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

Re: Mass bug filing for the removal of freetype-config and freetype.m4

On Thu, 01 Feb 2018 at 11:07:42 +0000, Hugh McMaster wrote:
> Freetype-config has been considered deprecated for several years [1].

By us, or by upstream?

> With this in mind, I removed freetype-config and built all reverse 
> build-dependencies. I have also searched codesearch.debian.net for use of 
> AC_CHECK_FT2 in configure.ac and configure.in. (Thanks to Simon McVittie 
> for the suggestion.)

Is there a reason to prefer removing AC_CHECK_FT2, rather than patching
it to provide enough of its historical functionality for (I'd guess) 90%
of packages? Something like this should work (untested):

    PKG_CHECK_MODULES([FT2], [freetype2 >= $1], [$2], m4_if([$3], [], [:], [$3]))

(This doesn't do the sanity-checks that current AC_CHECK_FT2 does,
and it respects PKG_CONFIG_PATH instead of --with-ft-prefix,
--with-ft-exec-prefix and FT_CONFIG, but this shouldn't matter most of
the time; and it seems better if simple packages still compile than if
they don't.)

Does Freetype's upstream developer consider AC_CHECK_FT2 to be deprecated

On Fri, 02 Feb 2018 at 11:38:02 +0000, Hugh McMaster wrote:
> On Friday, 2 February 2018 12:15 PM, Paul Wise wrote:
> > Will you be asking upstream to remove it too?
> I hadn't considered that, but it might be a good idea. Why do you ask?

If we ask the upstream developers of various packages to make a change
because otherwise their package won't compile on Debian, some of them
will say "well, that's Debian's fault for removing APIs provided by
Freetype's upstream developer" and do nothing. If we ask them to make a
change because Freetype upstream has officially deprecated the macro/tool
they're using, or because otherwise their package (eventually) won't
compile against newer upstream Freetype releases, it seems more likely
to happen.

Not carrying long-term patches to the build systems of a large number of
packages seems a good goal.


Reply to: