On Sun, Jan 04, 2009 at 05:20:56PM +0300, Richard Hainsworth wrote: > I posted an email to per6-all asking about how one should go about > reporting bugs. That message has appeared on the list. > > So again: how can bugs be reported?
See the "Reporting bugs" section of README file in languages/perl6. (Honestly, I don't know why we call such files "README", since nobody ever does. I think I'll rename the file to "IGNOREME" -- maybe it'll get some attention then. :-) > Eventually, I found reference to perlbug and opened an account. But I > cant find a way to submit a bug. > > Some requests: > a) When I strike a bug, perl6 (rakudo) dumps alot. Could this behaviour > be modified with a verbose option on the perl6 command line? I'm working on this. Parrot doesn't give us a lot of control over backtrace behavior at the moment -- it's either all-or-nothing. I might implement a switch somewhere. > b) Also, what needs to be stated? I used svn to download the latest > parrot. So, I should be stating the latest revision number. "perl6 -v" > doesnt provide this. We haven't come up with a good way to fetch the latest revision number for use with a -v option to perl6. If you know of one (and that will work even if someone obtains Rakudo via tarball or git or a non-svn mechanism), we can certainly add it. > I was going to describe the bugs, but I just checked the test results in > S05 and found the things I cant get working are not yet implemented. > > There isnt any table of these anywhere. Could we get one? http://www.perlfoundation.org/perl6/index.cgi?rakudo%20feature%20status Feel free to modify this list. > > Patrick! Do you want feed back about things that cant be done, but > should be done? > > If so, I would like to have m:i/ / and m/ a {say 'yes'} / m:i/ ... / can currently be done with /:i ... / . Getting closures to work inside of regexes is a bit tricky at the moment, because it implies that PGE (our regex engine) has to know how to call the Perl 6 parser and do the right thing with whatever it gets back. That's coming in the not-too-distant future, but it's by no means "trivial" to implement. Pm
