Ah!
Allan and I seem to share a similar opinion/conjecture/analysis of the
problem.
Pushing the "Cancel" button in the finding control points window
yields an error because the control point generator has successfully
completed *and quit* (That is, there is no longer anything left to
"Cancel.")
But what I'm calling the "hand off" to Hugin never happens. For some
reason, Hugin misses the "I'm done, here are those control points you
asked for" event.
And bothering to go a step further than I did before, in
/var/folders/Hg/HgTVbhBpG7GpGWUU2vjMhE+++TI/-Tmp-
are files such as ap_resj0h6w4, ap_resovgvgE, ap_resR4vOEu and so
forth that are indeed control points files. (Thanks Allan!) At least
the ones for Autopano-sift-C and Panomatic. There are also some files
with that naming convention that are empty, but I'm guessing those are
from patfree-panomatic (which this time around was crashing).
So! This ought to be easy to fix! :) ... right?
If the problem only happens on PPC machines, I don't suppose it might
have something to do with byte order, would it? (He said naively.)
eo
On Apr 23, 2010, at 8:24 PM, AKS-Gmail-IMAP wrote:
Harry,
On Apr 23, 2010, at 12:24 PM, Harry van der Wolf wrote:
About 15-16 months ago we needed to create a workaround for wxmac
on Tiger/ppc versus Leopard/ppc as leopard did not behave correctly.
For this exact same error as mentioned above the work-around was
created (in collaboration with a dev from the wxwindows team). It
now seems that tiger/ppc misbehaves the same, so I now applied the
same leopard/ppc workaround also for tiger/ppc.
I have no idea whether it works or not.
For all of you on Tiger on G4/G5, please download <http://hugin.panotools.org/testing/hugin/Hugin.app.zip
> and see whether it works or not.
This problem has become a PPC issue regardless of having Leopard or
Tiger. I see 2010.0.0 and 2010.1.0-svn5031 as the only versions
running on a PPC that do not exhibit this failure to know that the
Auto CP generators have successfully finished.
As you recall, that is basically what happens. The completed points
file winds up stranded in a tmp folder and the process that created
it unloads. I would guess that the Canel button on the "finding
control points" pane results in the "failed to kill" message because
it is trying to kill a process that no longer exists. One can watch
that process come and go on the Activity Monitor. Comparing the
differences seen on the Activity Monitor between svn5031 and svn5116
(any version), I see the "Control point detector results" pane
appear before the generator process disappears from the Activity
Monitor process list when running the svn5031. When running svn5116
one sees the generator process finish in the status pane and then
instantly disappear from the Activity Monitor list as one would
expect knowing that completed points files are left stranded in tmp.
Allan
--
You received this message because you are subscribed to the Google Groups "hugin and
other free panoramic software" group.
A list of frequently asked questions is available at:
http://wiki.panotools.org/Hugin_FAQ
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to
[email protected]
For more options, visit this group at http://groups.google.com/group/hugin-ptx
