hplip printing fix (partially solved)

Zbigniew Luszpinski zbiggy at o2.pl
Wed Nov 14 22:01:46 CET 2007

Monday 12 of November 2007 22:38:33 Kok, Auke wrote:
> any idea what caused this to break?

fix: run hp-toolbox this way:
LC_ALL=C hp-toolbox
or if you are non US English user:
LC_ALL=$LANG.UTF8 hp-toolbox

the problem exist only when non UTF8 codepages (e.g. ISO-8859-* ) are used.

hp-testpage still is broken so avoid running it (to not block printer queue). 
Nobody needs to print testpage often so it is not so important to fix.

Long description for those who are not bored yet:

This is source of problem (I suppose):
From cups 1.3.4 changelog:
"The scheduler did not reject requests with charsets other than US-ASCII or 
UTF-8, and the CUPS API incorrectly passed the locale charset to the 
scheduler instead of UTF-8"

hp-toolbox fails if LC_ALL=pl_PL is set (Polish locale, usually ISO-8859-2 
codepage, my default locale)

general fix, run from console to set LC_ALL to default English:
LC_ALL=C hp-toolbox

when I still use pl_PL for every localization variable and type:
LANG=C hp-toolbox
then hp-toolbox says there is no printer and red error appears on console:
"error: Invalid locale: C.utf8"

fix for Polish locale (Polish locale remains Polish but default ISO-8859-2 
codepage is changed to UTF8:
LC_ALL=pl_PL.UTF8 hp-toolbox

Now the bad thing:
hp-testpage always fails:
LC_ALL=C hp-testpage
LC_ALL=pl_PL.UTF8 hp-testpage
so LC_ALL fix not works in this case at all.
However hp-testpage is not used very often so is not very important. However 
people may start reporting false bugs when they realize hp-testpage fails.

The problem may appear to all people using ISO-8859-* codepages selected by 
locale. It only exist since cups 1.3.4


More information about the Lunar mailing list