Re: Undef issues

2005-05-24 Thread Adrian Taylor
Hi Stuart, You (and Carl) are absolutely right... all these things are behaving as they should. Cool! :-) I'll fix the tests later on today. Thanks, Ade Stuart Cook wrote: >On 5/24/05, Adrian Taylor <[EMAIL PROTECTED]> wrote: > > >>eval_is('undef + 1'

Undef issues

2005-05-24 Thread Adrian Taylor
Hi, Over the weekend I added some tests on 'undef' behaviour (t/builtins/undef.t): These behave as expected: eval_is('undef * 2', undef, 'undef * 2'); eval_is('undef * undef', undef, 'undef * undef'); These don't (at least, according to my expectations...) eval_is('undef + 1', undef, 'undef +

Re: Method/attribute chaining

2005-05-24 Thread Adrian Taylor
Hi, I have seen similar weirdness and added a couple of tests over the weekend, whilst simultaneously forgetting to mention it on IRC: class Foo { method noargs () { 42 } method callsmethod2 { .noargs(); } } $val = $foo.callsmethod2() dies. (t/oo/method/instance.t) It appears to be someth

Re: [perl #35450] Null pointer dereference in classes/sub.pmc

2005-05-15 Thread Adrian Taylor
/pugs/pugs] ade% ../parrot/parrot/parrot --version This is parrot version 0.2.0-devel (r8105) built for ppc-darwin. On 15 May 2005, at 13:55, Leopold Toetsch wrote: Adrian Taylor <[EMAIL PROTECTED]> wrote: I can't reproduce the segfault, when running: rule name { Larry | Matz |