Perl 6 - the future is here, just unevenly distributed

IRC log for #rosettacode, 2011-01-24

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

All times shown according to UTC.

Time Nick Message
00:00 BenBE And? Any results?
00:20 mwn3d_phone left #rosettacode
00:21 mwn3d_phone joined #rosettacode
02:18 shortcircuit BenBE: Strangely, nothing IPv6 worked. I'll have to look into it at some point, but not right now.
02:19 shortcircuit However, ping6 -I eth0 -c 2 ff02::1 from a VPS node is entertaining. :)
02:25 BenBE -c 2 ???
02:26 BenBE OT: I kinda like http://ripe.net/projects/ris/rawdata.html ... Unfortutnally the whole dataset might be around 300TB
02:26 fedaykin "Projects - RIS"
02:29 shortcircuit Restarting squid and apache2; that crawl left me with 201MB of stuff in swap.
02:30 BenBE My server currently is at 1910/3957MB of RAM usage, no swap used.
02:34 shortcircuit BenBE: -c 2; limit to two ping packets.
02:34 BenBE Ah.
02:35 BenBE BTW: I just got ping on Windows to return print "Time=-1ms" yesterday ...
02:35 BenBE IDK what I did, it just printed negative response times for my file server sometimes ...
02:36 shortcircuit ff02::1 is a link-local multicast address, and everything on the local link listens to that address. ping6 isn't smart enough to realize that responses to that packet from different link-local addresses aren't duplicates.
02:36 shortcircuit If you told it -c1, it would quit after the first reply it got. By telling it c2, it stays open and listens for a little bit longer, so you see the multitude of responses.
02:36 shortcircuit BenBE: Multicore box?
02:37 BenBE Yes. Sending and receiving box.
02:38 shortcircuit BenBE: Probably a bug in the driver for the high-precision clock mechanisms on the Windows machine.
02:39 BenBE hmmm, that's quite a nice way of scanning a network for hosts :P
02:40 shortcircuit BenBE: Yes and no; if a host isn't configured to respond to ICMP echo (er...ICMPv6 echo, more specifically), you won't see it.
02:40 BenBE If you want some load:
02:40 BenBE ping6 -I eth0 -c 2 ff02::3
02:40 BenBE Be careful! REALLY produces QUITE some load ;-)
02:41 shortcircuit Still, it's the most effective solicited discovery approach I know of; I wanted to figure out which of my network's systems had any IPv6 functionality enabled.
02:42 shortcircuit BenBE: Re the -1ms ping time...From a low-level perspective, my favorite thing to point to on the subject is: http://blogs.msdn.com/b/oldnewthi​ng/archive/2005/09/02/459952.aspx
02:42 fedaykin "Precision is not the same as accuracy - The Old New Thing - Site Home - MSDN Blogs" http://rldn.net/EuC
02:44 BenBE The RIS data I linked above might be an interesting research subject for compressing huge amounts of data while still being able to allow for quick access ...
02:44 BenBE What do you think?
02:45 shortcircuit I think I might have missed the link.
02:45 BenBE http://ripe.net/projects/ris/rawdata.html
02:45 fedaykin "Projects - RIS"
02:46 BenBE That's a set of collected information on the routing of different AS on the internet.
02:46 shortcircuit Ah. Dunno. I'd have to look at it and think about it.
02:46 shortcircuit I've still got 184MB in swap, and I don't know what's holdin git.
02:47 shortcircuit er. holding it.
02:47 BenBE IDK. Maybe some unused file buffers?
02:47 shortcircuit some was memcached. Some was mysql. Down to 17M.
02:48 shortcircuit That crawler really sucked.
02:48 shortcircuit I think the remainder is probably irssi. That's not so bad.
02:48 BenBE Well, at least your cache of the wiki is now up-to-date ;-)
02:49 shortcircuit BenBE: Yes and no; squid caches a separate object for each supported client-side encoding string.
02:49 shortcircuit Not client-side encoding _type_, but string. That means that something like 'zlib; deflate'
02:49 shortcircuit er
02:50 shortcircuit 'zlib; deflate' will be a separate cached object from 'zlib;  deflate', even though they symbolicly resolve to the exact same thing.
02:50 shortcircuit So the only cache that's up to date is the exact encoding string the httrack instance specified.
02:51 shortcircuit (Incidentally, the string vs symbolic meaning difference is the same reason MediaWiki has a hard time purging squid caches; it has to specify the encoding string exactly, too.)
03:42 Coderjoe which is stupid
04:14 SoniaKeys joined #rosettacode
04:22 Mathnerd314 left #rosettacode
06:58 Polynomial left #rosettacode
07:05 Polynomial joined #rosettacode
07:28 MigoMipo joined #rosettacode
07:35 MigoMipo left #rosettacode
08:14 SoniaKeys left #rosettacode
11:22 Coderjoe_ joined #rosettacode
11:22 Coderjoe left #rosettacode
13:28 FireFly joined #rosettacode
14:40 MigoMipo joined #rosettacode
14:58 Polynomial left #rosettacode
15:06 parsleyfirefly left #rosettacode
15:21 parsleyfirefly joined #rosettacode
15:56 parsleyfirefly1 joined #rosettacode
15:57 parsleyfirefly left #rosettacode
16:01 parsleyfirefly1 left #rosettacode
16:01 parsleyfirefly joined #rosettacode
16:30 parsleyfirefly left #rosettacode
17:16 Polynomial joined #rosettacode
17:53 parsleyfirefly joined #rosettacode
18:46 slavik joined #rosettacode
18:46 shortcircuit slavik: It's not outside the scope of this channel, though. :)
18:46 slavik ^^
18:46 slavik java developers are (on average) the dumbest people in the world!!!
18:46 shortcircuit However, I'm a tad swamped at work, so my participation will be slow, at best.
18:46 shortcircuit Ok, _that's_ probably a little OT. :)
18:47 slavik and questions like "can you tell us if our objects are collected on every gc cycle." should get an "architect" fired
18:47 slavik I'm done ranting
18:47 shortcircuit Hehe
18:48 shortcircuit There's at least one low-level JVM coder in here.
18:48 * shortcircuit looks at kpreid
18:48 kpreid not me!
18:48 opticron slavik, I've encountered dumb java programmers as well...
18:48 slavik but I am of the opinion that java (the language) teaches people to be stupid ... because if it manages memory, why do I have to restart shit that goes OOM and doesn't recover?
18:49 shortcircuit slavik: Also, note that the channel is publicly logged
18:49 slavik shortcircuit: did I say something wrong? :P
18:49 shortcircuit slavik: Not formally, I don't think. :)
18:50 shortcircuit slavik: Is the complaint with Java, or with gc?
18:50 slavik here's another one: we will write a j2ee app that runs inside a j2ee server, but we refuse to use the jdbc services from the j2ee server and will instead call the db driver directly
18:50 slavik more with java causing dumb developers ...
18:50 shortcircuit Hm.
18:51 kpreid slavik: recovering from OOM is a Very Hard Problem; when attempted, often done wrongly
18:51 shortcircuit Languages affect the way a programmer thinkgs about a problem, but I don't know that languages create poor programmers.
18:51 kpreid failing at large granularity is more robust
18:52 shortcircuit At worst, a language tends to make a programmer complacent about certain things the language seems to make easy.
18:52 kpreid now if you like you can ... ah, I don't have time to argue this right now. got to go.
18:52 slavik kpreid: also: try {} catch (e) { printStacktrace(); } is a horrible way to handle exceptions imo
18:52 kpreid left #rosettacode
18:52 slavik shortcircuit: that's exactly the point
18:53 slavik java teaches people to not care what the jvm does
18:53 slavik it's like the difference between range() and xrange() in python
18:55 shortcircuit I see two sides to the issue. One, depending on language features to fight lazy programmers is fighting human nature itself. On the other side, depending on a language's behavior on a particular platform is a different problem than writing in the language itself.
18:56 shortcircuit For example, my day job is coding to C++. I pretty much always code to some x86 variant. My code's always compiled to target 32-bit x86, but may eventually be compiled for 64-bit x86.
18:57 slavik I think the former is due to use of a statically  typed language to write "business logic" ... and I hate that term, too.
18:57 shortcircuit I need to avoid making assumptions about how C++ behaves based on my experience with C++ on 32-bit x86, as I face a very strong risk of then writing code that fails badly on 64-bit platforms.
18:58 mwn3d_phone left #rosettacode
18:58 mwn3d_phone joined #rosettacode
19:02 mwn3d_phone slavik: don't sell code that just prints the stack trace but if you're working to figure something out the stack trace is a good place to start
19:02 mwn3d_phone And its frequently good in logs along with some short messages
19:02 slavik mwn3d_phone: that's my point, add a message or arguments that you were called with, but often times, the devs just dump the stacktrace ...
19:03 slavik it's like,. I know your code is fubar, but you don't put enough info there to fix it
19:05 kpreid joined #rosettacode
19:06 mwn3d_phone For some errors the exception type and line number are enough. There's definitely a place for it but adding an extra message is almost never a bad idea.
19:09 slavik mwn3d_phone: right, but if you are trying to communicate to outside, knowing where you are going to is nice to know along with the exception.
19:10 TimToady supposing I want to give talks that have many links into rosettacode, but want a local copy so I'm not dependent on network access.  What's the easiest way to clone RS locally?
19:10 TimToady (that maintains style info)
19:11 mwn3d_phone There's an XML export but I'm not sure what that does with style.
19:11 shortcircuit TimToady: Probably something like Firefox's save page, for the relevant pages.
19:11 TimToady what I'd really like is something like a local mirror
19:11 shortcircuit RC doesn't have a good, functioning export mechanism. :-|
19:11 TimToady seems like it'd be useful for a failover if your site went down, too
19:12 shortcircuit Yeah
19:12 TimToady maybe something based on rsync?
19:12 shortcircuit I've got a nightly backup which has mysql dump the backup in XML format. I've been meaning to script up a sanitizer to remove personal info from it, and publish the backups.
19:13 TimToady anyway, I've tried the FF save route, and it's quite tedious, esp if you're not sure in advance which links you'll want, so you have to save them all...
19:13 TimToady and none of the second-order links work right
19:13 * shortcircuit nods
19:13 TimToady and recursive wgets also don't work right
19:14 shortcircuit Understood.
19:14 * TimToady is leaving for India on Thu to give a talk, y'see
19:15 TimToady and the FF saves and wgets I did have all disappeared in my last disk crash, so not looking forward to doing it again...
19:15 shortcircuit Let me see if I can do a manual sanitization of one of the most-recent backups. It'll have to wait until tomorrow-after-work, though.
19:16 TimToady thanks
19:19 shortcircuit One of the things I've wanted to do, in order to get more server-admin help with RC, is set up a script that uses sanitized versions of the automated backups to create local site instances.
19:20 TimToady sorry for the short notice, and I will understand completely if we don't get there this week
19:20 shortcircuit Doing so from a non-sanitized backup would make spinning up fallback nodes a fairly quick process.
19:20 slavik wow ... rc is awesome!!!!
19:20 slavik shortcircuit: I blame you for not inviting me here before! :P
19:21 TimToady we've been dangling the bait in front of you for several months now...
19:22 slavik :(
19:22 slavik you're mean :(
19:23 TimToady http://irclog.perlgeek.de/search.pl?cha​nnel=perl6&nick=&q=rosettacode # even meaner than you think :)
19:23 fedaykin "IRC logs - Search" http://rldn.net/49m9
19:24 shortcircuit slavik: Two primary rules on RC, btw. First, I explicitly allow and endorse language advocacy, as long as it comes in the form of code. Second, whatever I say, goes, unless I change my mind.
19:24 shortcircuit I try not to interfere too much, though; the community works best when left to its own devices.
19:24 slavik shortcircuit: how does that differ from what TimToady's rules about Perl? :P
19:26 shortcircuit slavik: Answering that question would require me to be familiar with his rules on Perl. :)
19:26 slavik 1. He is always right, 2. in cases of conflict, see #1
19:26 slavik he == TimToady
19:26 TimToady no, the second rule is, if Larry changes his mind, he's still right.
19:27 shortcircuit Hehe. I just figure I'm usually wrong, which is why I usually keep my mouth shut.
19:27 slavik TimToady: I would argue the meaning is the same between the two #2 rules :)
19:27 TimToady that's the assumption behind leaving the door open to changing one's mind :)
19:27 shortcircuit I try to only step in when things seem to drift away from the direction I'd prefer, and when things get heated to the point of uncivility.
19:28 TimToady you #2 doesn't imply that I ever change my mind
19:28 TimToady *your
19:28 slavik ahh
19:28 slavik I see, in that case, I conceed.
19:28 slavik concede
19:28 shortcircuit I'm also prone to talking to myself, once I get on a roll...
19:29 shortcircuit (Which is another good reason to keep my mouth shut where I'm likely to be wrong. :) )
19:29 TimToady shortcircuit: yes, all successful dictators know that they really can only dictate what people want already most of the time, and violate that only occasionally, at the risk of their dictatorship.
19:29 slavik shortcircuit, we are not so different, you and I. :)
19:30 shortcircuit slavik: http://tvtropes.org/pmwiki/pm​wiki.php/Main/NotSoDifferent ?
19:30 fedaykin "Not So Different - Television Tropes & Idioms" http://rldn.net/Bll
19:32 shortcircuit TimToady: Success was a happy accident; I just didn't have time to get actively involved as much as I wanted, and had to check myself when I noticed things more or less working better when I didn't poke them. :)
19:36 shortcircuit slavik: If you haven't closed your open TVTropes pages now, do it, NOW. You'll thank me later.
19:36 slavik shortcircuit: yeah, I closed it
19:51 kpreid left #rosettacode
19:59 kpreid joined #rosettacode
20:03 BenBE shortcircuit You mean "Spambot ahead"?
20:03 shortcircuit BenBE: ?
20:04 BenBE [20:36] <shortcircuit> slavik: If you haven't closed your open TVTropes pages now, do it, NOW. You'll thank me later.
20:05 shortcircuit Ah. No, I was thinking more along the lines of TVTropes' rare talent as expressed in http://xkcd.com/609/
20:05 fedaykin "xkcd: Tab Explosion"
20:08 BenBE Ah, k. ;-)
20:52 kpreid left #rosettacode
21:02 duper joined #rosettacode
21:02 duper awesome wiki
21:02 snarkyboojum left #rosettacode
21:03 duper kinda reminds me of langref.org
21:03 fedaykin "langref.org"
21:05 kpreid joined #rosettacode
21:06 duper fedaykin: what about it
21:06 shortcircuit duper: It's a preview bot
21:06 duper oh, that's the title tag?
21:06 shortcircuit Anyway, check out [[Help:Similar_Sites]]
21:06 fedaykin http://rosettacode.org/wiki/Help:Similar_Sites
21:06 duper oic
21:11 * shortcircuit grumbles. He's about had it with Dia.
21:24 snarkyboojum joined #rosettacode
21:31 BenBE2 joined #rosettacode
21:32 BenBE left #rosettacode
21:57 Mathnerd314 joined #rosettacode
21:59 kpreid left #rosettacode
22:01 parsleyfirefly left #rosettacode
22:28 MigoMipo left #rosettacode
22:56 kpreid joined #rosettacode
22:59 kpreid left #rosettacode
23:18 FireFly left #rosettacode

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