Perl 6 - the future is here, just unevenly distributed

IRC log for #rosettacode, 2011-08-02

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

All times shown according to UTC.

Time Nick Message
00:01 BenBE My BF Interpreter does the same. With Single Stepping and Live Edit while Execution ;-)
00:03 BenBE k, found ONE bug in my program; fixing ;-)
00:03 BenBE Lines 54 and 55 are doing something wrong ...
00:06 rodt should have a self interpreter task :P
00:08 BenBE ?
00:09 BenBE IIRC there was a BF interpreter implementation in BF ...
00:09 rodt http://en.wikipedia.org/wiki/Self-interpreter
00:09 rodt yeah it apparently tells a bit about the lang, as to have trivial a self interpreter is in it ?
00:14 mwn3d_phone Self-interpreter/compiler?
00:15 rodt yeah lol
00:17 mwn3d_phone I never heard a reason that it was good to do that with a language
00:17 mwn3d_phone It never blew my skirt up
00:18 rodt you need it for ai for instance
00:20 BenBE You can do a self-interpreter in PHP quite easily ... although this still is roughly 2kLOC of code plus a lot of Logic for the operators ...
00:20 rodt most langs its would be fairly easy
00:20 rodt well it depends on the lang really
00:20 BenBE BF is a REAL task :P
00:21 rodt the solutios would be varied ;)
00:21 rodt i didnt say it wasnt :P
00:23 rodt anyway im offski, take care peeps :)
00:23 rodt left #rosettacode
00:24 parsleyfirefly left #rosettacode
00:24 sorear self-interpreter says a lot about many things and so "short" or "long" doesn't tell you much
00:24 sorear BF is very simple but its low conciseness makes the self-interpreter long
00:25 sorear strictly interpreted languages with an eval make it extremely short
00:25 sorear for Perl or Javascript or SNOBOL, you can imagine banning eval, but it's not always that easy
00:25 sorear in Lisp, Forth, and J the parser is a pretty integral part of the system library
00:28 BenBE Okay, the overflow part now should do properly ...
00:44 parsleyfirefly joined #rosettacode
00:48 BenBE Got it from 1 to 9; fixing the overflow handling right now ^^
00:59 BenBE Okay, seems I got it work now ...
01:02 BenBE Okay Coderjoe, could you recheck the BF code? Updated the Code sample on RC.
01:14 mwn3d_phone BenBE: looks like it works
01:14 Coderjoe appears to work
01:27 parsleyfirefly left #rosettacode
01:28 kpreid left #rosettacode
01:37 mwn3d_phone left #rosettacode
01:41 BenBE Okay. Had 3 distinguish bugs in the original version. The first one regarding the mod3/5 counters, the second with the overflow detection for the digits and the third was some other stupid bug. The first two were both bugs that required porting the fix two two locations.
01:45 mwn3d_phone joined #rosettacode
02:15 parsleyfirefly joined #rosettacode
02:19 kpreid joined #rosettacode
02:20 kpreid_ joined #rosettacode
02:20 kpreid left #rosettacode
02:20 kpreid_ is now known as kpreid
02:58 parsleyfirefly left #rosettacode
03:12 mikemol I'm not sure we'll be seeing markhobley back. He hasn't been active on RC in about a week, he's detached from following RC's twitter account (where he was a bit of a champion of RC for a while), and I haven't received answers to an emailed query.
03:13 mwn3d_phone He was back today
03:14 mikemol Oh? I missed it.
03:14 mwn3d_phone They were minor edits
03:14 mikemol Ah, I see them, now.
03:15 mwn3d_phone I think he mentioned that he works at a university. Could have been finals week for the accelerated summer classes?
04:00 BenBE left #rosettacode
04:05 BenBE joined #rosettacode
05:19 mwn3d_phone left #rosettacode
09:25 FireFly joined #rosettacode
09:35 rodt joined #rosettacode
13:06 mwn3d_phone joined #rosettacode
13:14 dagnyscott1 joined #rosettacode
15:03 mikemol Need moar algorithms: http://www.reddit.com/r/learnprogramming/comments/​j3h21/is_there_a_good_resource_with_actual/c28tm9x
15:04 rodt ironic ;)
15:05 rodt i spose theres a ton of tasks for common algs you could/prolly should add
15:06 rodt you need about 4times the current number of tasks prolly ideally...
15:07 rodt then youd have a wiki of algs in most langs... but depends how you go with the site... i mean as it its great for cheecking out langs etc...
15:08 rodt i spose something like that would naturally grow into such a wiki tho
15:08 rodt by wiki i just mean more complete...
15:09 Hypftier rodt: http://en.wikibooks.org/wik​i/Algorithm_implementation
15:12 rodt er thats teh same as rossetacode ?
15:13 rodt prolly wanna merge or something no ?
15:19 Hypftier It's much smaller than RC, from what I can see. A bit more organized maybe, too. Still, even if you'd created 200 more tasks on RC right now for every conceivable algorithm ... have fun writing the implementations, too. Tcl might be complete, but few other languages are.
15:19 rodt lol psuedo code and applescript both only have 1 alg... quicksort...
15:20 kpreid left #rosettacode
15:20 lambdabot left #rosettacode
15:20 rodt ?
15:21 rodt i dont understand you lol
15:21 Hypftier Besides, RC has also a broader scope, covering not only algorithms, but also cookbook-y tasks (i.e. everyday problems that you might want to solve or adapt a solution from there), introductory topics (what kind of literals are there, control structures, loop constructs, etc.) and general concepts (how to "throw" an Exception, etc.)
15:22 rodt so add 200 tasks ?
15:23 mikemol rodt: Simply put, add tasks, as needed.
15:23 Hypftier Feel free to propose them. But don't expect them to magically fill with 150 implementations each in a day. There are not that many volunteers participating regularly on RC; I also didn't do much for quite some time now.
15:23 mikemol RC will never be complete. It's just that kind of a thing.
15:23 rodt prolly need task cats
15:23 rodt oh ;)
15:24 Hypftier mikemol: I like the direction of RC in that it doesn't only encompass algorithms but also more topics :)
15:24 rodt so it would be good to copy the wiki alg book to rc ?
15:25 mikemol Hypftier: I'd love to have it cover the 'libraries' dimension, too, but there's not much interest in that.
15:25 Hypftier mikemol: probably boils down to very limited contributor time, I guess.
15:26 mikemol So we tend to only see tasks which have strong built-in support, or for 'solved' problems.
15:26 Hypftier tasks × languages is quite a lot already, if you add × libraries ... yikes.
15:26 mikemol Hypftier: I never figured RC would ever be 'complete', nor does it need to be.
15:27 Hypftier rodt: Try to find the set of implementations and/or tasks that are available there but not on RC. That Wikibook is very small, actually.
15:27 rodt oh
15:27 mikemol rodt: Mind you we can't just do wholesale copying.
15:27 Hypftier mikemol: *nods*
15:28 rodt isnt the wiki somekind of "opensrclol" doc anyway?
15:28 Hypftier define »opensrclol«
15:28 rodt "you have the right to copy" basically
15:28 mikemol no
15:28 Hypftier it's under CC-By-SA
15:28 rodt creative commons attribustion ?
15:28 mikemol You can copy, but there's still necessity of attribution.
15:28 rodt "copyed from wikibooks..."
15:29 mikemol Also, RC isn't CC-licensed, so CC-By-SA --> Rosetta Code would be a bad transfer.
15:29 Hypftier mikemol: I'd love it to be, though :)
15:29 mikemol It would have to be a small enough piece to fall under fair use.
15:30 rodt all this licensing stuff "is unfair"
15:30 mikemol Then there's questions of what the 'work' is. Is it the example program? Is it the source task page? Is it the entire source wiki?
15:30 Hypftier Although individual authors on RC make their contributions available under additional terms as well, sometimes.
15:30 mikemol For every conveniently-chosen answer to such a question, there's an equally inconveniently-chosen possible counter.
15:31 mikemol rodt: Life is unfair. But that's not a bad thing. :)
15:32 mikemol A great quote from a great TV series: "I used to think it was awful that life was so unfair. Then I thought, 'wouldn't it be much worse if life *were* fair, and all the terrible things that happen to us come because we actually deserve them?' So now I take great comfort in the general hostility and unfairness of the universe."
15:32 rodt lol
15:33 Hypftier Another problem is that RC started out with GFDL and migrating to a new license involves consent from *every* contributor. See the state in Wikipedia, for example, which says "We're CC-By-SA, except when not, like when the page was written under GFDL or incorporates content that was licensed under GFDL" which is the problem you have after such a migration.
15:33 mikemol The only clean solution would be a total re-implementation of the wiki.
15:34 mikemol Which isn't *impossible*, and could have other positive effects, but it's wholly impractical right now.
15:34 Hypftier mikemol: couldn't you apply for GSoC for such a thing? :)
15:34 Hypftier or is RC not really an Open Source project
15:34 rodt lol rc != OS :D
15:35 Hypftier Hm, it would require you to mentor a student for that, though, which takes time as well
15:35 mikemol RC would probably qualify as open-source.
15:35 Hypftier mikemol: well, I meant it's not actually *software*, more a project collecting code :)
15:35 mikemol Hey, I'd be happy to mentor students. I was a student tutor for five years, started RC while a student tutor, and never really lost the educator mindset. :)
15:35 Hypftier Dunno how Google's terms are in that regard
15:37 rodt hmm confusing lol
15:37 mikemol The big reason not to do it right now is that we are (and have been) (and may forever be) on the verge of being able to shoehorn MediaWiki into behaving much more database-like.
15:38 mikemol The slightly less big reason is getting people to come to a relative agreement on how to move things around and rearrange things.
15:39 mikemol Everyone who's been part of RC for more than a few weeks will have their own ideas for how the site could run better, and it'll be pretty messy moderating and managing all the conflicting opinions as they land on a fresh canvas.
15:39 Hypftier (Why does that remind me of today's Dbildert cartoon? :))
15:39 Hypftier *Dilbert ... argh
15:40 mikemol Migration of content isn't a big deal; it just doesn't happen. The old wiki would get locked down to just a few editors. Edits would be restricted to "this content is archived. See {here} on the new wiki for related content"
15:41 mikemol I'd leave the existing wiki up exactly where it is. The new wiki would end up with a URI scheme like rosettacode.org/wiki2/*
15:59 FireFly Interesting.. browsing to http://rosettacode.org/wikiMain_Page actually takes you to Main Page.
16:00 mikemol Hm. That tells me I'll need to fix some stuff in the apache configuration.
16:00 mikemol "wiki2" would get redirected to "/wiki/2", otherwise.
16:01 kpreid joined #rosettacode
16:37 MigoMipo joined #rosettacode
16:39 mikemol I need a program I don't have time to write. It needs to grab a list of all category members and return a random entry in that list.
16:39 mikemol It needs to record the entry it returned, and use the set of "previously returned entries" to cull from the set of choices in subsequent runs.
16:41 mikemol Right now, I've got 30 scheduled posts through August 31, each being a random task. I'd like to automate that. I'd also like to set up automated scheduled postings for random draft tasks, and for random languages.
16:42 mikemol So, there'd be three posts a day, each pulling a random entry out of a different category--except I don't want repeats until the sets are exhausted.
17:34 Hypftier No shuffling? :)
18:01 mikemol Would shuffling be necessary?
18:06 Hypftier nah, not for the numbers of items we're talking about here
18:06 mikemol Thing is, you're already picking randomly from a culled set. (Or didn't I include the random pick in the described algorithm?)
18:08 Hypftier somehow thought about picking random and comparing against the set of already chosen ones. My brain doesn't really work anymore today :)
18:08 sorear step 1: scrape the DB to get a list of all candidates
18:09 sorear step 2: shuf -n 90 /tmp/candidates
18:09 mikemol head -1 /tmp/candidates
18:10 mikemol This would be the preferred way to get category members, though: http://www.mediawiki.org/wiki/API:Categorymembers
18:59 mwn3d_phone left #rosettacode
19:12 Coderjoe preferably with a bot account so it can get 5k items per pull
19:30 p6eval left #rosettacode
19:30 p6eval joined #rosettacode
19:43 * BenBE just updated http://rosettacode.org/wiki​/User:BenBE/Schulze_method to include a (yet incomplete) implementation of the proposed task.
19:44 fedaykin "User:BenBE/Schulze method - Rosetta Code" http://rldn.net/UUb
20:01 mwn3d_phone joined #rosettacode
20:55 MigoMipo_ joined #rosettacode
20:58 MigoMipo left #rosettacode
21:05 dagnyscott1 left #rosettacode
21:22 lambdabot joined #rosettacode
21:52 MigoMipo_ left #rosettacode
22:17 rodt left #rosettacode

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