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

Re: RFS: avant-window-navigator 0.2.1



On 03/02/2008, Neil Williams wrote:
> What I meant was if dependencyA appears in the dpkg-shlibdeps list
> of "not needed" linkages for 'foo' but foo is nearly always
> installed alongside bar which uses symbols from dependencyA (i.e.
> dpkg-shlibdeps doesn't complain about dependencyA in the build of
> bar), then foo probably doesn't need to be altered. It's no
> absolutely necessary for foo to Depend on dependencyA but it doesn't
> hurt in the majority of cases.

I now see your point, but I still disagree. Say foo depends on bar,
bar depends on baz, and there's an extra link (hence Depends) between
foo and baz. baz gets a bumped SONAME, bar gets rebuilt (through
binNMUs). Fine. Err, no. foo has to be rebuilt as well. Too bad.

I'm not only speaking about testing migration here. That means extra
libraries installed whereas the old version on baz could have gone
away in a single round. There's no use having additional unneeded
dependencies, even if they appear to be legit.

> there are also cases where dpkg-shlibdeps reports an extra linkage
> where the library concerned is already packaged alongside a library
> that does need to be linked against the package - libdl is the most
> common.

I agree that there are some special cases, like libdl, pthreads, and
so on. But ISTR that whole pages were reported by dpkg-shlibdeps, not
only a few lines.

> Well, it's not something I am doing for my own packages, yet, so I
> don't want to push it onto those maintainers whom I sponsor when I
> am not comfortable getting it to work on my own packages.

I understand that, of course, but I wouldn't jump to conclusions (I'm
not sure we should…) anyway.

> I'm experimenting with a few things upstream where the upstream
> ./configure asks for the pkg-config data but then either does some
> parsing of it or substitutes a (shorter) replacement but pkg-config
> exists for a good reason and (IMHO) -Wl,as-needed isn't a complete
> solution to the problem (with or without zdefs).

I didn't say (or at least not in these words) it was a silver bullet,
but rather a workaround in the cases where simple patches against the
build systems seem unreasonsable (e.g. qmake?), and as a learning
experience.

> the complication is that tinkering with pkg-config data like this
> can trip up porters and cross builders and as my main workload is
> cross building, I'm not about to break my own work. :-)

I know what porting means, look in the BTS for kfreebsd-* bugs.

> Now I'm all in favour of reducing spurious dependencies - as anyone
> would expect whilst working on making Debian small enough for
> embedded devices - but not at the cost of making more work for
> myself when cross building the same package. I just want to be sure
> the changes actually work before recommending them to others. IMHO,
> -Wl,--as-needed -Wl,zdefs is not at that stage, yet.

In the thread (on -devel) where the use of those options were
discussed, I don't remind of people reporting broken packages due to
that. But if you could find examples of broken packages because of
that, I'm very interested (in having a look and eventually fixing them
if I can do that).

Cheers,

-- 
Cyril Brulebois

Attachment: pgp_FQNYyd38k.pgp
Description: PGP signature


Reply to: