> I've searched in the parser and viewer doc and in the list archive
> (http://www.mail-archive.com/plucker-list%40rubberchicken.org/) but I
> didn't find how I can manage to convert images in 65000 colors.
Until very recently, it didn't. You have to grab a snapshot of the
code from the cvs and build that (building the viewer is not necessary in
this case), and also grab the latest cvs versions of netpbm from Sourceforge
(Bryan's version). Bill Janssen has patched netpbm and PIL support into the
parsers to handle this. The viewer itself handles this natively, AFAIK.
> Here are some lines I'm using in plucker.ini:
> bpp=8
You probably want 16bpp, not 8. 8bpp == 256 colors.
As an aside note, I'm working with some options to get >256 shades
of greyscale in 8bpp mode. Some images look good in 4bpp greyscale, but
because of contrast and coloring, look *HORRIBLE* in 256 colors in 8bpp, so
getting even 256 shades of greyscale on the same image would make it much
more "normal-looking" on the device. A little ppmquant trickery was needed
to test this out, I've almost got a solution. When I figure it out, I may
suggest a 5bpp option (which isn't *REALLY* 5bpp, but it would be enough to
trick the parser into knowing that this was a >4bpp greyscale conversion).
> ...it converts images to 256 colors but not to 65000. If I set bpp value
> as "16", images are not parsed. The same thing happens if I try
> "-color=C65000" or "-color=C65K".
> What can I do?
That implies that your parser doesn't properly grok those options.
I've never seen the -color=C65000 or -color=C65K options. Are these
something in the Windows parser only? Dirk?
> Another problem is when I try to convert a pdf document with the
> converter included in Plucker main web page (direct link:
> http://plucker.gnu-designs.com/pdf2plucker/pdf2pl.pl).
I have purposely blocked the direct link, to stop that exact type of
url-sharing. You can only access it by clicking on it from the main page of
the website. Thanks to IE not sending proper headers in the request, I had
to hack a bit of IE-workaround code into the page.
> I tried with several pdf documents, but I always get Internet Explorer
> "File download" box: "you choosed to download a file from this place :
> pdf2pl.pl from plucker.gnu-designs.com / What do you want to do with
> this file? Open it/Save it on your disk". If I clic on "open" and "ok",
> I get a second messsage which is nearly the same but the file is then
> "pdf2pl"... If I download it that's not what I want.
You might want to impress upon Microsoft begin to properly support
MIME types. Apparently they don't. Here's an example of how I present the
file to the user, which works perfectly with the snapshot link on the
website.
sub build_pdf {
print "Content-Type: application/octet-stream\n", "Content-Disposition:
attachment; filename=$thingy.tar.gz\n\n";
[...]
[other irrelevant stuff deleted]
}
If IE can't support that content type, it's a Microsoft problem, and
probably should be reported to them as a bug. This is the accepted standard
way to do this type of "presented-to-user" file download.
You may also want to read up on the following articles:
http://support.microsoft.com/support/kb/articles/Q267/9/91.asp
http://support.microsoft.com/support/kb/articles/Q281/1/19.asp
http://support.microsoft.com/support/kb/articles/Q279/6/67.ASP
I just tested this with IE6, and the bug is still there. If you
rename the file to 'foo.tar.gz' or equivalent in your SaveAs dialog, you
will be able to open it with Winzip and get the contents out. The patch
Microsoft provides for IE5 may solve your problem, but it returned in IE6.
> How is it supposed to work exactly? Does the program exists in a
> downloadable version (even in Perl Script language)? (I have a lot of
> Pdf files to convert and read in Plucker) The Palm version of acrobat
> reader is very slow and the files are too big for 8 Mb... Plucker is
> really better.
> BTW, thanks again to the developpers for their work and support.
Thank you for your support (and bug reports) as well.
Incidentally, while we're on the topic of IE, I've now made a new
favicon.ico for the site (as well as for the other sites on that network),
so if you have IE or Konqueror, bookmark the site, and let me know what you
think (if you've already bookmarked it, the new icon won't show up until you
delete your IE cache and History in C:\Program Files\Profiles\%USER\History)
I also managed to fix a font-rendering bug in the way IE loads and
renders the page, which would show the fonts in different parts of the page
in a different font face. Now they're all consistant.
Keep the bug reports flowing. This weekend, I'll be putting up live,
the bug-reporting engine on the site, and adding a few other fixes here and
there to make the site faster and more iteractive.
Requests are always welcome.
/d
