Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

ISO: Tk 8.5a6 developers using Windows and Hummingbird Exceed 10

9 views
Skip to first unread message

Larry W. Virden

unread,
Jan 23, 2007, 12:50:38 PM1/23/07
to
Maybe I'm in a peculiar situation here. But I do nearly 100% of my
development on sparc solaris, but due to forces beyond my control, use
exceed and windows xp as my "window" to the sparcs.

If there are others using windows and exceed more as "x terminals" than
as development platforms, could you drop me an email? I'm curious
whether you've run the Tk 8.5a6 test suite in your environment.

I see about 3 times more test suite errors running in this combination
than I do with the identical source/binary but running against the Sun
X server. I don't know whether there is a Tk issue, an Exceed 10 issue,
or a local configuration of Exceed issue or a problem with the copy on
this particular machine.

Thanks!

Slaven Rezic

unread,
Jan 23, 2007, 4:00:26 PM1/23/07
to

It's about six years I used Exceed, but I still have this line in one
of my (Perl/Tk) applications for workarounding a canvas capstyle bug:

# Exceed-Bug
$capstyle_round = ($exceed ? "projecting" : "round");

I guess there are other bugs, too. Why not try the cygwin X11 server?

Regards,
Slaven

--
Slaven Rezic - slaven <at> rezic <dot> de
BBBike - route planner for cyclists in Berlin
WWW version: http://www.bbbike.de
Perl/Tk version for Unix and Windows: http://bbbike.sourceforge.net

Wojciech Kocjan

unread,
Jan 23, 2007, 4:39:29 PM1/23/07
to
Dnia 23-01-2007 o 22:00:26 Slaven Rezic <sla...@rezic.de> napisał(a):
> It's about six years I used Exceed, but I still have this line in one
> of my (Perl/Tk) applications for workarounding a canvas capstyle bug:
>
> # Exceed-Bug
> $capstyle_round = ($exceed ? "projecting" : "round");
>
> I guess there are other bugs, too. Why not try the cygwin X11 server?

I second that, although I personally find Xming easier to use and a bit
less problematic (cygwin tends to have a few issues here and there I don't
like).

URL: http://sourceforge.net/projects/xming

--
WK
email (rot13): jbwpvrpuNGxbpwnaQBGbet

Larry W. Virden

unread,
Jan 24, 2007, 7:51:18 AM1/24/07
to

On Jan 23, 4:39 pm, "Wojciech Kocjan" <wojciec...@gazeta.pl> wrote:
> Dnia 23-01-2007 o 22:00:26 Slaven Rezic <sla...@rezic.de> napisa³(a):


> >Why not try the cygwin X11 server?
>I second that, although I personally find Xming easier to use and a bit
> less problematic (cygwin tends to have a few issues here and there I don't
> like).

Ah, if this was a personal thing, where I was free to change things
like x servers, I would be using Linux or Intel Solaris, and it
wouldn't be an issue. This, however, is an issue on the work side of my
life, in a situation where those types of issues are not, alas, open
for negotiation.

Darren New

unread,
Jan 24, 2007, 11:33:59 AM1/24/07
to
Larry W. Virden wrote:
> whether you've run the Tk 8.5a6 test suite in your environment.

I noticed many years ago that running the Tk test suite while other
things were running on the machine caused many failures. There were at
the time a number of tests that involved changing something, waiting a
tenth of a second or so, then checking if it had updated properly.
Running the same tests on a non-loaded machine made them pass.

Perhaps it's mostly a timing issue?

--
Darren New / San Diego, CA, USA (PST)
You know it's time to trim your beard when your
wife stops calling you Husband and starts calling
you Husbollah.

Larry W. Virden

unread,
Jan 24, 2007, 3:31:22 PM1/24/07
to

On Jan 24, 11:33 am, Darren New <d...@san.rr.com> wrote:
> Larry W. Virden wrote:

> > whether you've run the Tk 8.5a6 test suite in your environment.I noticed many years ago that running the Tk test suite while other


> things were running on the machine caused many failures. There were at
> the time a number of tests that involved changing something, waiting a
> tenth of a second or so, then checking if it had updated properly.
> Running the same tests on a non-loaded machine made them pass.
>
> Perhaps it's mostly a timing issue?

I can't tell - but probably not. The kinds of errors being generated
are more fundamental problems - problems in select, etc. But thanks for
the thought though.

0 new messages