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

<GummiEnte>Well, that it should be...19:20
<arjanb>maybe another name for your log in case other one comes back?19:21
<GummiEnte>Ok, just make a proposal please.
<arjanb>_log or Gummi_log19:22
<GummiEnte>Ok, that should be better...19:27
You`ll find the log at http://confer09.condor-edv.com/nice@freenode19:28
<GummiEnte>Hopefully the executable will be stabnle enough. I`ll build some check scripts tomorrow...19:31
Have you also asked if it is possible to split the nice.lang package?19:35
<GummiEnte>Well, maybe Daniel comes back and you`re still here...
I`ll have to go now...
I wish you a good night...
..and Daniel too.
<arjanb>good night19:37
* GummiEnte leaves
<bonniot>arjan, in the documentation you quote, why isn't there a compilation error in B when it does not implement foo?20:31
anyway, I think the change to the compiler should not be too difficult, I will look at it tomorrow
<bonniot>your above quote about A and B
abstract A declares foo
concrete B implements A but not foo
I suppose it has to do with modifying A without recompiling B20:33
the quote is about the jvm, and not Java, right?20:34
<arjanb>i have no idea it's literally from the jvm spec
<bonniot>about the coverage: it is known that this is not done for methods declared in Java
<bonniot>i wonder if you can override a method by an abstract method in an abstract class
that would allow to create a testcase
it could be good to have another case where the offending bytecode is not obviously never used like in this one20:37
because here we could just do the optimization, but that would not solve the general case
<bonniot>also, the testengine could be modified to allow the compilation by gcj, which would make a real testcase
<arjanb>i'm know trying to fix the printing of loop statements
i'll look at the bytecode gen tomorrow20:39
<arjanb>good night
* bonniot leaves
* arjanb leaves21:37
* arjanb joins
* arjanb leaves22:23

Generated by Sualtam