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

Using timezone: Central European Time
* arjanb leaves02:23
* arjanb joins13:09
* fcb joins15:54
<fcb>I'm hoping to join the nice team
<fcb>I noticed the ad refering to the nicedoc project, but I'd be happy to do whatever really16:03
<arjanb>the compiler needs quite a lot of work too but it's easier to get started on something like nicedoc16:06
i assume you discussed about nicedoc with daniel16:12
<fcb>yes that's right
(apologies for the delay)
anyway, I'm not a very experienced programmer (I've just finished honours at university), so I'm just going to have a play around with the language first16:14
<arjanb>experience doesn't matter much i'm just a third years cs student16:19
<fcb>oh, cool16:21
have you been in the project long?
<arjanb>and you don't have to know all the details of the language but getting familar with the concepts of the language is important16:22
<fcb>Nice seems like an interesting language - I haven't done alot of functional programming before, so that will be interesting16:23
<arjanb>i have discovered Nice about a year ago and started working on the compiler in March
<arjanb>have you written anything in Nice yet?16:30
<fcb>no, not yet. I've only been looking into for a couple of days
that's my next task. I'
m just going to play around with the language to get used to it16:31
<arjanb>I see
* bonniot joins
<bonniot>are you Frank?
<fcb>yes, we have been communicating via email for a couple of days
<bonniot>great, pleased to meet on here!
did you start playing with Nice?
<fcb>no, but I've just downloaded it16:33
<bonniot>ok. what OS do you run? what editor do you plan to use?16:34
<fcb>I'm running redhat 9 at the moment - I have a windows xp machine too
however, redhat have just stopped doing free OSs, so I might have to switch to debian16:35
I dont
have a text editor that i'm particularly found of. I don't really like emacs and vi is a bit clumsy
<bonniot>i am a happy user of Debian myself :)
<fcb>the installer failled for me when I tried to install it16:36
<bonniot>so how do you edit files?
<fcb>vi or emacs - but I
<bonniot>the rpm package?
oh, sorry - the debian installer
as in, trying to install the debian OS16:37
nice was fine to install
<bonniot>i see
<fcb>I've been using Forte4Java to write my Java programs
<bonniot>the Debian package is better integrated in the system
for instance if you want to use ant16:38
so you don't dislike IDEs?
<fcb>no, not really16:39
<bonniot>in that case you can try Eclipse. there is a plugin for nice
<fcb>OK, I'll try it
<bonniot>Eclipse and emacs are the two editors that have some support at the moment16:40
more would be welcome of course, but that's already quite decent to have these two, especially as they are quite complementary
<fcb>I don't know what part of the world everyone else is in, but here in Australia it getting a bit late16:41
<bonniot>it's 16:42 here in Budapest16:42
<fcb>so without being rude, I might go to bed and have a play with nice tomorrow
<bonniot>it's fall too16:43
<arjanb>how much is the time difference?
<bonniot>ok, goodnight. see you soon!
<fcb>here it is 23:43
yes, 'nice' to talk to you all (I bet you get sick of that joke)16:45
<bonniot>no, it's a nice one :-)
<bonniot>better than for the PL beta
* fcb leaves16:46
<bonniot>people kept joking about its beta version state....
how do you pronounce Nice? as in english or as the city16:47
<bonniot>as in english16:49
what about you? :-)17:33
<bonniot>how do you pronouce it?
<arjanb>as in english but I doubted about that because it's a odd name for a language17:36
<bonniot>why? there is also a language called Clean17:41
the situation for debuging is improving fast18:36
after my changes to implement jsr45, jdb is already working18:37
<arjanb>never used that yet
<bonniot>i am now submitting two small patches for JSwat, which would make it support showing the Nice source
<arjanb>is bootstrapping the cvs version possible again?18:38
<bonniot>yes, me neither, but eventually it can be useful, and some people will want it (alex did)
as far as I know yes
<arjanb>JSwat looks good but i never found debuggers very useful18:45
<bonniot>you use println?19:01
<bonniot>the advantage of a debugger is that you don't need to modify your source19:26
<arjanb>it looks like i have ignoring of retypings finally working21:34
<bonniot>good, i'm looking at it22:42
<arjanb>could you try it with some real 1.4 retyping22:43
i can't get both 1.3 1.4 working at mine machine
<bonniot>does getSyntacticReturnType() has anything to do with the retypings?22:44
for some reason is the real returntype lost in case of an ignored retyping22:47
<bonniot>it's not too clean to have this cast in 'inInterfaceFile', i'll fix that22:52
can you access SF by ssh just now?23:00
it seems to work again23:09
<bonniot>is there a reason to have created a separate testcase for each case?23:36
<arjanb>not really but it's useful when writing the code23:39
<bonniot>i understand that23:40
the thing is that running one testcase with four lines will be about 1/4 of the time needed for four testcases
so when there are several PASS cases that are similar, they can be put together23:41
one shouldn't do that for FAIL cases, of course
otherwise, the testsuite will get too long to run :-/
<arjanb>yes maybe we should go after all testcases to find which could be merged23:42
<bonniot>at least, please merge those four23:43
of course, one must be very careful when touching any testcase, not to interfere with the case23:44
since the bug is normally already fixed, it would be easy not to test the feature anymore and not notice it23:45
if you can find a few testsuite files that have *many* similar *very simple* PASS cases, it could be good to merge them23:47
what do you think should happen when the program tries to called an ignored retyping?23:50
at the moment, it is not satisfactory23:51
<arjanb>at least some error that the method doesn't exist23:52
or better saying that it's a ignored retyping
you had plans to use a coverage tool some time ago have you found an alternative for that?23:56
it's possible to use that tool
<arjanb>what is the 'but' here?00:02
<bonniot>the report is actually on the web00:10
so it was used
i just don't think it's a priority now to optimize the coverage00:13
one problem is that the webpages for the report take a lot of space00:14
<arjanb>yes if it print the entire source in the reports00:23
<bonniot>ok, i have an improvement00:27
it allows to know if a symbol was ignored, and in that case will print a useful message
it also removes the "trick" of setting the type to void00:28
i will commit it now, except the reporting part, because that one is mixed with another change
<arjanb>you gave RetypedJavaMethod it's own symbol?
it's ok to me00:29
<bonniot>no, but I added a method isIgnored to symbols, which is redirected to the method00:32
it could be used in other cases where symbols are ignored too00:33
<arjanb>i see00:44
what to do you think of proposal for block call syntax?01:02
<bonniot>it's interesting. i will have to think about it again01:37
sorry, but i was going to bed now. goodnight!01:38
<arjanb>me too
good night
* bonniot leaves
* arjanb leaves01:41