Hi Thomas!

Thanks for your explanations.

On 2 Apr., 16:53, "T. Modes" <thomas.mo...@gmx.de> wrote:

> No, that was not the indention. We want to keep the make file approach
> for stitching. The make file is created in
> PanoramaMakefileExport::createMakefile. But this method mixes
> stitching logic with make file syntax.

Ok, but may I ask whats the main reason for using make here?
It obviously causes trouble with escaping and quoting. Is any special
make feature used,
like dependencies, and changed files detection ..?
Some rules depend on each other, in the generated makefile. But is it
important to rebuild
only files with changed dependencies? Or is it more like a static
script invoking one tool after
the other? At least the output of a make run looks like this, but i
haven't looked at every line
in detail.
For example if you use make to run cp detectors, the generated data is
part of the pto file,
afaik. In that case, make has nothing to use as a generated file for
that rule and check it's
timestamp.
But of course make is a well established tool, and it allows easy
batch processing once the
makefile is generated.

So the parts of the project idea would be:
- develop/port an object oriented abstraction of all the items
existing in a makefile,
  similar to bruno's pearl scripts. With the goal to have a
bulletproof makefile generation
  lib, and never have to fight with quoting/escaping and other flaws
again.
- transfer cp detection to using make, which makes it "batch-able".
- use cp detectors and image metadata to autodetect complex panoramas
  (or is that already somewhere?)
- Extend PTBatcherGUI do handle those new batch jobs. (I think it
already queues makefiles,
  doesn't it for batch stitching?)

Am I right?

> I tested also your second exercise. It works good. But some comment:
> your approach will fail if the project filename or image filename
> contains special characters (like # or : and other). Therefor the
> filenames need to be escaped or quoted. This could be easily done
> using some variables and in the target only using this variables.

mhm, I think in my case it's not that critical, because the filenames
are only arguments to
echo, not used directly in the makefile. So a # wouldn't trigger a
comment. But a \ would be
critical. And there are always surprises with quoting, according to my
experience :)..

> Now concentrate better on your proposal. I hope I have give you an
> overview with my above descriptions.

Yes, that should be done soon. I'm not sure, what the proposal exactly
should look like.
I'm working on a draft.
I'll discuss it with you.

Flo

-- 
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

To unsubscribe, reply using "remove me" as the subject.

Reply via email to