tcltls hanging in self-tests? (Re: cvs commit: ports/devel/tcllib Makefile)

[Kris Kennaway]

On Wed, Jan 24, 2007 at 06:47:53PM -0500, Mikhail Teterin wrote:
> On Wednesday 24 January 2007 16:40, Kris Kennaway wrote:
> = > This is the first time I hear about the problem -- there is nothing
> = > unusual or unreasonable about asking the problem-reporter for the
> = > stdout/stderr, is there?
> = 
> = No, but you have to also accept that I can't oblige, as previously stated.
> 
> Indeed:
> 
> = > Can I see the build-log, please? Which test is hanging?
> 
> = I don't know, I haven't had the time to debug it (the build log does
> = not exist since it does not complete).
> 
> I accept that, and shall wait until you find the time to collect the output.
> 
> A semi-educated guess: the problem may have to do with the build-machine 
> dropping (rather than nacking) attempts to connect to a closed port. Is 
> pointyhat (mis)configured this way?

I answered this earlier too...the build machines do not behave in this
way (you can verify yourself, they just build in a jail with randomly
assigned 127/8 address).  Maybe it's another assumption about
127.0.0.1.

Kris