[KLUG Members] rpm trouble

Adam Tauno Williams awilliam at whitemice.org
Mon Nov 15 21:22:38 EST 2004


> I can't print to my Brother HL720 (a Windows printer, grr) in RH8; 

Sorry to be the bearer of bad news, but the RPM database in RH8 is well
known to be really unstable and die into seg fault mania at the
slightest flick of the pony's tail.  But we are talking about VERY old
software.

> I've downloaded newer versions of ghostscript, like 7.05 and 8.14, and 
> tried 'rpm -Uvh' on them.  I get
> +++++++++++++++++++++++++++++++
> error: Failed dependencies:
>          libijs.so is needed by (installed) gimp-print-4.2.1-5
> +++++++++++++++++++++++++++++++

Have you tried installing the ijs libraries?

> Fine, I figure; I never print from gimp, 

It doesn't work that way.  gimp-print is a library used by lots of
things related to printing,  it provides functionality relating to color
matching, dithering, and various other rasterizer type of things.

> I'll just erase that.  (And 
> grr, stupid rpm's and their dependencies.)  So I try 'rpm -e 
> gimp-print', and get
> +++++++++++++++++++++++++++++++
> error: Failed dependencies:
>          libgimpprint.so.1 is needed by (installed) ghostscript-7.05-20
>          gimp-print = 4.2.1 is needed by (installed) 
> gimp-print-utils-4.2.1-5
> +++++++++++++++++++++++++++++++

Right, you can't have ghostscript without libgimpprint.

> Grr.  So I try erasing gimp-print-utils and all I get a segmentation 
> fault, every time.  

Sounds very much like RH8,  your only chance is to run the
removes/adds/updates with scripts disabled; and you won't like the
results of that.

> (Though no core is dumped--why is that?)

Dumping to core only occurs when requested.  Saves us from the days when
systems were littered with often very large core files.

>   So, 
> frustrated, I try a '--nodeps' option to install the newer ghostscript. 
>   It tries, and prepares, and shows hash marks and everything, but then 
> gives a seg fault.  Removing gimp-print with 'rpm -e --nodeps' results 
> simply in a seg fault with no preparation.

Does rpm --rebuilddb succeed?



More information about the Members mailing list