Perl 6 - the future is here, just unevenly distributed

IRC log for #rosettacode, 2011-07-19

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

All times shown according to UTC.

Time Nick Message
00:55 kpreid left #rosettacode
01:01 kpreid joined #rosettacode
01:27 kpreid left #rosettacode
01:44 mwn3d_phone left #rosettacode
01:50 mwn3d_phone joined #rosettacode
01:52 BenBE Really interesting read!
01:52 BenBE Thanks for this link.
02:19 kpreid joined #rosettacode
02:19 kpreid left #rosettacode
02:19 kpreid joined #rosettacode
02:52 mwn3d_phone left #rosettacode
03:39 mwn3d_phone joined #rosettacode
03:49 BenBE left #rosettacode
03:53 BenBE joined #rosettacode
04:16 mwn3d_phone left #rosettacode
04:31 mwn3d_phone joined #rosettacode
04:35 kpreid left #rosettacode
04:36 kpreid joined #rosettacode
07:55 Util left #rosettacode
07:57 Util joined #rosettacode
10:49 FireFly joined #rosettacode
12:11 FireFly left #rosettacode
12:22 rodt joined #rosettacode
12:28 FireFly joined #rosettacode
12:32 dagnyscott joined #rosettacode
13:13 dagnyscott1 joined #rosettacode
13:13 dagnyscott left #rosettacode
13:13 dagnyscott1 left #rosettacode
13:14 dagnyscott joined #rosettacode
14:07 rodt hmm im thinking of plugging my language on rossettacode, but it looks like it only supports old linear squencial textual programming languages
14:10 rodt how would put this : http://imagebin.org/163485 on the hello world page ?
14:10 fedaykin "Imagebin - A place to slap up your images."
14:19 mwn3d_phone rodt: we have piet and labview examples on the site
14:19 mwn3d_phone They're graphical
14:21 rodt ah yeah sorry didnt notice them ;)
14:21 rodt nice one :D... i need to bulk out my lang anyway ;)
14:32 mikemol rodt: Interesting interface. Nice, clean. Should compress with GIF or PNG very nicely.
14:33 BenBE Looking forward to the GeSHi language file for it ;-P
14:33 mikemol rodt: Recommendation: Use as small an image dimension as reasonably possible; images that large won't flow well on most screens.
14:34 mikemol rodt: One other thing...consider an SVG export?
14:35 mikemol It looks like you're using vector ops to draw your UI, so the translation should be fairly natural.
14:36 mikemol Though to answer your initial question: http://rosettacode.org/wiki/Special:Upload
14:36 fedaykin "Upload file - Rosetta Code"
14:36 mikemol Then take a look at how example sections like this are constructed: http://rosettacode.org/wiki/Four_bit_adder#LabVIEW
14:36 fedaykin "Four bit adder - Rosetta Code" http://rldn.net/UQAt
14:50 BenBE Coderjoe Because you asked regarding the memory usage: Currently about 12.5 MiB in 27k Files.
14:58 BenBE About 2.4 MiB in 7.3k Files are sorted as "text" thus could theoretically be used to check if unsorted things are detected properly.
14:58 BenBE Or to check how stable detection is.
15:05 kpreid left #rosettacode
15:05 kpreid joined #rosettacode
15:08 kpreid left #rosettacode
15:51 trn left #rosettacode
16:14 kpreid joined #rosettacode
16:25 rodt thanks for all the help :), yeah its vector graphics anyway... i need to fix up a few things with it first anyway, sorry for late respone
16:44 realazthat is now known as trolzies
16:48 trn joined #rosettacode
16:54 axisys_ joined #rosettacode
16:59 r4m left #rosettacode
16:59 axisys left #rosettacode
17:00 busbusbusbus joined #rosettacode
17:05 busbusbusbus left #rosettacode
17:06 busbusbusbus joined #rosettacode
17:30 trn left #rosettacode
17:37 Coderjoe BenBE: hmm... perhaps you should have done some multilevel, like snippets/LANG/first2ofmd5/allofmd5. if there are too many files in one directory, many filesystems get a bit unhappy and slow down when accessing anything under that directory
17:37 busbusbus joined #rosettacode
17:38 busbusbusbus left #rosettacode
17:42 trn joined #rosettacode
17:57 mikemol No noticeable I/O delays right now.
18:07 trn left #rosettacode
18:26 trn joined #rosettacode
19:05 BenBE CodeBlock Ext3 and Ext4 aren't affected AFAIR. Was a problem with VFAT and FAT32.
19:06 CodeBlock ...
19:07 sorear was a problem with original ext2 also
19:09 sorear most early filesystems stored directory entries in an unsorted list
19:09 sorear newer filesystems use various forms of hash table or B-tree
19:11 sorear ext2 was retrofitted to support "directory hashing" via the dir_index filesystem option
19:12 mikemol It's still a problem on ext3 and ext4, to an extent. I had a directory with a few hundred thousand files in it, and touching that directory at all could be a slow affair.
19:12 mikemol Hm. Not sure if I had dir_index enabled.
19:48 sorear debugfs /dev/...; show_super_stats -h might DWYW
19:58 mikemol Dont have that folder any more. But I know my way around tune2fs. :)
20:00 sorear it's done on a filesystem basis...
20:00 * mikemol nods
20:00 mikemol tune2fs with the right parameters on the device with the filesystem in question would do what I'd need to do.
20:01 sorear I think lsattr -d /path will tell you +I or -I depending
20:01 mikemol k
20:01 * sorear doesn't have any ext*fsen on this box to test
20:02 Coderjoe I don't know what filesystems were used on each system, but some people doing archive work were seeing significant slowdowns (or outright failures, iirc) with a directory with several tens of thousands of subdirectories
20:02 Coderjoe (might have been hundreds of thousands)
20:03 Hypftier does accessing a file by name directly even care about the number of items in a directory?
20:03 sorear Hypftier: it does, if you're using an old filesystem
20:03 Hypftier (Ok, I'm on NTFS which usually copes well with 1e5 files in a directory, so no idea)
20:04 Coderjoe it would depend on how the directory's entries are stored
20:04 sorear Hypftier: the original nameless unixfs kept directory entries in an unsorted list, as do most of its imitators
20:05 Hypftier hmm, although BenBE mentioned something of 12 MiB of data so far. Which should fit in almost any cache, I guess.
20:13 BenBE I pulled a tar.gz which was about 6MB. But given TAR organizes small files in blocks most of the thin air came from the TAR structure.
20:18 BenBE Okay: WinRAR gives 41% remaining archive size for normal operation and 30% remaining size for solid archive (place everything after another and compressing without boundaries).
20:49 FireFly left #rosettacode
20:52 FireFly joined #rosettacode
20:58 dagnyscott left #rosettacode
21:05 busbusbus is now known as r4m
22:11 rodt left #rosettacode
23:30 axisys_ left #rosettacode
23:30 axisys joined #rosettacode

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