[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: debconf, tex-common and the management of ls-R and .cnf files



Norbert Preining <preining@logic.at> wrote:

> The second could be taken as is:
> -------------
> # ls-R handling (taken from old config script, copyright?)
> db_input low tetex-bin/lsr-perms || true
> db_go
>
> db_get tetex-bin/lsr-perms
> if [ x"$RET" = x"true" ]; then
>     db_beginblock
>     db_input low tetex-bin/groupname || true
>     db_input low tetex-bin/groupperm || true
>     db_input low tetex-bin/userperm || true
>     db_endblock
> fi
>
> db_go
> --------------
>
> Maybe we should change the tetex-bin to tex-common, too.

That would require some additional tweaking:  
- Read tetex-bin/item,
- if it was set, set tex-common/item accordingly and fset
  tex-common/item SEEN, 
- unregister tetex-bin/item

> The first part is a bit complicated:
> ----------------------
> # suggest a user to rename old cnf files (copyright Atsuhito?)
> # (upgrade from woody, keep for etch?)
> if [ -d /etc/texmf/texmf.d ]; then
>     if ls /etc/texmf/texmf.d 2> /dev/null | egrep -vq '(\.cnf|\.dpkg-.*|~)$'; then
>         db_input medium tetex-bin/cnf_name || true
>         db_go || true
>     fi
> fi
> ----------------------
>
> Here I am not sure wether we can take it as is.

I see no reason not to take it, if we decide to keep it for etch.  What
kinds of doubt do you have?

Regards, Frank
-- 
Frank Küster
Inst. f. Biochemie der Univ. Zürich
Debian Developer



Reply to: