Re: flag day for 64-bit?

[Roland McGrath]

> This doesn't actually break source compatibility, but it does break
> binary compatibility, AFAICT.  

You are either naive or have vast trust in the cleanliness of all user code.

> Sure, but that's what I thought a flag day was. :)

Well, there's little flags and there's big flags.  The plan I described
would have the (small) set of consequences I described.  Your plan would
require recompiling everything everywhere yet again.  Jeff might kill you.

> Seriously, I think we should make the end-state as pretty as possible.

I don't think there is anything really wrong with the LFS interface.
People expect it.  I suppose we could one day change the default to
_FILE_OFFSET_BITS=64.

> The "correct" end state is one which might not follow the broken Linux
> definitions.  Still, compatibility (in that same end state) is also
> important, and we'd like one day to have generic programs for the Hurd
> and for Linux differ only in the library they link against.

To get that we need to stick with the current LFS interface plan.  That is,
code compiled with -D_FILE_OFFSET_BITS=64 will produce references to foo64
when the source says foo.