On Wed, 9 Apr 2008, Martin Costabel wrote: > Viv Kendon wrote: > [] >> Note: I have xorg installed on this machine: >> i system-pkgconfi 1.4.0.90-1 [virtual pkgconfig package representing >> x... >> *i* xorg 6.8.2-1044 Free X11 implementation >> *i* xorg-shlibs 6.8.2-1044 Free X11 implementation >> >> but no info files for it (because it was installed before they disappeared >> from the 10.5 version of the tree I think). >> I also have the Leopard X11, probably 2.1.4, with the X11R6 simlink removed >> so they can coexist. > > If there is a safe recipe for disaster, then this is it.
Indeed...which is why I've been compiling on a 10.4 PPC that doesn't have hybrid X11 stuff and installing the resulting debs. That machine is about to be upgraded to 10.5 in a way that still gives me a clean build environment (I hope). What I'm trying to do with my runtime environment has already worked on one PPC machine without any errors like this...I didn't embark on this route lightly, or without testing everything I use regularly. To tell the truth, I didn't expect it to work as well as it does, but very few executables I run have problems when asked to work with the "other" xserver (I guess because they were compiled in clean environments and know where their correct libs live). But I think I kept the other machine updated more frequently as the 10.5 tree developed, and thus avoided this problem of several interdependent confused packages. Or maybe not, since fink list -t -o doesn't show *i* packages, I may just not have noticed... > As for the fontconfig problem: The "*i*" is there because you have a package > installed that doesn't exist (in this version or at all) for 10.5. I would > remove the fontconfig-config, perhaps using dpkg Yes...but the outdated fontconfigs are all interdependent on each other, and on other packages that don't exist in 10.5, and it won't let me purge them. I think it's time to compare with what I have on the machine that works, rather than looking for a quick fix...I can always swap in the entire fink tree from the machine that works, if fiddling around with forced purges gets too messy. many thanks, -- Viv ------------------------------------------------ Dr Viv Kendon http://quantum.leeds.ac.uk/~viv Quantum Information Physics & Astronomy University of Leeds ------------------------------------------------------------------------- This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone _______________________________________________ Fink-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/fink-users
