# IRC log for #rosettacode, 2011-01-31

All times shown according to UTC.

Time Nick Message
00:04 BenBE joined #rosettacode
00:07 BenBE left #rosettacode
00:16 BenBE joined #rosettacode
00:23 BenBE left #rosettacode
00:30 kpreid joined #rosettacode
00:32 BenBE joined #rosettacode
00:38 BenBE left #rosettacode
00:47 BenBE joined #rosettacode
00:54 BenBE left #rosettacode
00:59 Mathnerd314 idea: have some guidelines on what library functions are allowed (unless the question is about comparing libraries)
01:01 BenBE joined #rosettacode
01:05 BenBE left #rosettacode
01:13 BenBE joined #rosettacode
01:16 BenBE left #rosettacode
01:26 BenBE joined #rosettacode
01:32 BenBE left #rosettacode
01:41 shortcircuit Mathnerd314: Libraries are allowed unless the task in question prohibits them, in which case it's then a gray area.
01:41 BenBE joined #rosettacode
01:46 Mathnerd314 shortcircuit: but then you can't really compare implementations, because you have to look at the library too
01:47 shortcircuit Mathnerd314: Multiple solutions in a given language for a given task are allowed.
01:47 shortcircuit Mathnerd314: Also, the veracity of you statement depends on the context of the comparison; what kind of comparison are you doing?
01:48 Mathnerd314 syntactical; "these do exactly the same thing, but one has 10x less line noise"
01:50 shortcircuit Mathnerd314: In general, on RC, you can't make that comparison anyway; the requirements for an example is that they meet the requirements for a task, not that they do exactly the same thing.
01:51 shortcircuit Furthermore, variations in approach and idiom usage would lead to wildly different results for two languages with nearly identical capabilities.
01:53 shortcircuit Mathnerd314: The reason for the roughly anything-goes policy is simple; if that's how it's done, then that's how it's done.
01:54 Mathnerd314 but you can't compare between languages where the solution is "use this library" and languages where the solution is "run this code"
01:54 BenBE left #rosettacode
01:55 shortcircuit Mathnerd314: At a syntactical level, you are correct. However, that kind of problem should only be true for non-trivial scenarios.
01:56 shortcircuit There are a few tasks which attempt to expose the fundamentals of languages' idioms and syntax. search for 'conditional statements', 'loops', etc.
01:58 shortcircuit We certainly need more of those, I'll admit. But that kind of end is generally best met by limiting the scope of the task such that (for most languages) loading a library is more trouble than solving the problem with it is worth.
02:02 BenBE joined #rosettacode
02:08 Mathnerd314 but you don't exclude libraries, so some of them "cheat"
02:10 shortcircuit Mathnerd314: Generally, if they didn't "cheat" in the sense you're describing, the example never would have been written in the first place.
02:11 shortcircuit I can think of very few contributors to Rosetta Code who willingly write code which is forced by task description to be tortiously unidiomatic.
02:14 shortcircuit Mathnerd314: Generally, the choice is between "cheating" code and no code at all. There are tasks which get written that proscribe language non-builtins/non-bundleds, but they tend to have the fewest solutions.
02:17 Mathnerd314 well, I was looking at this: http://rosettacode.org/wiki/Matrix-exponentiation_operator#Ruby
02:17 fedaykin "Matrix-exponentiation operator - Rosetta Code" http://rldn.net/EU5
02:18 shortcircuit Looks like they inlined the library contents.
02:18 Mathnerd314 the solution for R is completely built-in: http://rosettacode.org/wiki/Category:R
02:18 fedaykin "Category:R - Rosetta Code"
02:19 Mathnerd314 * http://rosettacode.org/wiki/Matrix-exponentiation_operator#R
02:19 shortcircuit I don't know R, but what's the "library(Biodem)" line?
02:19 fedaykin "Matrix-exponentiation operator - Rosetta Code" http://rldn.net/EU5
02:20 Mathnerd314 I'm guessing this library: http://cran.r-project.org/web/packages/Biodem/index.html
02:20 fedaykin "CRAN - Package Biodem" http://rldn.net/12ynF
02:21 shortcircuit Mathnerd314: If you want, you can omit examples which use libraries from your analysis...
02:22 Mathnerd314 (and that library apparently provides matrix exponentiation)
02:23 Mathnerd314 but there's no way to tell from looking at the task; the OCaml implementation is basically from scratch
02:23 shortcircuit Why are you looking at the task in lieu of looking at the example itself?
02:24 Mathnerd314 because the task is "Demonstrate how to implement matrix exponentiation as an operator."
02:24 shortcircuit Then any task which doesn't show how to do that is *incorrect*.
02:24 shortcircuit At which point it needs [[Template:Incorrect]].
02:24 fedaykin http://rosettacode.org/wiki/Template:Incorrect
02:25 shortcircuit (Although the C example delves into that gray area I mentioned earlier.)
02:27 shortcircuit It up-front points out that the task requirements as specified is impossible in the language, and then provides an implementation for the parts that still are possible.
02:28 shortcircuit Ideally, the code could handle a separate task which only asked for matrix exponentiation without adding the operator requirement, but that kind of separation doesn't _usually_ work out in practice on the site.
02:29 shortcircuit If we had example transclusion working using Semantic MediaWiki properties for inclusion constraints, I'd jump on that approach in a second.
02:31 shortcircuit It's doable and desirable, but fragile; we're still trying to figure out how to get the templates to work right.
02:35 Mathnerd314 you ought to have a separate "operator overloading" page, since it tends to be an all-or-nothing proposition
02:38 shortcircuit Mathnerd314: Create the task. I don't generally create tasks. There are too many that I sincerely believe ought to be made, and I don't have the time.
02:39 Mathnerd314 yeah, was just signing up to do that
02:39 BenBE left #rosettacode
02:39 shortcircuit For example, I don't think we have most of the fundamental and common OO idioms for tasks yet, yet it seems obvious that that kind of task would be useful.
02:47 BenBE joined #rosettacode
04:59 parsleyfirefly left #rosettacode
05:53 Mathnerd314 left #rosettacode
08:47 BenBE left #rosettacode
08:55 BenBE joined #rosettacode
10:29 BenBE left #rosettacode
10:36 BenBE joined #rosettacode
14:00 BenBE left #rosettacode
14:00 FireFly joined #rosettacode
14:08 BenBE joined #rosettacode
14:24 kpreid shortcircuit: You might find http://rosettacode.org/wiki/User:Kevin_Reid/Task_list_updater.e interesting
14:24 fedaykin "User:Kevin Reid/Task list updater.e - Rosetta Code" http://rldn.net/11rs4
14:25 kpreid Back when the unimp lists were built by a bot, I would read their changes and modify my classified lists like http://rosettacode.org/wiki/User:Kevin_Reid/E_tasks to match
14:25 fedaykin "User:Kevin Reid/E tasks - Rosetta Code" http://rldn.net/3dt
14:25 kpreid That broke when they started being calculated by MultiCategorySearch (since there were no longer diffs); now I have a script that takes care of it.
14:25 opticron :)
14:26 opticron my fault
14:26 kpreid this is a better solution anyway, because it doesn't rely on catching every diff
14:49 parsleyfirefly joined #rosettacode
14:50 shortcircuit kpreid: I'll check it out later. Monday's looking busy, Tuesday will probably be about as busy, but Wednesday isn't spoken for yet. :)
14:55 parsleyfirefly left #rosettacode
15:01 parsleyfirefly joined #rosettacode
15:42 MigoMipo joined #rosettacode
16:07 parsleyfirefly1 joined #rosettacode
16:09 parsleyfirefly left #rosettacode
16:10 parsleyfirefly1 is now known as parsleyfirefly
16:18 BenBE left #rosettacode
16:25 BenBE joined #rosettacode
17:01 BenBE left #rosettacode
17:09 BenBE joined #rosettacode
17:16 kpreid left #rosettacode
17:24 BenBE left #rosettacode
17:31 BenBE joined #rosettacode
17:49 parsleyfirefly left #rosettacode
17:51 kpreid joined #rosettacode
17:51 parsleyfirefly joined #rosettacode
17:59 mwn3d_phone1 joined #rosettacode
18:00 mwn3d_phone left #rosettacode
18:02 BenBE left #rosettacode
18:04 mwn3d_phone1 is now known as mwn3d_phone
18:09 BenBE joined #rosettacode
18:32 BenBE left #rosettacode
18:35 BenBE joined #rosettacode
18:35 BenBE left #rosettacode
18:35 BenBE joined #rosettacode
18:35 BenBE 2. Next GeSHi release tomorrow evening.
18:35 BenBE 20:10 01.02.2010 ist just a too good date to miss ;-)
18:39 BenBE left #rosettacode
18:44 BenBE joined #rosettacode
18:56 kpreid left #rosettacode
19:03 kpreid joined #rosettacode
19:35 shortcircuit BenBE: Just be careful about poking PCRE. ;)
19:36 Coderjoe do not taunt the PCRE
19:41 shortcircuit Don't use it on concrete. When done playing, place PCRE back in its secured container and refridgerate.
19:43 shortcircuit Productive programmers, Great Old Ones and newbie programmers should avoid prolonged exposure to PCRE.
19:46 shortcircuit Discontinue use of PCRE if any of the following occurs: Itching, vertigo, dizziness, tingling in extremities, loss of scheduling or sync objects, slurred typing, temporary files, profuse smoking or TSC drift.
19:47 shortcircuit If PCRE begins to speak, get away immediately. Seek shelter and cover ears.
19:48 mwn3d_phone I'm imagining all of those instructions being read by glados from portal (a newer video game in case you haven't heard of it)
19:48 Coderjoe if you haven't heard of portal, perhaps you have heard the cake is a lie and that you're still alive
19:49 shortcircuit Glados is PCRE with speech...
19:49 mwn3d_phone Does PCRE offer cake?
19:50 BenBE left #rosettacode
19:50 kpreid left #rosettacode
19:54 Coderjoe the cake is a pie
19:55 BenBE joined #rosettacode
19:58 BenBE shortcircuit Clinical studies have shown PCRE to be effective - just not for what it's used.
20:02 kpreid joined #rosettacode
20:02 BenBE left #rosettacode
20:02 BenBE joined #rosettacode
20:49 kpreid left #rosettacode
21:12 kpreid joined #rosettacode
21:27 * shortcircuit thinks he's not going to run 6to4 for the prgmr node.
21:28 shortcircuit Aggregate over a couple days, pings to the 192.88.99.1 anycast address have reached a 35.6% packet loss.
21:28 shortcircuit It's not a gradual loss, by the looks of it, so I'd guess it's much more likely to be stale routing information as an anycast responder gets overloaded and goes down.
21:31 BenBE Yeah. that's still the problem with 6to4
21:33 shortcircuit What's odd is how stable my 6to4 from home appears to be.
21:33 shortcircuit In any case, it's going to be largely dependent on where you are on the Internet.
21:34 shortcircuit From home, I had 0.0% packet loss over the course of 30k packets.
21:35 shortcircuit What I'd love to see, personally, is for hosting providers like prgmr, Linode, LiquidWeb, Slicehost, etc, to have a network-local 6to4 relay router.
21:37 shortcircuit For hosting providers which dish out public IPs, that's a pretty simple deployment mechanism.
21:41 BenBE Well, I go tproviders which offer native IPv6.
21:41 BenBE That way I can complain if the quality of the v6 connection is below useable ;-)
21:41 BenBE And it's not that expensive either: 5� per month for a /64 ...
21:50 Coderjoe shortcircuit: why not just have the hosting provider provide native ipv6?
21:51 shortcircuit Coderjoe: Primarily because I'm less confident in planning and configuring for IPv6 than for IPv4, and would prefer a solution that builds more off of areas I'm familiar with.
21:53 BenBE shortcircuitFor IPv6 there's not much configuration to do. 3 lines in /etc/network/interfaces, one additional line on /etc/hosts for rDNS, and copying all vHosts from IPv4 to IPv6 (for NVH, no change for IPVH).
21:53 BenBE Changes done in <30 minute+tests.
21:54 BenBE If you need help, feel free to ask.
21:56 shortcircuit prgmr offers experimental, opt-in IPv6.
21:56 shortcircuit I'm probably just about ready to roll out 6to4 at home, though.
21:57 BenBE shortcircuitWhat do they charge for the opt-in?
21:57 shortcircuit Nothing, IIRC.
21:57 shortcircuit BenBE: Did you see this, per chance? http://mmol-6453.livejournal.com/260143.html
21:57 fedaykin "Michael Mol - Got Linux? Curious about IPv6?"
21:57 BenBE Then: Opt-In, configure Service by serve and announce v6 on separate domains for now, until things work properly.
21:58 BenBE Not yet, just gonna read it ...
21:58 BenBE sec
21:58 parsleyfirefly left #rosettacode
21:59 shortcircuit Related and expanded: http://shinobu.grlug.org/pipermail/grlug/2011-January/008672.html
21:59 fedaykin "[GRLUG] Testing a host's IPv6" http://rldn.net/6ka
21:59 BenBE Yeah, someone mentioned it ...
22:00 shortcircuit That ping6 to ff02::1 was from the prgmr node; there's at least 86 machines on the local layer 2 responding to ICMPv6 echo.
22:00 shortcircuit Anyway, I'm off for most of the evening.
22:12 Mathnerd314 joined #rosettacode
22:32 waldwolf joined #rosettacode
22:32 waldwolf left #rosettacode
22:32 waldwolf joined #rosettacode
22:37 kpreid left #rosettacode
22:38 MigoMipo left #rosettacode
22:54 BenBE left #rosettacode
23:03 BenBE joined #rosettacode
23:07 BenBE left #rosettacode
23:17 BenBE joined #rosettacode
23:17 CodeBlock left #rosettacode
23:21 CodeBlock joined #rosettacode
23:27 BenBE left #rosettacode
23:27 BenBE joined #rosettacode
23:27 asdtoob joined #rosettacode
23:29 kpreid joined #rosettacode
23:29 asdtoob_ joined #rosettacode
23:32 CodeBlock left #rosettacode
23:36 asdtoob left #rosettacode
23:36 CodeBlock joined #rosettacode
23:40 BenBE left #rosettacode
23:46 lambdabot left #rosettacode
23:48 asdtoob_ left #rosettacode
23:49 BenBE joined #rosettacode
23:53 BenBE left #rosettacode