[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: pdftk -

Frank Küster wrote:
Rob Munsch <rmunsch@solutionsforprogress.com> wrote:

No, I'm saying that from this changelog entry:

pdftk (1.12-11) unstable; urgency=low

  * Mixing different versions of g++ and gcj does not work, explicitely
    use gcj-4.1 and g++-4.1.

I read that there were problems with different versions of these
packages being installed, and used during build, and that the explicit
dependency was added to avoid that.  I think that it should also work if
both are -4.0, or 3.3.

Ok, sorry i misunderstood that.  Interesting...

it still FTBFS in sarge.  I haven't tried to build in a sarge-backports
environment, though.
Weird - it built for me, and AFAIK no one's had problems with it yet.
The only oddity was this warning:

eport.o(.text+0x1888): In function
`ReplaceXmp(com::lowagie::text::pdf::PdfReader*, std::basic_string<char,
std::char_traits<char>, std::allocator<char> >)':
: warning: the use of `tmpnam' is dangerous, better use `mkstemp'

but it built just fine and seems to run.

I assume you've tried a different patch than I have?

Not using any patch - as per the author's site, i edited the Debian.Makefile by hand to change the version suffix and off i went (after pulling in libgcj5-dev and the 30 dependencies that triggered automagically).

Both g++ and gcj are 3.4.4 here, the current plain-sarge releases.

I guess at this rate, both g++/gcj and pdftk will be out of testing together long before i would figure out how to make a proper backport package. :)

Rob Munsch
Solutions For Progress IT

Reply to: