> Sure, it will take more time and effort, but this is something that
> each upstream author should really do, not something we should do in
> their name to hide a problem which is really an upstream problem.

In an ideal world, these would all disappear upstream.

But the sheer scale of small and tedious fixes required -- not counting
the energy required liasing and convincing upstreams who may not even
care or exist anymore -- make this really untenable IMHO if we want to
make progress as a distribution in the short to mid-term.

I understand that this line between local and upstream patches is
somewhat arbitrary, but these particular macros definitely are on the
"longer-term" fix list for me.

We would, of course, always push for upstreams not to use __DATE__ or
__TIME__ and I don't think this is really condoning it. Actually that's
a good point - the documentation for this feature in GCC should mention
that you probably shouldn't be using this anyway.


     : :'  :     Chris Lamb
     `. `'`      la...@debian.org / chris-lamb.co.uk

Reproducible-builds mailing list

Reply via email to