[Kde-print-devel] [Bug 140006] Create standard procedures to test KDEPrint (and all components it relies on) for regressions, and KDE applications' interaction with KDEPrint

[Kurt Pfeifle]

------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.
         
http://bugs.kde.org/show_bug.cgi?id=140006         




------- Additional Comments From pfeifle kde org  2007-01-14 18:23 -------
OK, so now I have my first additional CUPS instance running here.    :-)

I've implemented its creation by adding an "include Makefile.cupsclone" into the main Makefile. So you can compile CUPS as you would normally do.

But you have more options now: "make [tab] [tab]" now shows my additional build targets. You can run "make cups_clone", "make_printers_clone" and "make cups_clone_start". Heh... And it works pretty well once it is up.

make cups_clone copies all required files into a separate subdir-tree in the user's home, and symlinks for the rest which point to the build directory. So every time the build directory is updated, the running clone also has new bleeding edge binaries at its disposal (if the cupsd binary itself got rebuilt, it will of course only run if the old one is stopped or killed and the new one started up). If you wan to de-couple the new instance from the build directory (to safeguard yourself against b0rken updates), just copy the whole clone-tree to a third place, and use the --dereference ("follow symlinks") switch for the cp command.

The more I play with it, the more I'm convinced it could be a really useful tool for everyone who is interested to look a bit deeper into all the un-loved printing stuff (not just limited to KDEPrint).

It surely is not yet release-able for now. 

 1. The Bash code is horrible and uncommented. "Horrible" will probably never
    change, but some comments I can add for sure. 
 2. And it also needs at least a README which outlines the most important steps
    and procedures to follow and the envrionment to set up when someone
    seriously wants to test applications against a CUPS server that runs on a
    non-standard port such as 23631. 
 3. And last, I need to test the hack at least once on a virgin system. You 
    *always* find issues when you "install" your own software for the first
    time on a completely different system.

I cobbled this together in half a day. But of course, the initial "make cupsclone" had a few things missing which I only discovered when trying to run the new instance from its own clone tree. So I copied them over manually, and updated the Makefile accordingly -- but I may have missed stuff.

So this required cleansing and testing will still take some time. I also need to put it into some version control system (at least locally, for me), so I have it "organized" every time I can work on it. (I doubt there would be many people wanting to contribute on this particular thingie, but we'll see).