[hugin-ptx] Re: segmentation fault when stitching

2009-01-12 Thread Kornel Benko
Am Monday 12 January 2009 schrieb Lukáš Jirkovský:
 Fortunately someone did more testing on the gcc side
 (http://gcc.gnu.org/bugzilla/show_bug.cgi?id=38625) and the cause was
 found. I'm attaching small patch to configure.in against newest cvs
 which adds necessary flag to CXXFLAGS when compiler is GCC.
 Can someone test it? It's still a little bit hackish, but it works for me.


Yes, but it works. I tried some projects, and it did not crash.

Very nice bug indeed.

Kornel

-- 
Kornel Benko
kornel.be...@berlin.de


signature.asc
Description: This is a digitally signed message part.


[hugin-ptx] Re: GSoC 2009 (was: *Your* Opinion is requested about Google Summer of Code 2009 and Community Funding)

2009-01-12 Thread prokoudine

On Jan 11, 6:20 pm, Yuval Levy goo...@levy.ch wrote:

 2007 I was admin and you (Alexandre) were backup admin. 2008 we inverted
 roles. My wish for 2009 is to bring in fresh people. If you want to back
   the new admin up, I'll gladly take a further step back. I can always
 mentor a student. We need to spread the know-how in the community so
 that others can continue after us.

Well, whatever comes, you can rely on me as either primary or backup
admin.

 I've a son now and I hope he will be applying as a student in 2027 ;-)

:-)

Alexandre
--~--~-~--~~~---~--~~
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 hugin-ptx@googlegroups.com
To unsubscribe from this group, send email to 
hugin-ptx-unsubscr...@googlegroups.com
For more options, visit this group at http://groups.google.com/group/hugin-ptx
-~--~~~~--~~--~--~---



[hugin-ptx] Re: 0.61 update of Panini perspective tool available

2009-01-12 Thread Bart.van.Andel

 Aaaarghh!  Mac users are reporting all sorts of problems with cubic
 format, you are not alone...

 But why only on Macs?  And with 3 different video systems at least?

For the record, I don't have a Mac (I only wish I had one, haha). I'm
running WinXP SP3 and just updated my graphic card drivers (ATI Radeon
9600SE) to the newest version. I just installed the drivers, not that
piece of memory consuming bloatware called Catalyst.


 I wouldn't have
 any idea how to set up mapping of multiple texture images onto one
 surface.  If you do know, please tell me.

Well, basically they're just 6 rectangular projections 90 degrees
apart, which shouldn't be too difficult to backproject onto the
panosphere. I guess you're already doing the same thing when loading a
rectangular image?


 Am I safe in assuming that you can display the non-cubic formats OK?

Correct. Only trouble with cubics.


  I'll see if I can figure out more details later.

 Any help you can give would be most welcome.  I am not a Mac developer
 and have no Mac to test on.

Me neither, unfortunately. I've read through the source code a bit and
couldn't find anything wrong there. Moreover, I compared your code to
some of the code available on other websites (by googling), and if not
pretty much an exact copy, it's still fairly the same amount and order
of operations. Strange.

If you look closer at the picture I linked to in my previous post,
it's not only that the same image is shown on every face. It's shown
with a *different* error on every face. The images are shifted in both
X and Y directions (in texture coordinates), and have border issues
which differ from face to face. It popped into my mind that something
might be wrong with texture coordinate generation internally, but this
does not quite explain the cloning behavior.

By the way this is the first time I dive into OpenGL, so don't expect
me doing wonders just yet ;)

Cheers,
Bart
--~--~-~--~~~---~--~~
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 hugin-ptx@googlegroups.com
To unsubscribe from this group, send email to 
hugin-ptx-unsubscr...@googlegroups.com
For more options, visit this group at http://groups.google.com/group/hugin-ptx
-~--~~~~--~~--~--~---



[hugin-ptx] Re: Community Funding (was Re: *Your* Opinion is requested about Google Summer of Code 2009 and Community Funding)

2009-01-12 Thread J. Schneider

Yuval Levy schrieb:
 Hi all again, and thanks for the feedback about this subject.
 
 First of all thanks to the Mac and Linux users for their display of 
 generosity. Did I miss something, or Windows users were conspicuously 
 absent? with the exception of
So far I was only reading. But not conspicuously.

One thing is that I would happily donate. And if I get a download link 
as thank you, that's OK for me. On the other hand: If I knew I would 
never get a windows build anymore without donation, it would actually be 
buying, not donating. Nevertheless, for me it's worth it.

A completely different thing is that this is an open source project 
driven by volunteers and volunteers do the work that they choose, of 
course. And if there is no volunteer do to the windows work, then it is 
not done. It's as simple as that. I would find it a pity but I could not 
complain.
Then there is commercial software development and why shouldn't somebody 
do this work for money alongside with all other work he does for a 
living? There is nothing wrong about that.

The arguments Rich brought have to be considered by you developers but 
if you choose to loose some windows users as a trade-off for not having 
to do this particularly complicated and unappealing work, that's your 
decision. And a decision I can understand.
(My guess is that in the long run some windows user who is able to do 
that will step in as it happened with PPC.)

regards
Joachim

--~--~-~--~~~---~--~~
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 hugin-ptx@googlegroups.com
To unsubscribe from this group, send email to 
hugin-ptx-unsubscr...@googlegroups.com
For more options, visit this group at http://groups.google.com/group/hugin-ptx
-~--~~~~--~~--~--~---



[hugin-ptx] Re: segmentation fault when stitching

2009-01-12 Thread Felix Hagemann

2009/1/12 Lukáš Jirkovský l.jirkov...@gmail.com:
 Fortunately someone did more testing on the gcc side
 (http://gcc.gnu.org/bugzilla/show_bug.cgi?id=38625) and the cause was
 found. I'm attaching small patch to configure.in against newest cvs
 which adds necessary flag to CXXFLAGS when compiler is GCC.
 Can someone test it? It's still a little bit hackish, but it works for me.

Works for me. A test case that was segfaulting reproducibly when
compiling without debugging symbols now completes. And four to fives
times faster as well, finallz  making pano stitching and blending fun
again!

Thanks for sorting this out,
Felix

--~--~-~--~~~---~--~~
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 hugin-ptx@googlegroups.com
To unsubscribe from this group, send email to 
hugin-ptx-unsubscr...@googlegroups.com
For more options, visit this group at http://groups.google.com/group/hugin-ptx
-~--~~~~--~~--~--~---



[hugin-ptx] Re: Image geometry data into EXIF?

2009-01-12 Thread Daniel M German

 Bruno Postle twisted the bytes to say:

 Bruno On Sat 10-Jan-2009 at 10:09 -0800, Klaus wrote:
  
  As I do see a vast majority of panoramic images not being of the 4\pi
  or 360 degrees type, information like hfov, projection type, position
  of horizon line should allow viewer programmes to properly display the
  hugin/enblend output.
  
  Is there maybe even a standard way of including such info into EXIF?

 Bruno Only for rectilinear images, we could add fake focal length and 
 Bruno sensor size info, but this would be wrong if the image is cropped.

 Bruno PTmender sort-of does what you want, it puts a full copy of the 
 Bruno stitching script in the exif data.  We could modify the hugin 
 Bruno exiftool command to do the same, but currently no tools are able to 
 Bruno make use of it.

In my opinion the entire script is the best solution with respect to
embedding information about how the stitch is done.

Each of the images that is output by PTmender have this information in
an TIFF field (I don't recall which one, I think it is the TIFF
comment field). The TIFF field that indicates the page number and
total pages have the corresponding image number from the script. So
you can easily lookup the i line that generated that image.

All PTtools keep the script in the file. This data is lost in JPegs.

--dmg



--~--~-~--~~~---~--~~
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 hugin-ptx@googlegroups.com
To unsubscribe from this group, send email to 
hugin-ptx-unsubscr...@googlegroups.com
For more options, visit this group at http://groups.google.com/group/hugin-ptx
-~--~~~~--~~--~--~---