Camelia, the Perl 6 bug

IRC log for #ironperl, 2009-03-17

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

All times shown according to UTC.

Time Nick Message
03:50 diakopter joined #ironperl
03:50 cj joined #ironperl
05:48 cj joined #ironperl
05:49 cj joined #ironperl
05:49 cj__ joined #ironperl
17:35 diakopter <sigh>
17:35 diakopter maybe it could be a better (by way of being more feasible) idea to implement Perl 5...
17:47 diakopter cj: you around?
17:47 cj sorta
17:48 cj I'm at a coffee shop on an island.  I'm going to head out and pick up the big one from school shortly
17:48 cj with a working perl5, the current perl6-on-perl5 approach would at least enable something on the dlr
17:49 cj have you seen the gcc-cil (or do they call it gcc-cli?) back end?
17:49 cj http://gcc.gnu.org/projects/cli.html
17:49 cj they probably have libm working now
17:50 cj it wouldn't be a first-class citizen out the door, but it would be perl on a cli
17:50 cj ooh, binutils!
17:50 diakopter what current perl6-on-perl5 approach?
17:51 cj isn't there one? :)  maybe I was thinking of flavio's parser/emitter
17:51 diakopter no work on it in more than a year
17:51 diakopter and it was abysmally slow, like all the other implementations on Perl 5
17:52 diakopter and even if Perl 5 were working on the DLR, that would be (likely) a few orders of magnitude slower.
17:53 cj step 1: make it work.  step 2: make it work right.  step 3: make it fast
17:54 diakopter no, I mean unusably (ever) slow, so slow that even testing during development is unreasonably slow.
17:55 diakopter implied component of your step 1 is to be rfasfast enough to make it work.
17:55 cj so be more iterative.  make one test.  make it pass.  improve speed.  add another test.  be sure they both past.  improve speed.
17:56 cj s/past/pass/
17:56 cj contract ending before lang.net.  still planning on coming? :)
17:57 diakopter your contract ending?  are you still going?
17:57 cj y
17:57 cj bby in l hnd
17:57 cj I just told said bby that I'd say goodbye and we can go :)
17:58 diakopter last two msgs opaque to me
18:02 diakopter besides, even Perl 5 would be nigh infeasible to try implementing with a much larger team
18:02 diakopter without a much larger team, I mean
18:12 * diakopter follows your suggestion to read about gcc-cli

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