[GTALUG] Fedora 22 Live Workstation Install - no fglrx - no pdftk

[Russell Reiter]

I recently acquired another older MB, ASUStek M3A78-EM, Athlon 64 X2
Dual-Core, 4Gb ram & onboard Radeon and I thought I'd give the latest
Fedora Live a try. It installed well enough and Gnome pretty much
worked except for the lack of fglrx in the open source drivers. In
fact KDE Plasma doesn't work at all.

In the Gnome desktop however, any intensive video apps such as VLC etc
overheat the CPU and crash the unit. It looks like I'd have to
backport X in order to install the proprietary Catalyst drivers for
this particular GPU. HTML 5 videos on the net do work, but eventually
the unit overheats and dies. I switched to lightDM and that's a little
better but still not up to speed.

It seems that yum has been dandified for this release but when I tried
the new dnf update it failed to connect to the repositories. However,
yum-deprecated did update the kernel and after that update dnf was
able to function as intended.

As I am always a sink or swim kind of guy and because a few of the
individuals I support have switched to Fedora, even though I usually
install Debian, I decided to try this as my workhorse for a while. I
am looking around for a more recent PCIe video card in my price range,
which usually hovers around free or the price of a hamburger, but
failing that I'm pleased overall with the performance of the business
end of things.

Well there is one exception, no more pdftk support. (Licensing issues
with iText5+)

My problem.
I often have to combine several complete and already numbered pdf
documents into one consecutively numbered tabbed document. I use a
shell script like this one below to number the combined documents.

I include this in case others on the list might find it handy. I
didn't author it, I found it on the net somewhere and would credit the
original author, if I could find my notes this morning.

enscript -L1 -F --header-font=Times-Roman10 -b '||$%' -o- < <(for i in
$(seq "$(pdftk "$1" dump_data | grep "Num" | cut -d":" -f2)"); do
echo; done) | ps2pdf - | pdftk "$1" multistamp - output
"${1%.pdf}-header.pdf"

This adds a header with the page number in 10pt text to the upper
right corner of each page. Then I use pdftk to burst the document
again and then blend in the tabs like this

pdftk TabA.pdf file1.pdf file2.pdf TabB.pdf file3.pdf file4.pdf cat
output newfile.pdf

after each page is separated and numbered. It's a bit kludgey on
larger docs but for small quick work it's ok. I was hoping to automate
it more eventually.

So if anyone has a recommendation on a more linux friendly video card
of somewhat older vintage, or a command line tool to replace pdftk I'd
appreciate it.

I note that there are several GUI tools now available to assist with
sorting pdf pages but I was really hoping to find something in the
command line arena which I could then poke into this script.

Thanks
Russell