Re: File names (and synchronization)

[Peter Simons]

Bastiaan Veelo writes:

 > I see two reasons why we may want to rename files now:

 > 1) complyance with policy point 3
 > (http://www.gnu.org/software/ac-archive/policy.html)

You realize that just renaming the file won't suffice,
right? Someone would have to edit the macros too. I wouldn't
want to keep a macro called AC_FOO in a file called
ax_foo.m4.


 > 2) if the released macros are going to be generated files
 > [...] then it would be an advantage if the generated file
 > and its original have different names. The original of
 > "macro.m4" may be called "macro.src.m4", or something.

 > There is a downside on 2) though, until the two archives
 > are unified. As Guido's archive releases the pristine
 > sources as they are on Savannah, he may have to build
 > renaming into his synchronising procedures. Does this
 > count as an argument to only distribute pristine sources
 > and no generated macros?

I think it counts as an argument against having a different
name for the archive entries than for the generated m4
release files. They are in a separate repository in a
separate directory; that should distinguish them
sufficiently enough, IMHO.

About synchronization: Wouldn't it be beneficial if Guido's
archive would generate its presentation from the _same_ CVS
repository as gnu.org does? The problem I see right now is
that there are still numerous macros in the sf.net
repository that are _not_ on gnu.org, and I don't see any
technical reason why that would be. Guido, what do you
think? Could we agree to _both_ use the repository on
Savannah? And to commit new entries there and only there?

Peter