Perl 6 - the future is here, just unevenly distributed

IRC log for #rosettacode, 2011-07-16

| Channels | #rosettacode index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:17 BenBE Hi mikemol
00:18 BenBE Regarding your mail: Could you check which language file it's referring to?
00:32 thundergnat left #rosettacode
00:39 realazthat_ joined #rosettacode
00:43 realz left #rosettacode
00:43 realazthat_ left #rosettacode
00:43 realazthat_ joined #rosettacode
00:56 ttmrichter joined #rosettacode
01:11 kpreid left #rosettacode
02:11 Mathnerd314 joined #rosettacode
02:24 kpreid joined #rosettacode
03:47 BenBE left #rosettacode
03:51 BenBE joined #rosettacode
04:16 mikemol BenBE: I don't have that data. The best I have are those referrers, it might be possible to narrow it down based on which languages are present on all those pages.
04:18 mikemol Or, rather, the pages referred to by all those referrers. So, kinda difficult to narrow down.
04:27 CRGreathouse[RCo joined #rosettacode
04:33 kpreid left #rosettacode
04:43 BenBE left #rosettacode
04:47 BenBE joined #rosettacode
04:53 kpreid joined #rosettacode
05:24 CRGreathouse[RCo left #rosettacode
05:59 dagnyscott left #rosettacode
06:33 Mathnerd314 left #rosettacode
07:07 r4m left #rosettacode
07:08 r4m joined #rosettacode
08:01 realazthat joined #rosettacode
08:01 realazthat_ left #rosettacode
08:02 realazthat left #rosettacode
08:02 realazthat joined #rosettacode
10:34 BenBE mikemol I guess something like PariGP but i couldn't reproduce it locally ...
12:17 Util left #rosettacode
12:19 Util joined #rosettacode
12:51 dagnyscott joined #rosettacode
14:39 BenBE $case_sensitive = $this->language_data['CASE_SENSITIVE'][$k];
14:39 BenBE Now lets find the offending language ;-)
14:39 BenBE And luckily this should be visible through the langcheck ^^
14:54 BenBE Interestingly: There's no language file wrong according to the langcheck.
14:57 BenBE mikemol Any idea where we could get some more information which language file is causing the warning?
15:14 BenBE I propose a new RC task: Find the Programming Language that is causing the following problem given the following list of referrers: http://slexy.org/view/s20Nt2X4bn
15:14 fedaykin "List of referrers that were seen when accessing a page causing trouble // Slexy 2.0"
16:20 ttmrichter OK, all y'all might be getting a bunch of SNOBOL4 entries soon.  I emailed the community and it seems at least one has taken up the challenge.  :)
16:20 mwn3d_phone Whoo!
16:20 mwn3d_phone o/
16:28 TimToady ttmrichter: you might be interested in the forced use of a cons list in http://rosettacode.org/wiki/Singly-linked_list/Traversal#Perl_6
16:28 fedaykin "Singly-linked list/Traversal - Rosetta Code" http://rldn.net/14vHe
16:29 TimToady as with Lisp, of course, a cons list is mutable, and does run into your reversal limitations to do right association
16:32 TimToady which are two good reasons to encapsulate lists behind an abstraction layer, as most FP languages try to do, with varying levels of success
17:19 ttmrichter left #rosettacode
17:28 kpreid left #rosettacode
17:52 mwn3d_phone left #rosettacode
17:58 mwn3d_phone joined #rosettacode
18:14 kpreid joined #rosettacode
19:15 mbishop man that was a weird error...that pythagorean triples task helped find a bug in cm3, a modula-3 compiler...turns out the bug was in in some code that converted the "L" literal for LONGINT into regular INTEGER, and someone typo'd one of the numbers
19:15 mbishop so 6L became 4, and 4 became 0
19:15 mbishop in the format code
19:17 mikemol BenBE: I'll give it some thought. I've got ideas, but it'd involve making PHP throw exceptions on such notices, and I'd then have to do a *lot* of cleanup of existing code to catch those exceptions.
19:32 BenBE Well, you could basically start with a small patch at geshi.php line 3383 logging some notice if CASE_SENSITIVE[$k] is undefined and log the language name.
19:35 BenBE Another thing: Could you do a file_put_contents($langname.'/'.md5($source)) when highlighting an uncached entry? (As RC has loads of example sources accross languages this should be quite an easy way to get a list of source samples ;-))
19:43 mwn3d_phone mikemol: not that it'd help much with this problem but there might be a PHP upgrade available too. Another thing on the pile.
20:15 mikemol mwn3d_phone: Actually did the php upgrade yesterday.
20:17 Coderjoe i'm going to go out on a limb and say the language is a basic derivative, possibly liberty basic
20:17 mikemol BenBE: The code you put in geshi-live is run live, so put whatever you need in there. Re the source examples, just keep in mind that the code is licensed GFDL.
20:17 Coderjoe hmm
20:17 Coderjoe or not. referred to from there
20:19 Coderjoe can we get a list of pages with the error as well as the list of referrers?
20:20 Coderjoe or perhaps someone check which of these two pages is on the list of errored pages:
20:22 Coderjoe Query_Performance (or Time_a_function), Compare_sorting_algorithms'_peformance (or Measure_relative_performance_of_sorting_algorithms_implementations)
20:27 mikemol Coderjoe: Everything I have is in what BenBE posted. (I assume)
20:27 mikemol I don't know the URIs of the pages which triggered the error.
20:28 thundergnat joined #rosettacode
20:32 thundergnat I may be totally out in left field here, but I wonder if the errors are caused by a bad interaction with SVG files that have <text> elements in them? I had problems over the past couple of day where a SVG with text elements have errors while generating previews.
20:34 TimToady thundergnat: I tweaked your queue a bit
20:34 thundergnat Ouch! my queue!!
20:34 thundergnat :)
20:39 Coderjoe binary search reveals: http://rosettacode.org/wiki/User:Coderjoe/Sandbox2
20:39 fedaykin "User:Coderjoe/Sandbox2 - Rosetta Code" http://rldn.net/EmX
20:40 Coderjoe BenBE: it appears to be Pari/GP, based on my binary searching
20:43 mikemol thundergnat: Different part of the codebase. These errors come from within GeSHi, which handles data found <lang>in here</lang>. The SVG is through MW's file upload mechanism, which is totally separate.
20:44 thundergnat Apparently, I'm WAY out in left field. :-/
20:45 mikemol thundergnat: np. Not much I can do about your svg issues, either, until I can update rsvg.
20:46 Hypftier You could switch to Inkscape or Batik :P
20:46 * Hypftier hides
20:46 * mikemol is tempted to play with /kick right about now...
20:46 mikemol It took quite a while to find a usable svg renderer.
20:47 mikemol Inkscape and some java lib are reputed to be the two best renderers.
20:47 Coderjoe indeed it is. Pari/GP's file has keys of 1,2,3 under keywords, but only GESHI_COMMENTS, 1, 3 under 'CASE_SENSITIVE'
20:47 mikemol Inkscape would require me to have X libs(!) on the server.
20:47 thundergnat Not a problem. I am certainly not complaining. And Hypfier was kind enough to help me out.
20:47 Hypftier Batik is said Java lib. And yes, it's probably one of the most complete SVG implementations out there.
20:47 mikemol I was going to have a guy write a good daemon for me to use the java lib without killing my server resources, but that didn't wind up happening.
20:48 mikemol I should poke him again.
20:50 Hypftier I still wonder why rsvg failed to render the SVG. The Wikimedia projects use it too and they only have the usual share of rsvg bugs; text rendering works fine. Although it crashed here too, but I'm tempted to say that's a broken Windows port of rsvg-view (and probably a horribly old library).
21:15 BenBE mikemol As long as I don't link against the source I don't have to have it be licensed GPL compatible.
21:16 BenBE CodeBlock I currently suspect PariGP or PicoLisp to be the culprits. See the list of referrers I put in the channel earlier.
21:16 BenBE CodeBlock Unfortunately we only have the list of referrers ATM.
21:18 BenBE mikemol Jep. I posted everything you mailed me, except for the exact paths of the error message (to protect the server system's internals)
21:19 BenBE Okay, I'll try to update some files on the server.
21:19 CodeBlock ?
21:19 * CodeBlock growls. Learn to tab complete, people. :(
21:21 BenBE CodeBlock Sorry ;-)
21:23 BenBE k, upload of my updated files started.
21:24 BenBE The files I'm just uploading are clean according to the LangCheck on my dev system: http://benbe.home.omorphia.de:43815/geshi-misc/profiling/geshi-trunk/contrib/langcheck.php
21:25 fedaykin "GeSHi Language File Validation Script" http://rldn.net/HoAc
21:26 Coderjoe BenBE: this one sets it off: http://rosettacode.org/wiki/User:Coderjoe/Sandbox2
21:26 fedaykin "User:Coderjoe/Sandbox2 - Rosetta Code" http://rldn.net/EmX
21:26 Coderjoe (before your updates, at least)
21:27 BenBE Currently updating the language files.
21:27 BenBE Currently at diff.php
21:34 BenBE mikemol Is it okay if I put the following lines in geshi.php (Function parse_code()? http://slexy.org/view/s20NeZYJVr
21:35 Coderjoe does anything ever collect the garbage?
21:36 BenBE Coderjoe I'll only leave it in for about a week, grab the source snippet folder than and remove the snippet again.
21:37 BenBE I currently already have a small collection in the GeSHi code repository, but that is much too incomplete for one script I wanted to develop from it.
21:38 BenBE Thus the aim basically is to get a plain dump of source snippets I can easily feed to a script (also in the GeSHi SVN) to auto-learn the various languages for detection.
21:38 Coderjoe ah
21:38 BenBE If you like I can make the whole collection available under the GFDL for download.
21:42 Coderjoe it isn't my say. I was just concerned about disk space usage
21:44 BenBE It shouldn't be much more than taken up by the database already. (upper bound).
21:45 BenBE That's also why I'm waiting for confirmation by mikemol.
21:45 BenBE Because this also introduces some additional IO to the disks which may put some strain on the server performance.
21:48 BenBE Coderjoe Could you recheck if the Notice is gone now?
21:49 Coderjoe it appears to be gone
21:49 BenBE Okay. Sounds good.
21:49 Coderjoe last one was at 21:22:47 UTC
21:50 Coderjoe (and I forced my sandbox page to refresh
21:50 BenBE Okay, seems to be fixed then.
21:51 FireFly joined #rosettacode
22:03 BenBE Could someone have a look: http://rosettacode.org/wiki/User:BenBE --- Somehow some languages are missing ;-)
22:03 fedaykin "User:BenBE - Rosetta Code"
22:06 mwn3d_phone BenBE: you have to spell and capitalize them right
22:06 mwn3d_phone Look at the languages category listing for the right names
22:07 mwn3d_phone E.g.: we censor "brainfuck" as "brainf***"
22:07 mwn3d_phone I know its lame to censor things but we wanted to make sure we wouldn't get blocked by academic networks
22:08 Hypftier I hope it's more schools than actual universities ... if the latter censor then somewhere education is in a very weird shape ;)
22:09 mwn3d_phone I don't think we got reports of it ever
22:09 mwn3d_phone Just trying to be proactive
22:13 BenBE k, fixed some ...
22:13 BenBE But still missing the rest.
22:31 BenBE Okay, just added the Delphi solution for the Ackermann Function: http://rosettacode.org/wiki/Ackermann_function#Delphi
22:31 fedaykin "Ackermann function - Rosetta Code" http://rldn.net/SNFS
23:01 FireFly left #rosettacode

| Channels | #rosettacode index | Today | | Search | Google Search | Plain-Text | summary