i should be working (but i'm fixing my computer).

So, as I left off yesterday, gscan2pdf would not work and was borked.  I installed v. 0.9.30 from source and now it works.  Yay!

...

[update note:  there are two hints I should add to this post.  Lately, I have been pretty successful with getting gscan2pdf to work if (a) I add the devel>languages>perl repository and (2) I download the src package from sourceforge and run an rpmbuild --rebuild on it, then install the new rpm. YMMV.]

This is on OpSu 11.2 i586.  And, some hints for anyone else trying to get this working, well -- first I should mention I had o.9.29 already installed from the Build Service, and this may be significant because I may have installed a bunch of necessary Perl dependencies which would not have been there had I started from scratch with .30.  I'm no good at explaining things (mostly because I forget what I did), but here are some clues:

Ver. .29 did not install via "1-click" from the build service for me.  There are/were two versions -- I don't think the build service finder locates them now, but they seem to still be in the repositories, so find the "shapiro" and "illusio" repositories.  Why?  Well, I couldn't get all the dependencies from either (or both) installs, but on version installs some of the dependencies and the other version installs some other dependencies.  Which gets you 90% there.  I think what has happened is that these packagers must have had some of the dependencies already installed on their computers and so were not aware they needed to include them in the package.

But, that did not entirely work.  There remained one or two dependencies I could not locate -- which as I recall, I determined what I needed by error messages upon trying to run gscan2pdf.  So, what I did was go straight to the illusio repository and search through the folders/directories, whereupon I actually found the missing perl modules.  (Thanks illusio and shapiro!)

I will also note that I forgot to remove the .29 version before installing .30.  Works anyway, though.

Now, if you are starting with ver. 29 (and that's the only way I know the story at the moment), the sourceforge noarch rpm for gscan2pdf for SUSE won't actually work.  It can't find dependencies for Perl(Goo::Canvas) and Perl(Proc::Kilfam).  So, we now have a mess on our hands.

I did not have this computer set up to install from sources, so I had to install all the base and perl "development" packages.  I then ran CP(from console, of course). CPAN is a pretty good tool for finding and downloading CPAN modules.  A command like "i  /Goo::Canvas/" will locate the module.

Only problem is, we need goocanvas, as the perl module is just an interface.  there is a 1-click goocanvas at OpSu Build Service, but it doesn't work.

So ... now we have to install from sources, so we download the source file, put it in our /usr/src/sources or whevever and run ./configure, then ./make and then ./make install from a console in the directory where we have placed the source.  (Oh, you may have to extract it first.)  Not being a command-line maven, I open a "kdesu konqueror" window (dolphin doesn't usually work well for me), browse to the sources, do the extract, then use "tools" to open a console in the correct directory.

Okay, so we've got "goocanvas" if all goes well.  Back to CPAN and try and "install Goo::Canvas".  Did not work for me, I don't know why.  Claimed it could not find /usr/bin/perl which was, um, not true.  I mean, it was there.  So, now back to konqueror, over to /root/.cpan/build/[whatever is the goo::canvas directory.]  Now, open console, run "Makefile.PL" or "/usr/bin/perl Makefile.PL" (we are assuming here that you do in fact have Perl installed!), then ./make then ./make install.

Now, your choice, you can use CPAN to install Proc::Kilfam (I think I have that right, but gscan2pdf will tell you what is missing if this isn't right).

Next, we are back to our /usr/src/[etc] to install gscan2pdf.  We've downloaded from sourceforge the v.30 source rpm for SUSE, which when "installed" will place a tar.bz  package in our SOURCES directory.  Extract, open console, Makefile.PL, ./make, ./make install (I think these are the commands) and we've now got gscan2pdf.  Run it from the console, because we'd like to see the feedback if something is still missing.

With any luck, we are now good to go and can run gscan2pdf from a run "box" (<alt><f2&gt or from a menu icon.

If, like me, you need a pretty decent scanning interface and aren't too fond of xsane, hope this helps.  I would package it -- I use gscan2pdf on a number of computers, so it would help ME out to package it -- but I don't have a clue how to build rpms or use the build service.  Maybe something to consider someday in my idle time.

[yeah, right.]

Thank you for visiting and have a pleasant tomorrow.