On 25.12.10 Hilmar Preusse (hille42@web.de) wrote: > On 25.12.10 Norbert Preining (preining@logic.at) wrote: Hi, > > It is not there, it is in tex-common: I adapted the postinst with > > loads of debug and what did I found: > > > Well, I expected somthing like that. > > > Sooooooooooooo ... what is the solution? The simple idea would be > > to call updmap-sys --nohash, but on initial installations that > > might lead to problems as the psfonts.map etc file generated by > > updmap might not be present *before* the run. > > > Hmm, there is only one mktexlsr/texhash call in /usr/bin/updmap. > Hence we could simply run "updmap-sys --nohash" and an additional > "mktexlsr ...." after it. > I didn't check yet if the ls-R update is needed during the run. > However I didn't check if an updated ls-R database is needed during > the updmap-sys run. In this case we can only patch updmap to update > only our dist texmf trees...and file an wishlist bug to upstream that > the texmf trees to be updated during updmap can be specified. > Of course this is non-sense. The normal system admin would expect that updmap-sys refreshes the ls-R db as needed which would not be the case after the change. H. -- sigmentation fault
Attachment:
signature.asc
Description: Digital signature