31 Aug
2011
31 Aug
'11
4:45 p.m.
On 31/08/11 3:32 PM, Dave McGuire wrote:
On 08/31/2011 08:03 AM, Toby Thain wrote:
Understood. That is very sad. I guess they "go through the motions"
so they can claim they've released the spec, but it's not complete
enough to enable someone to truly take advantage of the format.
Is it at least the case that someone can "see the damn image" (which
is usually the end point of digital imaging anyway) even if a lot of the
metadata, layering, etc etc aren't usable?
Unless the "flattened" (composited) image is baked in, then you can see
most of the constituent parts (e.g. image layers, masks) but you can't
see what the whole would actually look like. Which is why my tool is
really aimed at scavenging what can be scavenged
( http://telegraphics.com.au/svn/psdparse/trunk - there's a very fast
command line converter to Gimp XCF in there, too).
As for metadata, some of the structure of it is documented, but detailed
interpretation is another matter. Again -- this speaks to a desire to
hinder competition & interoperability. If the Gimp knew exactly how to
render a PSD, then this would probably eat away at Adobe's extortionate
licensing :)
--Toby
Riiiiight. Like PSD for example, which is documented. Perhaps not
WELL, but it's documented. (as you know intimately from your work on
that parser)
As I also said: It's not documented sufficiently to reconstruct any non
trivial image. Only Photoshop can rasterise a PSD (and even then, it
must be undamaged, and you have to find a sufficiently new version of
Photoshop, etc). Calling what Adobe provides "documentation" isn't fair
to real documentors. :) -Dave