[pdftex] pdfcrypt
Ricardo Sanchez Carmenes
carmenes at bioinf.medicina.uniovi.es
Fri Oct 4 10:49:46 CEST 2002
On Wed, 2 Oct 2002, Thierry Bouche wrote:
> Le mercredi 2 octobre 2002 à 11:12:53, Martin Schroeder écrivit :
>
> MS> On 2002-10-02 10:39:21 +0200, Thierry Bouche wrote:
> >> Next problem: gs' pdfopt works on the unencrypted version, not on the
> >> encrypted one...
>
> MS> Of course. You want your encrypted files to be safe from changes,
> MS> don't you? And pdfopt surely changes the file...
>
> sure, but I noticed that gs doesn't obey so much security settings (you
> can extract text when copy is forbidden, e.g.).
Acrobat specs state someware that it is the responsibility of reader
implementors to make sure their software obeys those settings, because
once the text is available in any way, it can be used for anything if
the reader software allows it, as it is the case with GS.
I already stated some time ago that *standard* PDF encryption cannot be
considered a strong protection at all. It has its uses, however, that's
why I added it to pdftex, and I have received several mails of people
concerned by the removal. They asked if I could mantain a paralell
pdftex version with the pdfcrypt facility. My whish would be to be able
to mantain a patch to apply to the current pdftex version, but I am
afraid that I don't have much time right now.
> MS> Best regards
> MS> Martin
>
> MS> P.S.: From the NEWS of the next pdfTeX version:
> MS> - The extensions for pdf encryption have been removed,
> MS> since they make the pdfTeX code overly complex. Those
> MS> needing pdf encryption are encouraged to produce a
> MS> standalone program for encrypting pdfs.
>
> Well, fine for me, as I checked that the encryption mode acts OK as a
> postprocessor, but doesn't really interact properly with pdftex working.
Well, to be fair, the problems I have found so far have not that much
to do with pdftex itself (it integrates very well into pdftex, because
the only necessary thing is to encrypt the streams just before they are
output), but with other postprocessing. Usually, the tools that work on
the resulting pdftex file (e.g. pdfthumb) expect (or require) a
non-encrypted file. In this case, encryption shouldn't be activated until
the very last run. This can be easily automated, though.
> What I (we) really need is some batch postprocessor that would add
> protection and optimization to the PDF files from pdftex.
Of course the postprocessor approach is a good one, especially for linux
users.
> Apart from using acrobat's batch mode (which implies windows) there seem
> to exist no way to do that (preferably under linux). Wouldn't it be
Yes, there is at least one, but it is not freeware, although you can
get a free fully-functional (I think) linux evaluation copy at
<http://www.sanface.com/archive/pdfcryptl.tar.gz>.
I found it by chance after developping the pdftex-pdfcrypt addition, and
in spite of the coincidence in the name, it has nothing to do with my work.
> quite feasible starting from the existing work by R.S. Carmenes?
I guess so, at least for the encryption stuff. I don't know about
optimization. As I said above, I don't have the time myself, but if there
is any volunteer, I would happily assist him or her (if [s]he needs so).
>
> --
> Thierry Bouche
More information about the pdftex
mailing list