Perl 6 - the future is here, just unevenly distributed

IRC log for #rosettacode, 2012-03-16

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

All times shown according to UTC.

Time Nick Message
00:44 ttmrichter joined #rosettacode
01:14 Bursihido joined #rosettacode
01:43 mwn3d_phone1 joined #rosettacode
01:50 ttmrichter joined #rosettacode
02:09 mwn3d_phone joined #rosettacode
02:10 ttmrichter joined #rosettacode
02:43 ttmrichter joined #rosettacode
04:00 ttmrichter joined #rosettacode
04:50 mwn3d_phone1 joined #rosettacode
04:51 mwn3dsphonesucks joined #rosettacode
06:18 ttmrichter joined #rosettacode
08:19 ttmrichter joined #rosettacode
08:30 kpreid joined #rosettacode
09:34 mischi joined #rosettacode
12:11 FireFly joined #rosettacode
12:33 bearophile joined #rosettacode
12:33 bearophile Hello
12:33 bearophile I have a small problem, maybe you are able to help
12:34 bearophile The D language is young and it changes quickly still, almost every day there are some improvements, even for the language syntax
12:34 bearophile And almost every week some of the D programs in Rosettacode break
12:35 bearophile So I'd like to keep a list of most of the D programs and run them every few days to see what one breaks
12:36 bearophile Do you have a simple way to download all D entries?
12:36 bearophile If you don't have suggestions, then I will probably write a screen scraping script.
12:37 bearophile Assuming you allow me to run such script (but it has to run only once, so it's not a big problem for the server).
13:10 mikemol bearophile: Someone's written a screen-scraping script, but I don't remember who.
13:10 * bearophile nods
13:11 bearophile No direct download then
13:11 mikemol Also, should be no problem running such a script on the site. If the server can't handle it, I haven't configured it properly.
13:11 bearophile OK
13:11 bearophile Thank you
13:11 mikemol np
13:11 mikemol Thank _you_ for taking ownership of a language on RC. :)
13:12 mikemol We don't have very many cases of that. Off the top of my head, I know that Python, Perl 6 and Tcl examples are actively maintained. Not sure which others, or by who.
13:13 bearophile D doesn't have many people that write examples, so there's low risk of stepping on other people toes
13:13 bearophile And most of the D entries are written by me anhy ay now
13:13 bearophile any way
13:13 mikemol I generally leave it to each language's community to figure out how they want to do things.
13:14 * bearophile nods
13:15 opticron I've written a few D entries, but I trust bearophile to take care of it :)
13:15 bearophile I have not written all D entries
13:15 bearophile I have improved/updated some of them
13:16 bearophile But in the end it's also a matter of personal tastes in coding
13:16 opticron and whether you're going to write examples in D1 or D2
13:16 opticron since when examples were first written, D1 was the norm
13:16 mikemol Is there a significant difference?
13:16 opticron now that has likely changed to D2
13:17 opticron the base language is very similar, but D2 adds quite a bit of syntax and features
13:17 bearophile There are now many differences
13:17 mikemol I mean, is this the kind of distinction that might warrant separation as with Perl vs Perl 6, or annotation as with Python2 vs Python3?
13:17 bearophile D1 programs usually don't work in D2
13:17 opticron mikemol, probably
13:17 bearophile Even D2 programs one month old usually don't work now
13:17 opticron D2 is still a moving target
13:18 bearophile That's why I have asked for all the D entries, to run them often and look for breaking changes.
13:18 mikemol bearophile: I recommend you become familiar with [[Template:workswith]]
13:18 fedaykin http://rosettacode.org/wiki/Template:workswith (Doesn't exist.)
13:18 mikemol er
13:18 mikemol [[Template:works with]]
13:18 fedaykin http://rosettacode.org/wiki/Template:works_with
13:19 mikemol I'd probably also recommend keeping D1 examples present (and even adding more), and adding D2 examples for contrast.
13:19 opticron D2 also has a lot more in its std lib like json and xml parsing which was noticably missing from the D1 std lib
13:19 mikemol The D1 examples sound like they'll be more stable, and D2 more illustrative.
13:19 bearophile That means duplicating all D entries :-)
13:19 bearophile It means a LOT of work, and I don't program in D1 any more
13:20 bearophile And D designers (like Andrei) want to essentially kill D1
13:20 mikemol Well, as I said, I like to leave things up to the communities directly involved. That's just how I'd recommend it, if you want both stable code and "here's the difference between the two versions" code.
13:20 bearophile D1 will be fully discontinued in some months
13:20 opticron the D community gets pretty silly at times
13:20 mikemol I won't be playing by fiat here. :)
13:20 bearophile So I don't think it's a good idea
13:21 bearophile But if someone wants to add D1 entries, tagged with works_with I will not remove them :-)
13:21 bearophile But I don't think I will add hundreds of D1 versions
13:22 bearophile Thank you for your answers
13:22 bearophile See you later.
13:23 * mikemol shrugs
13:23 * mikemol thought he was recommending keeping working D1 examples in place, not adding more.
13:24 opticron you may have been
13:25 opticron if they're really going to kill D1 like it seems, it won't be worth it to maintain the examples, but leaving them there can't hurt
13:26 mikemol I'm trying really hard not to second-guess their decisions, but it seems really weird to kill off the stable version while the replacement is still a fast-moving target.
13:27 opticron you're not the first one to suggest this
13:27 opticron the biggest issue is that it's splitting the community
13:28 opticron they seek to fix that and the only way that'll happen is by dropping D1 entirely and forcing the move
13:28 opticron and there are other splits as well
13:29 mikemol I've been watching the Perl/Perl6 evolution, and that seems to have been handled very sanely.
13:29 opticron but like I said, the D community gets pretty silly at times
13:29 mikemol Perl 6 has been developing at its own pace, and Perl 5 has been backporting the useful things it can.
13:29 opticron there's also a stdlib war that is pretty close to death (but still might come back with D2)
13:29 opticron yeah
13:30 opticron that never happened with D1/D2
13:30 mikemol Similarly, Python 2 vs Python 3; Python 2.7 contains a lot of Python 3 backports.
13:30 opticron they locked down D1 and said ALL NEW DEVELOPMENT GOES TO D2
13:30 opticron no new feature backports
13:30 mikemol Sounds like enough of a mess that I really need to be keeping my opinions to myself.
13:31 opticron heh
13:32 mikemol Hm. You see the backporting behavior in C++, too. A bunch of the draft features of C++11 appeared in boost long before C++0x finally settled.
13:32 opticron the biggest thing that gets me is the lack of real closures in D1 (they exist in D2) and it's a simple #ifdef'd block to enable that feature
13:33 mikemol Why didn't someone take the D1 codebase and fork it?
13:33 opticron and of course both compilers still use exactly the same codebase, just with different defines
13:33 opticron it's not possible
13:34 opticron the backend is still proprietary even though the code is available
13:34 mikemol Huh. I didn't know that.
13:34 opticron the frontend was just recently (in the last year) made FOSS
13:34 opticron maybe 2 years
13:34 opticron but GCC 4.7 will be getting D2 support
13:34 opticron maybe 4.8
13:34 opticron but it's close
13:37 mikemol Hm. I need an example of an academic programming language.
13:38 mikemol Pascal, I suppose.
13:38 mikemol Hm, perhaps not.
14:14 ttmrichter mikemol: It doesn't get much more academic than Haskell or Mercury or Ur or Oz or Curry or …
14:20 mikemol I thought Oz was supposed to be a commercial package?
14:21 mikemol I'd have a hard time arguing Haskell as academic, though. I know too many people who use it for production purposes. And there's stuff like bzr.
14:22 mikemol Hm. [[Category:Ur]] doesn't exist.
14:22 fedaykin http://rosettacode.org/wiki/Category:Ur (Doesn't exist.)
14:58 mischi pascal? what? where? how come?
15:05 mikemol Yeah, I forgot about Borland's compiler, and that people used Pascal outside of educational purposes. (Ignoring extended versions like Delphi)
15:36 mwn3d_phone1 joined #rosettacode
15:37 TimToady heh, I rewrote the compiler for a discrete event simulator in the early 80s; it was definitely a commercial thing, and ran on the Big Iron of the day
15:37 TimToady (in Pascal)
15:54 kpreid joined #rosettacode
16:07 mwn3d_phone joined #rosettacode
16:24 mikemol I'm very much liking that we haven't had any spam in the last couple days.
16:57 mikemol Edit counts don't appear to have dropped, but I'll be able to measure that better after a week or two.
20:04 mischi joined #rosettacode
21:01 mwn3d_phone1 joined #rosettacode
22:16 mischi joined #rosettacode
22:37 mwn3d_phone1 joined #rosettacode
23:04 mischi joined #rosettacode
23:07 mischi joined #rosettacode
23:22 mischi joined #rosettacode

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