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

Re: Comments on FHS testsuite run

Previously Santiago Vila wrote:
> The empty /etc/opt may well be inside the foreign package using it.

No, since /etc/opt should be used by software that is not part
of the distribution. The distribution is responsible for making
sure the infrastructure for such third party apps exists (ie dirs like
/opt, /usr/local, /etc/opt and /var/opt must exist).

> bug in the test suite. /usr/local/src is part of base2_2.tgz

Well, it's definitely not present on my woody systems so I suggest
the base-files postinst creates it if it's not present.

> > * Reference 4.7-6: /usr/share/nls not present
> >   Reason: bug in Debian
> Again, FHS does not say it must exist.

As I said I also looked at tests that were marked as unsupported, not just
those failed. I'ld like to have the optional stuff present as well; it
certainly won't hurt us.

> > * Reference 4.7-6: /usr/share/tmac not present
> >   Reason: bug in Debian
> Same as before.

Indeed. Same comment from me as well.

> > * Reference 4.7.30-2: /usr/share/misc/ascii missing
> >   Reason: weird standard, possible bug in Debian
> > 
> Same as before. This time FHS even more clear, since it says:

See my previous comment.

> > * Reference 5.1-1: /var/account missing
> >   Reason: bug in Debian
> > 
> In cases like this we let the corresponding package to ship whatever
> directories are required to provide the features.

I'ld like to have base-files create all the directories that the FHS
says should be present. That way we won't break third-party software
that is not shipped in the form of packages.

> If FHS says "foo should be here, bar should be there", then we would
> violate FHS by placing foo there or bar here, but this does not mean
> we have to create empty `here' and `there'.

No, but it doesn't hurt us either. Lets put it the other way around:
third party software will rely on directories that the FHS and LSB
specification lists. Not having those will break things.


 /       Nothing is fool-proof to a sufficiently talented fool     \
| wichert@cistron.nl                  http://www.liacs.nl/~wichert/ |
| 1024D/2FA3BC2D 576E 100B 518D 2F16 36B0  2805 3CB8 9250 2FA3 BC2D |

Reply to: