Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2007-07-30

Perl 6 | Reference Documentation | Rakudo

| Channels | #perl6 index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:55 nekokak_ joined #perl6
01:06 sunnavy joined #perl6
01:13 nipotaway changed the nick to nipotan
01:22 jisom joined #perl6
01:27 nipotan changed the nick to nipotaway
02:00 Tene joined #perl6
02:15 dduncan joined #perl6
02:18 drbean_ joined #perl6
02:25 justatheory joined #perl6
02:27 nipotaway changed the nick to nipotan
02:28 buetow joined #perl6
02:34 Aankhen`` joined #perl6
02:38 * diakopter calls evalbot forth from the tomb...
02:38 diakopter it replies "hey, I'm not forth"
02:44 * Aankhen`` stabs diakopter in a friendly manner.
02:45 Schwern joined #perl6
02:46 diakopter ouch.
02:47 diakopter well, apparently I'm not Jesus.
02:48 spinclad nor evalbot a zombie
02:48 diakopter this Lazarus appears to have stopped to build pugs from scratch before it tries to find the tomb's egress.
02:48 diakopter (as well it should)
02:49 spinclad evalbot: this way to the egress
02:50 spinclad evalbot: 'i'm not over egre to go'
02:51 * diakopter snorts
02:55 bcmb_ joined #perl6
02:57 pugs_svnbot r17176 | diakopter++ | 'cuz I like it that way.
02:57 pugs_svnbot diff: http://dev.pugscode.org/changeset/17176
02:57 lambdabot Title: Changeset 17176 - Pugs - Trac
02:59 bcmb_ I've got a question about one of the test scripts...
03:14 diakopter bcmb_: yes?
03:14 diakopter (sorry for the delay)
03:15 bcmb_ no problem. i'm looking at t/syntax/interpolation/strings.t
03:15 bcmb_ it was missing a semi-colon, so i fixed that, but the tests on line 26-27 don't run and was wondering how best to handle that?
03:17 bcmb_ sorry, i mean lines 36-37
03:28 diakopter bcmb_: taking a look...
03:28 bcmb_ diakopter: ok
03:30 perlDreamer joined #perl6
03:31 perlDreamer So I was saying, I've been trying to run the tests from the pugs repo on parrot.
03:31 diakopter perlDreamer: indeed.
03:31 perlDreamer and finding that parrot can easily run fractions of tests, but very few whole tests due to unsupported syntax
03:32 perlDreamer It doesn't sound good for the idea of using them wholesale
03:33 diakopter by "test" you mean ".t file containing one or more tests"?
03:33 perlDreamer yes, test == test file
03:33 perlDreamer some tests don't compile
03:33 diakopter what would you call each is(), ok()?
03:33 perlDreamer a test
03:34 perlDreamer It's context sensitive
03:34 diakopter ok
03:35 * diakopter takes off the RFC-reading mode hat.
03:36 diakopter perlDreamer: what options do you see for helping parrot/languages/perl6 by making the test suite more incrementally runnable?
03:36 perlDreamer hm
03:36 perlDreamer Option 0: Don't try.  Write parrot tests from scratch.
03:37 perlDreamer Option 1: Slice and dice the current tests.  This would make for messy repositories.
03:37 diakopter msw on Option 0: I kinda like this option at first glance.
03:38 perlDreamer Option 2: Hack parrot's fetchspec script to modify the tests to suit parrot on checkout
03:38 perlDreamer Option 3: Import the current suite wholesale, hack out uncompilable code.  Periodically, manuall resync the tests.
03:38 diakopter O0: (except, while still storing the refactored test suite in pugs' svn or whichever spot becomes agreed-upon next)
03:39 diakopter O2: blecchhh
03:39 diakopter :)
03:39 diakopter O3: blecchhh
03:39 perlDreamer 2 and 3 give you some bootstrapping
03:40 perlDreamer for existing tests
03:40 perlDreamer re your comment on O0, why would parrot specific tests be put into the pugs svn repo?
03:41 diakopter O3: manually resyncing would be a huge pain.  Unless it's mostly one-way.  That's quite hard to predict, contrary to (perhaps) popular belief.
03:41 diakopter why would there be parrot-specific tests?
03:41 perlDreamer well, pugs doesn't implement the e constant yet
03:42 diakopter disregarding the (difficult to trace) historical literature on the origin of the current todo marking system... what do you suggest for that?
03:42 perlDreamer I don't think parrot can do named params yet, so the current marking system is no good.
03:43 diakopter p/l/nqp can, so I assume p/l/perl6 could trivially....?
03:43 perlDreamer what is p/l/nqp?
03:44 diakopter parrot/languages/nqp
03:44 weinig joined #perl6
03:44 perlDreamer perl6 still uses tge?
03:45 diakopter no, I mean it's probably a copy-and-paste possibility.
03:45 perlDreamer it's hard to write tests for possibilities, unless you're talking about junctions :)
03:46 diakopter I wasn't suggesting writing tests for the possibility I conjectured.
03:47 diakopter I was suggesting that you or pmichaud could merge down more of nqp's functionality to perl6
03:47 perlDreamer When perl6 uses nqp, that might work.
03:47 diakopter why wouldn't it work before then?
03:48 diakopter (sorry if that's a dumb question)
03:50 perlDreamer You asked about the todo marking system.  It uses named params,   perl6 (parrot) doesn't do named params.  You mentioned nqp (which perl6 (parrot)) doesn't use.
03:50 perlDreamer The questions isn't implementation, it's function.
03:52 amnesiac joined #perl6
03:59 hakobe joined #perl6
04:04 drbean joined #perl6
04:12 thoughtpolice joined #perl6
04:12 diakopter so, back to O0..O3.  Let me be a lot more explicit about the design parameters, here.
04:15 diakopter allison declared pugs' svn the location of the official 'Perl 6' test suite, here: http://use.perl.org/articles/06/12/01/2234220.shtml   ...
04:15 lambdabot Title: use Perl | Perl 6 Design Meeting Notes for 29 November 2006
04:17 diakopter perlDreamer: your comment on my comment on O0 implies that you think tests that are pugs:TODO but are implemented in parrot shouldn't be in the official test suite
04:17 diakopter er, shouldn't be in other implementations' test suite
04:18 * obra begs the community to try to keep a single test suite together across all implementations.
04:18 diakopter TPF's goal is a test suite that is releaseable under AL2
04:18 diakopter (which requires the contributor agreements, etc)
04:18 diakopter (another reason why I still prefer O0)
04:19 perlDreamer diakopter: only because I don't think (but haven't tried) the :todo's will compile under parrot.
04:19 perlDreamer aside from that, it should be no sweat
04:19 diakopter (because whoever does the refactoring can release/commit them under AL2.
04:19 diakopter )
04:19 perlDreamer Can't take someone else's code and re-release under a different license.  That's bad.
04:19 perlDreamer even after refactoring
04:20 diakopter I guess I'm imagining a somewhat more severe refactoring than you are, then.
04:20 diakopter I think what you're imagining is closer to O1
04:21 perlDreamer there are 18000 tests in the pugs repo
04:21 diakopter sort of.
04:21 perlDreamer rewriting that from scratch, by hand, would take large amounts of time
04:21 obra I believe  patrick has indicated a preference to improve the pugs versions of the tests to be useful to parrot.
04:21 diakopter many of them are redundant or generated
04:22 obra rather than to fork the test suite
04:23 diakopter obra: right; I guess I'm trying to take into account TPF's perspective.
04:23 diakopter what with the claim of copyrighting "Perl 6" on the home page.
04:24 obra I think that for now, I'd not worry about that.
04:24 obra we have full version history.
04:24 obra if there is ever a license issue and a contributor isn't comfortable making a grant of rights AND that grant of rights is necessary, the en corrective action can be taken
04:27 obra TPF has been working on copyrights with larry's blessing and involvement. (But TimToady can better speak to that ;)
04:29 diakopter obra: my other (related) point is that the test suite needs not only a (1) drastic reorganization [which tests are in which file, which test files are in which dirs, what the files are named, what the dirs are named] but also a (2) quite severe functionality-rejiggering [there are only two closed dependency strata, currently: pre-sanity and post-sanity], not to mention
04:30 obra ok. that sounds like a good thing
04:31 diakopter (3) integration of the "code" sections of the Synopses as tests that can be back-smart-linked to the Synopses.
04:34 perlDreamer what if we had a branch for parrot tests in the pugs repo
04:34 diakopter so I guess my hands-thrown-up-in-surrender mindset/conclusion is to say that the organically grown test suite needs formalized, and in the process, refactored, reorganized, cross-dependency-reduced..... in order to meet the goal I stated way above (helping parrot/languages/perl6 by making the test suite more incrementally runnable)
04:35 obra perlDreamer: why a parrot branch specifically?
04:35 diakopter perlDreamer: other than not having the named parameter markers, why do they need a separate branch?
04:35 obra I suspect kp6 is going to need something similar in the near future too.
04:35 perlDreamer what parrot needs today, is pugs' test suite from a long time ago
04:35 obra (this discussion might be worth having on p6l)
04:35 perlDreamer that sounds like version control
04:35 perlDreamer which makes me think branch
04:36 obra perlDreamer: well, the language has evolved considerbly
04:36 diakopter what was more usable/runnable about the test suite long ago, and how long ago?
04:36 perlDreamer yes, but currently parrot can't do named params, object methods on variables, and many other things that pugs does today.
04:37 perlDreamer and are in the test suite
04:37 obra is it the test harness that needs these? or is it more advanced test files?
04:37 perlDreamer it's across the board
04:37 perlDreamer have a peek at t/builting/math/trig.t
04:37 perlDreamer it fails on test 2 because parrot doesn't handle unary builtins without parens yet.
04:38 perlDreamer the substr tests won't run because parrot only supports substr with no parens
04:38 perlDreamer the join tests will run with and w/o parens, but not as string methods
04:38 perlDreamer and all those things are interleaved through the tests
04:39 diakopter perlDreamer: for those several things you just mentioned, are there tests that test *just* those things?
04:39 perlDreamer no
04:39 perlDreamer that's why I suggested a branch
04:39 diakopter perlDreamer: then that's what you add to pugs' test suite: in new files.
04:40 perlDreamer or as a branch.
04:40 perlDreamer either would work
04:40 obra I think that new test files would have the greatist utility. And lead to the most people actually looking at, working on and running them
04:40 perlDreamer then the parrot people would have to keep a list of which test files they can run, and which they can't
04:41 obra perlDreamer: having a script that "runs all useful tests for parrot" in the pugs repo would certianly be welcoem
04:41 obra welcome, even
04:41 perlDreamer pmichaud has such a script, but thinking about scaling to hundreds of names isn't pleasant
04:41 obra and it might even be a good thing to start turning something like that into "test this tier of functionality"
04:42 diakopter perlDreamer: what about putting all these new files into new dirs?
04:42 obra it sounds less bad then having parrot tests out on a branch and starting to diverge. that feels like it could lead to a _lot_ of duplicated effort and diverging implementations
04:42 diakopter new dirs that start with digit-pairs, perhaps.
04:43 perlDreamer obra, how is this different from branching and merging source code?
04:43 obra also, if the tests are organized by what functionality they need in the interpreter. (or even annotated with those features), you wouldn't need to maintain a list.
04:43 diakopter perlDreamer: then the import script could import all the files in just particular dirs
04:43 perlDreamer divergence is controlled by dev's
04:43 diakopter perlDreamer: b/c pugs' svn doesn't have branches.
04:43 perlDreamer is that by policy?
04:43 diakopter and why would there need to be merging?
04:43 obra perlDreamer: the zen of pugs is not big on long-lived branches.
04:44 perlDreamer ah
04:44 perlDreamer well, then dirs sounds better then
04:44 obra branches are mostly used for short feature work.
04:44 nekokak joined #perl6
04:44 diakopter perlDreamer: why would they ever need to be merged anywhere?
04:44 obra *nod*
04:44 perlDreamer well, if someone added string methods to parrot, you could merge in those tests from the respective test files
04:45 perlDreamer but it sounds like a poor idea, compared to dirs
04:45 diakopter why merge?  why not just move?
04:45 diakopter instead of copying
04:45 perlDreamer I'm assuming that wholesale moving wouldn't work, due to other features that may not work yet
04:46 perlDreamer maybe it's just nomenclature
04:46 diakopter I think I'm following
04:46 diakopter definitely don't ever duplicate individual test lines, even partially.
04:47 [particle] please take this discussion to a list.
04:47 [particle] btw there is a syntax for marking tests as todo with a comment
04:47 diakopter [particle]: backlog; discussed :{
04:47 [particle] #todo :parrot<0.4.1> (or similar)
04:47 diakopter :P
04:47 [particle] yes, i see that.
04:48 diakopter the problem is that parrot's perl6 can't parse that, according to perlDreamer
04:48 [particle] yet.
04:48 perlDreamer agreed, yet
04:48 diakopter right, obvious.
04:48 perlDreamer which list would be appropriate to move this discussion to?
04:48 [particle] once the tests are in 01-sanity, it can be a requirement for Test.pm
04:48 [particle] p6c, i think
04:49 diakopter [particle]: once which tests are in 01-sanity?
04:49 [particle] tests for anything deemed a requirement for running Test.pm
04:50 perlDreamer I need to go stand on a board, I'll watch p6c for more discussion
04:50 diakopter ah, named parameters.
04:50 [particle] either they go in t/01-sanity or t/02-test
04:50 [particle] also, right now 01-sanity tests are not sufficient to run Test.pm
04:51 [particle] don't have 100% coverage of Test.pm features, that is.
04:51 * [particle] has to run off
04:52 diakopter right, that's why there're only 2 strata, and not 3 (including the Test.pm level).
05:01 Schwern joined #perl6
05:06 xinming_ joined #perl6
05:06 pasteling "./pugs_r17176" at 193.200.132.135 pasted "Pugs build failure" (6 lines, 278B) at http://sial.org/pbot/26554
05:07 pro21ms4 joined #perl6
05:12 pasteling "./pugs_r17176" at 193.200.132.135 pasted "Pugs build failure" (6 lines, 278B) at http://sial.org/pbot/26556
05:13 diakopter <sigh> evalbot: grrrr
05:23 dduncan left #perl6
05:33 jisom joined #perl6
06:24 anatolyv joined #perl6
06:34 hakobe changed the nick to hkb_lab
06:40 elmex joined #perl6
06:56 masak joined #perl6
06:58 devogon joined #perl6
07:01 kane_ joined #perl6
07:04 kane_ joined #perl6
07:11 iblechbot joined #perl6
07:12 diakopter Haskerl: we take the "ask" out of "Haskell" and "FAQ"
07:28 masak diakopter: "Hell Frequently 'ed Questions"?
07:32 diakopter er, something like that.
07:32 flokuehn joined #perl6
07:34 agentzh joined #perl6
07:36 drrho joined #perl6
07:36 agentzh PCR's emitter often generates single boolean expressions for nodes in the regex AST...i'm wondering how to add debugging code to such expressions
07:37 agentzh maybe something like "expression ^ do { print $pos; 0 }" ?
07:37 agentzh ^ will never short-circuit, unlike && and
07:37 agentzh ||
07:38 f0rth joined #perl6
07:39 agentzh for AST nodes that emit code blocks instead of single expressions, adding tracing code is staightforward :)
07:40 f0rth joined #perl6
07:40 diakopter heehee: http://www.dcs.gla.ac.uk/~partain/haskerl/wall-1.html
07:40 lambdabot Title: A message about Haskerl
07:47 masak agentzh: looks ok, but if you anticipate doing that construct a lot, consider embedding it into a well-named subroutine
07:47 agentzh masak: i've been avoiding closures or subs because it's costy...
07:47 masak agentzh: ok
07:48 agentzh masak: but there's still a problem...
07:48 agentzh masak: i can't get the result of expression in the do block
07:48 agentzh only $pos is available :(
07:49 agentzh masak: how about this? $expr && ($bool = 1) xor do { print "$bool $pos\n"; 0 }
07:49 masak eh.
07:49 masak well, it works
07:49 agentzh yay
07:49 masak is this code that anyone will ever read? :)
07:49 agentzh it's the code generated by the PCR emitter
07:49 agentzh i think only PCR developers will read such code
07:50 agentzh hehe
08:08 masak agentzh: I guess if speed trumps maintainability/readability, anything goes
08:09 agentzh masak: okay, i'll wrap the code with a do block then. fighting with logical operators is no fun at all :)
08:09 masak true. good luck. :)
08:09 agentzh masak: thanks.
08:12 agentzh i think it's acceptable for PCR to slow down a bit in tracing mode...for production code, there will be no performance impact since there's only some special comments in non-tracing mode. :)
08:13 agentzh heh, PCR's emitter is already a bit hard to maintain...there's so many optimization tricks...
08:13 agentzh kp6 may make our lives easier.
08:14 masak hopefully.
08:14 agentzh :)
08:17 meppl good morning
08:17 masak meppl: morning
08:17 agentzh morning, meppl
08:18 meppl good morning masak and agentzh
08:35 daxim_ joined #perl6
08:36 anatolyv left #perl6
08:37 rgs joined #perl6
08:47 toly joined #perl6
08:56 franck___ joined #perl6
08:58 IllvilJa joined #perl6
08:59 drbean left #perl6
09:09 iblechbot joined #perl6
09:10 agentzh p6 grammar => p5 code with smart comments => p5 code with smart comments expanded as tracing code
09:11 agentzh does this compilation route look sane?
09:13 rindolf joined #perl6
09:22 ruoso joined #perl6
09:39 rfordinal joined #perl6
09:50 agentzh hmm, ANTLR's ParseView looks interesting.
09:52 agentzh is there any good visual debugging tool that i can learn from?
09:54 agentzh supper &
10:01 pro21ms4 joined #perl6
10:03 Aankhen`` joined #perl6
10:31 chris2 joined #perl6
10:33 nipotan changed the nick to nipotaway
10:44 kanru joined #perl6
10:49 cognominal_ joined #perl6
10:55 buetow joined #perl6
11:13 penk joined #perl6
11:20 buetow joined #perl6
11:20 Blwood joined #perl6
11:29 Southen joined #perl6
11:47 bonesss joined #perl6
11:56 cognominal_ joined #perl6
11:56 cognominal_ joined #perl6
11:57 Casan joined #perl6
12:10 agentzh left #perl6
12:16 xinming_ changed the nick to xinming
12:16 Aankh|Clone joined #perl6
12:48 lichtkind joined #perl6
12:49 drupek12 joined #perl6
13:17 jhorwitz joined #perl6
13:23 sunnavy joined #perl6
13:25 TJCRI joined #perl6
13:39 iblechbot joined #perl6
13:40 revdiablo joined #perl6
13:48 ruoso joined #perl6
14:29 thepler joined #perl6
14:39 amnesiac joined #perl6
14:43 cognominal_ joined #perl6
14:52 cognominal_ joined #perl6
14:55 cognominal_ left #perl6
14:55 cognominal_ joined #perl6
14:59 franck____ joined #perl6
15:04 pugs_svnbot r17177 | masak++ | [Prim/Code.hs]
15:04 pugs_svnbot r17177 | masak++ | * <anon> coderefs are now treated properly
15:04 pugs_svnbot r17177 | masak++ | * I have no idea what I am doing, am not a Haskell programmer, I just play one on TV. Code review appreciated
15:04 pugs_svnbot diff: http://dev.pugscode.org/changeset/17177
15:04 lambdabot Title: Changeset 17177 - Pugs - Trac
15:14 weinig joined #perl6
15:16 Eidolos_ joined #perl6
15:23 moritz_ masak++ # pugs hacking
15:24 masak thx :) haskell is alien but quite fun
15:25 drupek12 joined #perl6
15:30 moritz_ http://www.google.com/trends?q=perl6%2C+perl+6&amp;ctab=0&amp;geo=all&amp;date=all&amp;sort=0
15:30 lambdabot Title: Google Trends: perl6, perl 6, http://tinyurl.com/238f73
15:31 cmarcelo joined #perl6
15:32 masak wonder what the spikes at the end of 2005/beginning of 2006 were?
15:32 masak something to do with releases of Pugs, probably
15:35 moritz_ or christmas and thus p6 release being near ;)
15:36 masak that must be it
15:37 diakopter SPJ's Oscon Haskell primer; I dunno if there's a video: http://conferences.oreillynet.com/cs/os2007/view/e_sess/14016 http://conferences.oreillynet.com/presentations/os2007/os_peytonjones.pdf
15:37 lambdabot Title: O'Reilly Open Source Convention 2007 &#8226; July 23-27, 2007 &#8226; Portland,  ...
15:43 pugs_svnbot r17178 | masak++ | [Prim/Code.hs]
15:43 pugs_svnbot r17178 | masak++ | * corrections from auto-review ;)
15:43 pugs_svnbot diff: http://dev.pugscode.org/changeset/17178
15:43 lambdabot Title: Changeset 17178 - Pugs - Trac
15:47 cerridwen joined #perl6
15:57 justatheory joined #perl6
15:58 iblechbot joined #perl6
16:04 ruoso joined #perl6
16:16 rindolf joined #perl6
16:17 daxim joined #perl6
16:17 melissa_ joined #perl6
16:19 franck___ joined #perl6
16:22 IllvilJa joined #perl6
16:36 REPLeffect_ joined #perl6
16:43 Snake707 joined #perl6
16:44 melissa_ joined #perl6
16:49 Snake707 left #perl6
16:50 jisom joined #perl6
16:53 hkb_lab changed the nick to hakobe
16:56 dduncan joined #perl6
16:57 Eidolos_ changed the nick to Eidolos
17:12 araujo changed the nick to araujoIsNotarmin
17:20 Psyche^ joined #perl6
17:32 Psyche^ changed the nick to Patterner
17:41 cognominal_ joined #perl6
17:42 cognominal_ joined #perl6
18:02 perlDreamer joined #perl6
18:02 rindolf joined #perl6
18:02 Jamtech joined #perl6
18:05 araujoIsNotarmin changed the nick to araujo
18:05 semifor joined #perl6
18:12 Schwern joined #perl6
18:39 lumnar joined #perl6
18:42 aindilis joined #perl6
18:44 pugs_svnbot r17179 | bcmb++ | Update smart links and correct 1 test script.
18:44 pugs_svnbot diff: http://dev.pugscode.org/changeset/17179
18:44 lambdabot Title: Changeset 17179 - Pugs - Trac
18:46 moritz_ bcmb++ # cage cleaning
18:47 cernd joined #perl6
18:47 [particle] kennel cleaning?
18:47 moritz_ [particle]: they are all just containers ;)
18:54 franck____ joined #perl6
18:58 aindilis` joined #perl6
19:11 pugs_svnbot r17180 | moritz++ | kp homepage: removed bogus </a>, added fglock++ as the main author
19:11 pugs_svnbot diff: http://dev.pugscode.org/changeset/17180
19:11 lambdabot Title: Changeset 17180 - Pugs - Trac
19:16 hakobe changed the nick to hkb_zz
19:30 moritz_ in the html of http://www.pugscode.org/ there are <nobr>...</nobr> tags, which are not html
19:30 lambdabot Title: Pugs - pugscode
19:30 apeiron joined #perl6
19:30 thoughtpolice hi moritz_
19:30 moritz_ any idea what they should be? span? or just nothing?
19:30 moritz_ hi thoughtpolice ;)
19:33 moritz_ I'll just delete those tags now
19:33 moritz_ afterwards you can hit me if you want ;)
19:33 pugs_svnbot r17181 | moritz++ | pugscode.org homepage: removed bogus tags
19:33 pugs_svnbot diff: http://dev.pugscode.org/changeset/17181
19:33 lambdabot Title: Changeset 17181 - Pugs - Trac
19:36 moritz_ thoughtpolice: so what's up?
19:36 thoughtpolice moritz_: not a whole lot actually. i've been more busy than i figured lately so I haven't had time to blog/program much in the past couple weeks :/
19:37 moritz_ thoughtpolice: same for me, too much $work (it's mostly perl, though ;-)
19:38 moritz_ but last weekend I managed to implement fulltext search in the logs ;)
19:38 thoughtpolice good stuff. :)
19:39 thoughtpolice i have been reading through the new erlang book (really good) and spj's old 'implementation of functional languages'
19:39 thoughtpolice i must say spj is quite eloquent and writes/speaks very nicely. i watched the OSCON talk on transactional memory systems.
19:40 * moritz_ pities himself for living so far from OSCON ;-)
19:41 thoughtpolice i watched it on blip.tv :p
19:41 moritz_ ;)
19:41 pugs_svnbot r17182 | moritz++ | kp homepage: fixed link to examples
19:41 pugs_svnbot diff: http://dev.pugscode.org/changeset/17182
19:41 lambdabot Title: Changeset 17182 - Pugs - Trac
19:42 diakopter thoughtpolice: I'm reading the same (SPJ 1987)
19:44 thoughtpolice diakopter: SPJ 1992 here (.pdf download from his site)
19:45 diakopter hrm
19:45 thoughtpolice http://research.microsoft.com/~simonpj/Papers/pj-lester-book/
19:45 lambdabot Title: Simon Peyton Jones: book
19:47 diakopter oh, that's the much-abridged version.
19:49 diakopter Relationship to The implementation of functional programming languages:
19:49 diakopter An earlier book by one of us, [Peyton Jones 1987], covers similar material to this one, but in a less practically oriented style. Our intention is that a student should be able to follow a course on functional-language implementations using the present book alone, without reference to the other.  The scope of this book is somewhat more modest, corresponding to Parts 2 and 3 of [Peyton Jones 1987]. Part 1 of the latter, which ...
19:49 diakopter ... discusses how a high-level functional language can be translated into a core language, is not covered here at all.
19:50 diakopter in other words, Part 1 of the 1987 work contains lots of explanatory background/theory that would be helpful to someone who's, say, writing a new (say, for Perl 6) frontend for GHC.
19:52 diakopter I don't mean to imply that the tutorial isn't good or anything... I'm just saying, it's different from what I'm reading (and I haven't gotten to needing a tutorial yet.. I will soon, I hope).
19:52 thoughtpolice ah. ty. :)
19:53 weinig_ joined #perl6
19:56 weinig joined #perl6
19:59 sysdef joined #perl6
20:02 * moritz_ wanders off to bed &
20:05 thoughtpolice later moritz_
20:09 Teratogen joined #perl6
20:59 buetow joined #perl6
21:01 cerridwen changed the nick to julchen
21:01 julchen changed the nick to cerridwen
21:11 meppl good night
21:28 riffraff joined #perl6
21:35 cernd joined #perl6
21:55 SamB joined #perl6
21:56 demq joined #perl6
22:14 thorat joined #perl6
22:16 polettix joined #perl6
22:16 Jamtech joined #perl6
22:21 lichtkind joined #perl6
22:28 weinig joined #perl6
22:42 pugs_svnbot r17183 | bcmb++ | Correct smart links in several test scripts
22:42 pugs_svnbot diff: http://dev.pugscode.org/changeset/17183
22:42 lambdabot Title: Changeset 17183 - Pugs - Trac
22:46 Limbic_Region joined #perl6
22:53 Caelum joined #perl6
22:56 weinig_ joined #perl6
23:06 REPLeffect_ joined #perl6
23:07 REPLeffect_ changed the nick to REPLeffect
23:23 perlDreamer TimToady: In S29, there are only 3 argument forms of substr defined.  Does that mean that the 2 and 4 argument forms from perl5 do not exist in perl6, or that since they are unchanged they remain?

| Channels | #perl6 index | Today | | Search | Google Search | Plain-Text | summary

Perl 6 | Reference Documentation | Rakudo