On 26/08/08 at 14:48 +0200, Michal Suchanek wrote:
> On 26/08/2008, Lucas Nussbaum <[EMAIL PROTECTED]> wrote:
> > On 23/08/08 at 11:46 +0200, Michal Suchanek wrote:
> > > Package: ruby
> > > Version: 4.2
> > > Severity: normal
> > >
> > >
> > > The attached code fails after about 100 iterations of invoking popen.
> > >
> > > However, on this particular interpreter version inserting a comment into
> > > the code makes it not fail.
> > >
> > > On stable the code fails regardless of comment.
> >
> >
> > It's not a ruby-defaults problem, but a ruby1.8/ruby1.9 one (both are
> > affected).
> >
> > It seems that there's a probably when several IO objects point to the
> > same fd.
> > Changing:
> > t = Thread.new( (IO::new analyzer.fileno),res){|fd,ary|
> > To:
> > t = Thread.new(analyzer, res){|fd,ary|
> > makes the problem disappear.
> >
> > Could someone raise this issue on ruby-dev?
>
> If you mean the Japanese list then don't look at me.
>
> I have posted about the issue on the English ruby-core and got no replies.
>
> It is expected that the two objects point to the same fd, that's what
> the code does.
>
> However, the issue I tried to raise here is that the Ruby in Debian
> behaves completely differently when a comment is added in the code.
> The latest 1.8.7 release fails a bit later with the comment but still
> fails.
Well, the comment might cause a slightly different scheduling to happen.
Here, it fails both with and without the comment.
--
| Lucas Nussbaum
| [EMAIL PROTECTED] http://www.lucas-nussbaum.net/ |
| jabber: [EMAIL PROTECTED] GPG: 1024D/023B3F4F |
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
