09:20
raydiak yeah, trying to match colors on perl6.org kinda painted me into a corner (almost a pun)
ah, will have a look at that in the morning 09:21
moritz maybe use yellow for the footer? 09:22
09:22 hoverboard left
moritz or the light blue 09:22
(tweaking perl6.org isn't forbidden either)
raydiak internet connection is flaking out, sorry if I'm even slower than usual :)
the colors seem to work well on perl6.org, it's just that it has a very different layout, so they don't work quite as well for large, unbalanced combinations of boxes 09:24
moritz raydiak++ # much improved layout
raydiak lue++ also suggested I think more about color selection, which I will have to give some thought to
raydiak has not ever attempted to mix colors like this before 09:26
I usually don't play with more than 2 basic hues, so it may take a few revisions to get it well-tuned 09:27
am also planning on removing the old search page and simplifying the generation and css accordingly, unless you have objections
moritz using the blue combination from the perl6.org header for the doc.perl6.org footer might work 09:28
09:28 grondilu joined
moritz also, the doc.perl6.org header uses a different styiling (that I like), clinging to the top; maybe for symmetry the footer should stick to the buttom in the same way 09:29
(or maybe the footer could simply be the same green as the header)
raydiak I had it that way originally, but it caused issues when the page is shorter than the window...I actually like it more as a separate box, though that could be just b/c it's that too-dominant red, as you say 09:30
s/separate box/fully bounded box/
09:31 tgt joined
raydiak there was also jnthn++'s idea to add a left nav box with anchor links to the subsection headingson the current page, and make that blue 09:33
09:36 tgt left
raydiak but it'd look really odd on pages with less content...ashort blue box and then empty space for the rest of the page in that column 09:36
09:39 berekuk joined
raydiak will try more permutations of colors, and bleeding off the footer again (assuming net connection improves) 09:41
right now, just typing these messages requires near-superhuman patience :)
well, I better get to bed...thanks for the ideas, we'll see where it goes tomorrow 09:44
if you have any more thoughts, I'll catch upon the backlog. good morning and goodnight, #perl6 09:45
09:50 spider-mario joined
moritz good night raydiak, and dream of beautifully-colored websites 09:55
09:56 daniel-s left 09:59 Rotwang joined 10:00 tgt joined 10:03 pochi left, pochi joined 10:06 virtualsue joined
FROGGS o/ 10:21
nwc10 \o 10:22
10:24 rurban joined
moritz |o| 10:27
10:39 dmol joined 10:43 denisboyun joined, berekuk left 10:50 rurban left 11:01 berekuk joined 11:04 tgt left
FROGGS jnthn: declaring an enum triggers MVM_sc_wb_hit_st in Attribute.compose (the call to add_method in Attribute.pm line 79) 11:10
err, line 78 actually
jnthn FROGGS: Ah, so you got it to a one-line thing? :) 11:14
yoleaux 00:25Z <lue> jnthn: doing s:g/regex/token/ (as well as removing useless backslashes in single-quoted things) makes it not eat memory on perl6-m anymore, *but* perl6-m doesn't catch nearly as many hits as perl6-p and perl6-j from the parsed files (-p and -j still work as before with the changes)
00:27Z <lue> jnthn: for example, the test string "# L<S82/With Frag/QQW LIST-LIKE>" works on perl6-p and perl6-j, but not perl6-m
FROGGS jnthn: well, it triggers the said function, but I don't get the STable exception (yet)
jnthn lue: Ah, then you're probably hitting a regex compiler bug, and the memory eating is it just building up more and more state due to a match fail. 11:15
lue: Please either file it as a rakudobug as is, or if you're willing to golf it some that's even better.
FROGGS: Oh...but removing the STable exception we added does...what?
11:16 virtualsue left
FROGGS jnthn: I just have a dump_backtrace(tc) in that function instead of the throw, because the thrown exeption gets swallowed by a try/catch somewhere 11:16
11:18 tgt joined
jnthn FROGGS: Oh...but we do reach that place? 11:18
FROGGS yes
we reach the place where the oops-exception was
jnthn ok, great. 11:20
And an enum A < b c >; is enough to recreate it?
FROGGS yes, "perl6-m -e 'enum Foo <bar>'" triggers the STable reposession 11:22
11:22 Funkwecker left
FROGGS now I even have a bt where it explodes in panda: gist.github.com/FROGGS/ba492ddbaa8...tfile1.txt 11:22
jnthn FROGGS: The other thing you could do is set a gdb breakpoint where the dump_backtrace is, and see what's on the C stack (which tells us what lower-level thing is guilty) 11:24
FROGGS k
jnthn FROGGS: I'm going to plug on with the IO stuff for a bit, but will give this enum thing some tuits later today. 11:25
FROGGS cool :o)
jnthn I've reached the point where I try to migrate tty, then pipe, over.
FROGGS the op before the dump_backtrace if setmethcacheauth 11:27
gdb bt: gist.github.com/FROGGS/60c7938ae23...6013e/raw/
jnthn oh...hm.
FROGGS yeah, I've read about tty/pipe in the clogs 11:28
bbiab (need to fetch food so the pups wont starve) 11:29
11:41 tgt left 11:46 virtualsue joined 11:53 virtualsue left, darutoko- joined 11:56 darutoko left 11:58 tgt joined 12:02 darutoko joined 12:06 darutoko- left 12:11 tgt left 13:14 rindolf joined 13:16 immortal left 13:20 SamuraiJack joined 13:21 virtualsue joined, erkan joined, erkan left, erkan joined 13:34 Rotwang left 13:48 virtualsue left 13:53 Rotwang joined 13:55 kaare_ left 14:01 virtualsue joined 14:02 lizmat joined
lizmat is in a holding pattern for Ijmuiden harbour in 90 miles/hour winds 14:05
jnthn Well, that blows... 14:06
timotimo lizmat: maybe you want to fix rakudo to use the new ast-based names while you wait? :)
i think it just involves search/replace for two words that are used as named arguments somewhere 14:07
i could actually do it myself, even.
lizmat well, that was the plan, but we need a version upgrade for nqp for that first, no?
timotimo yes
i just noticed that :) 14:08
lizmat I'm not sure how to upgrade the minimal nqp version for rakudo
timotimo i think i shall do it
timotimo starts working
lizmat timotimo++ 14:09
seeing the waves go over deck 6 sort of distracts
as well as having to hold on to your chair
and the load banging of the waves 14:10
:-)
it appears our 16 hour ferry is turning into a 25 hour ferry
timotimo has never been on a ferry that went for longer than ~30 minutes 14:11
FROGGS has never been on a ferry 14:12
colomon used to hop on the ferry to get hard ice cream in Canada when he was a wee lad 14:13
colomon admits he's never been on a ferry ride longer than a few hours (crossing Lake Michigan to Wisconsin) 14:17
lizmat it was either taking the ferry or driving an axtra 500km in high winds
*extra 14:18
with a fully loaded van
colomon ferry++
lizmat in any case, we're close enough to shore to be able to use my regular ISP's network 14:19
instead of the HTTP-only satellite of the boat
timotimo mhhh, kate highlighting definitions are shared by QtCreator and probably work with anything that uses some specific Qt widget perhaps? 14:20
14:30 mtk left 14:34 mtk joined
lizmat land ho! 14:37
FROGGS \o/
14:38 denisboyun left
timotimo \o/ 14:45
FROGGS .oO( Ho land ) 14:46
lizmat that too, finally within the inner piers of the harbour
timotimo hoelzro: you are our highlighting expert, are you not? 14:47
hoelzro I guess so?
timotimo heh.
have you looked into kate's xml highlighter definition files yet?
hoelzro nope
timotimo is currently doing so
hoelzro I'm thinking of enhancing Vim support next
timotimo that would be lovely, too
hoelzro the hardest part (imo) is bracketing characters 14:48
timotimo qtcreator shares support for the highlighters and that is the one IDE i could imagine myself writing plugins for, hint hint
hoelzro and re-entering the Perl 6 language from within another (ex. regexes)
the pygments lexer was actually not that hard to do
but my hat goes off to the pygments folk; they made a simple-to-extend but powerful system for lexers 14:49
FROGGS um... what about SCiTe/Notepad++ ? 14:50
err, SciTe
hoelzro I think if I did another higlighter, it would be for ACE 14:51
sine that's what GH uses for their online editor
*winc
timotimo aye.
hoelzro *since
14:51 berekuk left
timotimo well, at least the SciTe highlighters are c++ code, so you can go nuts there :P 14:52
14:53 lizmat left, Rotwang left
timotimo like ... embed a moarvm and use the actual perl6 grammar :D 14:53
moritz ... in an extern "C" { ... } block 14:54
hoelzro hahaha
timotimo that part is easy, moritz :)
hoelzro: can you suggest a roadmap for implementing my own lexer? 14:55
where should i start, etc?
hoelzro well
when I started with pygments, I just matched '.', I believe
which is kind of the base rule
and then I added feature by feature
using a few Perl 6 files as a reference point 14:56
so I'd go through the output, and spot things that looked messed up
and implement those piece by piece
then, after I got the basics done (keywords, etc), I created a regression test
to make sure that things remained properly highlighted 14:57
timotimo fair enough
hoelzro and just worked out the kinks
timotimo this should be a decent exercise for my flu-ridden brane
hoelzro you should take a look at the pygments lexer
timotimo oooooh
Contexts may be generated dynamically to allow the usage of instance specific data in rules.
hoelzro it contains a lot of data/helper routines that you could use 14:58
timotimo i shall have a look, aye.
14:58 Rotwang joined, Zoffix joined, Zoffix left
hoelzro let me know if you have questions =) 14:58
timotimo dynamic if true, the context remembers strings/placeholders saved by dynamic rules. This is needed for HERE documents for example. Default: false. 14:59
<3
not quite powerful enough for bracket pairs, though, i fear :(
hoelzro for Vim, I'm going to go for common bracketing pairs
timotimo aye.
i might build a script that creates rules for a list of bracket pairs actually
15:01 cognominal left 15:03 xinming joined 15:06 xinming_ left 15:20 berekuk joined
timotimo i'm ready to spectest my changes to rakudo that make it work with the newest nqp 15:36
FROGGS cool 15:38
timotimo twas easy ;)
15:41 Jimmy4 joined
Jimmy4 hi，I am -1 to s/ast/made/ 15:43
timotimo Jimmy4: well, so far it's not a substitution, but an addition
but i also liked .ast
dalek kudo/nom: 06fee3d | (Timo Paulssen)++ | / (3 files):
bump NQP_REVISION and fix ast/past incompatibility
 15:45
FROGGS I like .ast too, but I have to admit that .made makes some sense from a user perspective
timotimo yes 15:46
it's okay to have both
it's hardly any overhead at all, aye?
FROGGS I dunno these things
Jimmy4 I don't like the comment ' # for historical reasons' 15:47
good night
15:47 Jimmy4 left
timotimo good point. 15:50
15:59 Psyche^ joined 16:03 Psyche^_ left
timotimo hoelzro: you should introduce EVAL to the pygments lexer :) 16:34
until is a keyword in perl6? 16:37
oh, loop ... until i guess 16:38
TimToady no
repeat...until
loop (;;) {...} is C's for loop
but you can just loop {...}
until foo {...} is just them same as while not foo {...} 16:39
same for the modifier
moritz does anybody have a good example for parameterized roles? 16:40
hoelzro timotimo: is EVAL new? 16:41
FROGGS we have it since two months or so 16:42
humm.... I can't build nqp-p atm
hoelzro well, that would explain its absence 16:44
any other new syntax/keywords introduced in the last year?
timotimo hm, what about the parallelism stuff? 16:45
hoelzro good question
hoelzro should just do a git diff on the specs
FROGGS umm, eager and lazy? 16:46
timotimo what is ofs? o_O
moritz I think those are older
timotimo ofs, irs, ors?
they are not in perl6 any more, eh?
why would "oo" be a keycord? 16:47
keyword
moritz what would it do?
timotimo if that question was for me: "get highlighted by the pygments perl6 lexer" 16:48
haha 16:49
16:49 sqirrel_ joined
timotimo if you try, in the repl, to find out how many OUTER::s it takes to get to the setting ... 16:49
you'll infiniloop, because every line you execute adds an OUTER :D 16:50
16:50 hoverboard joined
FROGGS hehe 16:50
TimToady which it shouldn't, really, if we had a decent incremental hook into the parser 16:52
hoelzro 12,000 lines changed in the specs since I wrote the lexer o_O 16:53
jnthn That one's not really a parsing issue
timotimo yes.
jnthn I agree we should have a hook that lets you enter multi-line things though :) 16:54
timotimo so much yes!
hm. Str has methods indent and trim-*, but not center/ljust/rjust 16:55
those are kind of nice to have, though
16:58 smls joined
smls Hi #perl6 16:58
yoleaux 8 Feb 2014 06:38Z <TimToady> smls: Revised rosettacode.org/wiki/Brace_expansion#Perl_6 so it now passes all the tests I know about
timotimo hello smls 16:59
smls doc.perl6.org is looking much better all of the sudden... :)
moritz yes, raydiak++ did lots of work there 17:00
hoelzro whoa, I hadn't seen the changes 17:06
raydiak++ # indeed
timotimo kate is a surprisingly pleasant editor 17:11
FROGGS and somebody ported kate's syntax highlighting files to a P5 module, which is really nice also 17:14
17:14 rindolf left
timotimo oooooh 17:14
that's very useful, then!
.o( also, we could get a port to perl6 that way )
FROGGS I've used it lately to highlight soap messages for easier reading
17:16 virtualsue left 17:20 guru joined, guru is now known as ajr_
dalek href="https://perl6.org:">perl6.org: e64a28e | (Steve Mynott)++ | source/compilers/index.html:
correct perlito perl6 link
 17:21
href="https://perl6.org:">perl6.org: 3086d62 | (Konrad Borowski)++ | source/compilers/index.html:
Merge pull request #9 from stmuk/master

correct perlito perl6 link
timotimo has highlighted comments and keywords now :3
17:23 sqirrel_ left 17:55 denis_boyun joined 17:57 Adriaaan joined, zakharyas joined
timotimo my brane has the flu. can someone write a perl5 regular expression that accepts variable names? with - and ' and everything? 18:01
18:01 berekuk left, berekuk joined
timotimo preferably also stuff handling foreign characters 18:01
18:13 virtualsue joined 18:16 thou joined
timotimo this is discouraging :| 18:17
i.imgur.com/Ex55atX.png?1 18:19
hoelzro: github.com/timo/perl6-kate-syntax 18:24
jnthn hm, it thinks twigils are operators... :)
timotimo that's how i format it, aye 18:26
but it's not a good idea it seems
i.imgur.com/1pCDtco.png?1 18:28
it also thinks nqp::foo ops are classes :) 18:30
18:31 pippo_ joined
timotimo and it has no grasp of interpolated strings at all etc etc etc 18:31
pippo_ Hello #perl6!
timotimo hey pippo
pippo_ I was trying to use perl6-debug-m but had ===SORRY!=== Could not find Debugger::UI::CommandLine in any of: lib, ... 18:32
Nay clue on what to do to solve the pb?
*Any 18:33
I've compiled perl6 with: perl Configure.pl --gen-nqp --gen-moar --backends=jvm,moar 18:34
timotimo you install Rakudo::Debugger in your lib/
pippo_ No! How can I do it? 18:35
timotimo i think it ought to be enough to just copy the thingie
pippo_ Thank you timotimo! I'll try... &
18:43 sqirrel_ joined
pippo_ timotimo: :-( does not work. #!/usr/bin/perl6-debug-m\n use v6\n say "Hello world";\n ==> Too many positional parameters passed; got 2 but expected 1 18:51
FROGGS there is a ; missing after v6
pippo_: btw, your readallfh segfault should be gone when the moarvm eieio branch gets merged 18:53
pippo_ timotimo: Corrected ';' but same thing
FROGGS++ \o/
timotimo hasn't tried the new Rakudo::Debugger yet :( 18:54
FROGGS pippo_: jnthn++ for that, he refactors how IO happens in moar atm
pippo_ :-)) jnthn++ 18:55
FROGGS: did you use perl6-debug-m ?
19:07 dmol left 19:08 darutoko left
FROGGS pippo_: no, not yet 19:10
but I know it should work
pippo_ Are there instructions available out there to how getting it working? 19:11
jnthn Last time I used it (couple of weeks ago to do demos at FOSDEM) it was just a case of having a lib directory with the deubber UI module and Term::ANSIColor available. 19:12
cooking &
FROGGS I think the ast changes are breaking it 19:13
pippo_ Thank you jnthn. Bon appetit !!
FROGGS pippo_: let me check it locally and then I'll guide you
19:15 beastd joined
timotimo inspire me 19:19
moritz r: say 1.^HOW.defined
timotimo what kinds of things could be simplest to build IDE support for?
camelia rakudo-parrot 06fee3, rakudo-jvm 06fee3, rakudo-moar 06fee3: OUTPUT«===SORRY!=== Error while compiling /tmp/tmpfile␤Cannot use .^ on a non-identifier method call␤at /tmp/tmpfile:1␤------> say 1.^HOW⏏.defined␤ expecting any of:␤ …» 19:20
moritz r: say 1.HOW.defined
camelia rakudo-parrot 06fee3, rakudo-jvm 06fee3, rakudo-moar 06fee3: OUTPUT«True␤»
timotimo huh, that's defined?
r: say 1.HOW.perl
camelia rakudo-parrot 06fee3, rakudo-jvm 06fee3, rakudo-moar 06fee3: OUTPUT«Perl6::Metamodel::ClassHOW.new()␤»
timotimo ah, of course
it's not the type object, it's the meta-object
moritz right
the type object is .WHAT
timotimo the type object is WHAT? 19:21
that passes the attributes to WHO
mumble mumble third base mumble WHENCE mumble
moritz m: say 1.HOW === 2.HOW 19:22
camelia rakudo-moar 06fee3: OUTPUT«True␤»
moritz m: say 1.HOW === (2 but role { }).HOW
camelia rakudo-moar 06fee3: OUTPUT«False␤»
timotimo is this in response to anything? 19:24
moritz timotimo: I'm trying to write about the MOP
(for doc.perl6.org)
timotimo ah
sounds good :)
moritz m: say 1.HOW === Int.HOW 19:25
camelia rakudo-moar 06fee3: OUTPUT«True␤»
19:29 araujo joined 19:30 sqirrel_ left
dalek kudo/nom: 40b76ef | (Tobias Leich)++ | src/perl6-debug.nqp:
align perl6-debug to latest grammar change

statementlist does take an argument now, we just hand that along.		 19:32
FROGGS pippo_: try that ---^ 19:33
also I just put Debugger/UI/CommandLine.pm and Term/ANSIColor.pm in lib/ (you get both from modules.perl6.org) 19:34
timotimo FROGGS: what should my first IDE plugin try to do?
FROGGS timotimo: IDE plugin?
timotimo yes 19:35
QtCreator
i want to experiment with it
moritz provide tab/Ctrl+space completion for identifiers
timotimo and i need some inspiration
FROGGS execute the current script and show its output?
or just show the git diff 19:36
I dunno
timotimo QtCreator already has git support 19:38
FROGGS k 19:39
timotimo downloads a full Qt sdk 19:44
daxim fwiw, Kate is still missing perl6 support (syntax highlighting) 19:57
this is important because these highlighting definitions are reused in other projects
timotimo daxim: i.imgur.com/Ex55atX.png?1 20:00
github.com/timo/perl6-kate-syntax
:)
daxim put it upstream so it appears here <www.kate-editor.org/syntax/update-3.7.xml>
timotimo dude, no way
i only spent like 30 minutes on that so far
i was hoping hoelzro would pick it up and make it perfect :D 20:01
daxim oh, I see.
I'm going to install it for usage 20:02
timotimo please install it for fixage instead :) 20:06
lue Why does the test harness thing decide to fail if all tests pass but there's no plan, again? 20:07
moritz because the specification says that there must be a plan 20:08
(even if it's supplied afterwards) 20:09
mostly to catch things like unintentional uses of exit(0) or silent failures
20:10 grondilu left
moritz m: class A { method x() { } }; say A.^methods 20:10
camelia rakudo-moar 06fee3: OUTPUT«x␤»
moritz m: class A { method x() { } }; say A.^methods(:all)
camelia rakudo-moar 06fee3: OUTPUT«x infinite exists exists_key exists_pos delete delete_key delete_pos list flat eager hash elems end uniq squish pick roll reverse sort values keys kv pairs reduce combinations permutations classify categorize lol map tree Array push unshift grep first join…»
lue Does "plan *" not work? I just tried it and still got failures
moritz it works
unless there's a premature exit 20:11
or unless you'r doing something very wrong
20:12 Ulti left
FROGGS you have to call done() at the end, no? 20:12
p: use Test; plan *; ok 1, 'ohh'
camelia rakudo-parrot 06fee3: OUTPUT«ok 1 - ohh␤»
FROGGS p: use Test; plan *; ok 1, 'ohh'; done
camelia rakudo-parrot 06fee3: OUTPUT«ok 1 - ohh␤1..1␤»
moritz oh right
lue p: use Test; ok 1, 'ohh'; done; 20:13
camelia rakudo-parrot 06fee3: OUTPUT«ok 1 - ohh␤1..1␤»
araujo checks perl6 to see what's new in the community 20:15
timotimo heyo araujo :)
araujo hi timotimo :)
timotimo we're kind of excited about MoarVM
last week i started a "weekly perl6 changes" blog that may be of interest to you
(also since i told you about it now, i'll have more reason to keep it up :) )
araujo timotimo, any nice link at hand to update me?
20:15 thou left
araujo sure :) 20:15
timotimo 6guts.wordpress.com/ ← always an excellent source for happy thoughts 20:16
lue I personally think it's wrong for panda to refuse to continue when the test pass, but there's no plan-like thing, but I'll play along with it.
araujo timotimo, thanks!
timotimo p6weekly.wordpress.com/ ← this is my blog
araujo checks 20:17
timotimo, wow, that's nice... will you update weekly? 20:19
timotimo araujo: hopefully!
araujo hehe, nice, useful for people like me .. who just glances at perl6 from time to time :P 20:20
timotimo xfix aka GlitchMr used to do one of those, but his posts were more concise and a bit more technical 20:21
i'm hoping i can keep up the energy to write about changes in a bit more detail and give outlooks each week
moritz m: say 1.^mro
camelia rakudo-moar 06fee3: OUTPUT«(Int) (Cool) (Any) (Mu)␤»
moritz r-m build fails with "Unexpected named parameter 'deserialize_past' passed" during parse stage of the setting 20:24
oh, I forgot to pull
20:24 virtualsue left
moritz never mind 20:24
jnthn timotimo: Is Rakudo HEAD fixed against NQP master yet? 20:25
FROGGS yes
the debugger is fixed too
dalek c: 1c79ebf | moritz++ | lib/objects.pod:
start to talk about introspection and meta objects
 20:26
jnthn \o/
timotimo jnthn: yup
moritz lots more to write :-) 20:27
FROGGS moritz++ 20:28
20:29 telex left 20:30 telex joined 20:35 zby_home_ joined 20:46 SamuraiJack left
pippo_ FROGGS: FROGGS++. I works with your patch !!! Thank you very much!! 20:48
FROGGS pippo_: cool! 20:49
pippo_: have fun hacking! :o)
pippo_ :-)) Yes indeed with perl6 and its community it is really a fun programming :-)) 20:50
FROGGS yes, absolutely
pippo_ :-) 20:51
TimToady 'course, our idea of fun is a bit masochistic at times...
or perhaps, stoic...
dalek c: d0ea532 | Mouq++ | lib/Version.pod:
Add missing TITLE to Version.pod
pippo_ TimToady++ 20:52
timotimo "This is an action from MyPlugin!" 20:57
\o/
FROGGS \o/ 20:58
timotimo now i'm basically standing in front of a big wall :\ 20:59
FROGGS haha
yeah
a 'Toady
masak .oO( "if I have seen further, it's only because I have been standing on the shoulders of big Walls." ) 21:03
TimToady --Humpty Dumpty
21:04 thou joined
masak :P 21:08
pippo_ jnthn++ for perl6-debug-m 21:10
21:14 virtualsue joined
dalek p: c64c70f | jnthn++ | / (2 files):
Map socket ops for Moar; get Moar with sockets.

Or at least, client sockets.		 21:15
21:22 hoverboard is now known as YOLOBIRD
dalek kudo/nom: f35ca0b | jnthn++ | / (4 files):
Get client sockets support on Moar in place.

This also takes us to a MoarVM with heavily refactored/improved IO; of note, line-level reading should be more efficient, and cases where we read characters split over packet/buffer boundaries should be handled correctly.		 21:24
FROGGS cool! 21:25
timotimo hunts bumped revisions 21:32
21:35 ggherdov left 21:36 ggherdov joined 21:46 smls left 21:56 ajr_ left
pippo_ FROGGS: indeed the above patch from jnthn solved the problem of qx[] on my machine. jnthn++ 22:03
jnthn pippo_: yay :) 22:05
pippo_ :-) 22:06
FROGGS \o/ 22:07
timotimo i think the first "plug-in" to qtcreator for perl6 support will actually just be a wizard that creates projects/classes/modules for you 22:08
and after that probably a "pipe selection through a perl6 program" menu entry or something 22:09
and then perhaps "search with p6 regex"
the possibilities! they are endless! :)
22:11 zakharyas left 22:15 YOLOBIRD is now known as hoverboard 22:17 kaare_ joined, colomon left 22:18 smls joined
smls .tell raydiak I've made some mockups showing how I think the structure and design of doc.perl6.org could be improved further: imgur.com/a/y1dHo -- What do you think? 22:19
yoleaux smls: I'll pass your message to raydiak.
smls ^^ this also goes to everyone else who is interested in docs.perl6.org
timotimo ooooh, damn 22:20
that looks pretty
FROGGS ohh yes 22:22
jnthn smls++ 22:24
lue smls++ though the book icon doesn't fit with Camelia :) 22:28
smls I just took the address book icon from from the Tango icon theme, and tweaked it minimally... :P
if someone has the skills to create a custom icon, that would of course be better 22:29
timotimo bedtime o/
questhub.io/realm/perl/quest/52ffea...06670000e7 ← thoughts and comments very much appreciated 22:30
FROGGS gnight timotimo
pippo_ o/ timotimo
jnthn 'night, timotimo++ 22:34
22:35 pippo_ left 22:41 dmol joined 22:50 hoverboard left 22:54 dalek left 22:55 dalek joined, ChanServ sets mode: +v dalek 22:56 masak left, PerlJam left, PerlJam joined 22:57 masak joined, [Coke] left, [Coke] joined 23:06 hoverboard joined
jnthn What on earth... 23:10
The socket tests, on the server side, do this:
while my $client = $server.accept() { 23:11
...
FROGGS that is interesting
jnthn accept blocks. The client connects, tthe server replies. Then it goes and waits for another connection...which of course we're never going to make.
FROGGS a bit strange, but interesting
jnthn Because all the tests connect once.
On Windows, I think there's no timeout mechanism. 23:12
So the test will hang forever. Correctly, afaict.
TimToady well, but that's just how accepts always worked, originally
jnthn TimToady: Blocking? Yes.
TimToady: I'm not questioning accepts. I'm questioning why the tests look like this.
TimToady nodnod 23:13
jnthn TimToady: As in, "how could these tests have ever been right"?
TimToady is there any other way to terminate the proggie?
jnthn Well, we only ever send it one connection 23:14
The tests can still pass on anything that runs the .sh 23:15
'cus it contains a bunch of stuff to kill after 45s.
Conveniently, the tests then allow things to take 60s.
So anyway, we might have solved the mystery of why the sockets tests are so darn snow... :)
uh, slow
23:17 zby_home_ left
jnthn oh, hmm 23:17
both should unconditionally kill the server if the client exits. 23:18
masak .oO( you know nothing, Jon Slow ) 23:22
jnthn FROGGS: Still about? 23:28
23:32 virtualsue left 23:41 cognominal joined, smls left 23:42 Adriaaan left 23:43 xinming left, hoverboard left 23:44 xinming joined 23:54 denis_boyun left 23:58 spider-mario left, araujo left

Powered by Cro and the Raku® Programming Language.

Please report any issues / comments / feature requests as an issue on App::Raku::Log.

Thank you!