Log of the #nice channel on irc.freenode.net

Using timezone: Central European Time
* zzorn_sleep leaves03:04
<CIA-5>03bonniot * 10Nice/src/bossa/syntax/methodbody.nice: Avoid dumping cryptic exceptions on the poor user!04:08
* Cybertnt joins04:16
* Cybertnt leaves
<CIA-5>03bonniot * 10Nice/ (3 files in 2 dirs): Optim: avoid checking the type of default values for method parameters in compiled packages.04:24
* arjanb leaves06:11
<CIA-5>03bonniot * 10Nice/src/bossa/syntax/ (nicefield.nice niceclass.nice): Optimization: avoid checking the type of fields in compiled packages.06:54
* dreese9859 leaves07:19
* dreese9859 joins08:14
* bonniot leaves08:31
* bonniot joins08:32
* Bluelive joins09:09
parsing 0 as an octal is very common09:11
c and java do that
* ChanServ leaves09:58
* ChanServ joins09:59
<Bluelive>do you use antlr ?10:09
<bonniot>no11:02
although we might want to switch sometim
e
<Bluelive>ah just found a tool that will print a railroad diagram from the antlr syntax11:04
<bonniot>;-)11:08
<Bluelive>really reminds me off the pascal syntax that was in the back of all those books11:15
* arjanb joins12:34
<bonniot>hi arjan12:41
<arjanb>hi
<bonniot>last night's optims save about 15% on compiling sokoban (small swing app)12:42
are there no testcases for field value override in compiled packages, or did I miss them?12:53
<arjanb>they don't exist i think12:56
* zzorn joins13:04
<bonniot>would be good to add some13:11
i did not change value overrides because of that13:12
i guess they are not too much of a performance problem at the moment, but it's still important to have these tests for correction
i will be traveling for a week, so less present13:13
but hopefully i should be able to read emails
i guess the missing failure location warnings will block the new dev version until solved13:19
<arjanb>ok13:20
<bonniot>do you have plans?13:51
<arjanb>doing failure locations, tuples and maybe partial application14:11
<bonniot>sounds good14:16
:-)
for tuples, i guess the trick is to give share constraint to all components, and featch each component type in the tuple type14:31
after that it should be the same as a normal let
if it's problematic, you could restrict to monotypes tuples, without constraint14:32
bye14:36
<arjanb>bye
* bonniot leaves
* cm_ joins16:05
<arjanb>hi16:09
<cm_>hi16:46
* Bluelive leaves
* Bluelive joins17:13
* Bluelive leaves17:21
* Bluelive joins17:28
:) yes, got the pluk website up again :)19:23
* dreese9859 leaves20:29
* Bluelive leaves20:45
* Bluelive joins20:46
* cm_ leaves22:16
* zzorn_afk leaves23:17

Generated by Sualtam