[MinGW-dvlpr] Re: [ATT] Packaging: RFC

[Earnie B. <ear...@ya...>]

--- Paul Sokolovsky <pa...@is...> wrote:

I'll break my response into pieces.

> 
> EB> IMO you should have a binary distribution and a source distribution.  For
> EB> ports, the source as modified should be distributed with a differences
> file so
> EB> that modifications can be reviewed and reverted if necessary.
> 
>     IMHO, there should be pristine sources and diff.
> 

If one wants pristine sources they can simply apply a reverse patch.  Very few
will want to do this.  Why complicate the process of many to provide a benefit
to a few?  I'm not even sure that the patch file should be provided at all,
becomes old after multiple patches applied, useless to those without patch
binaries except for historical or documentational purposes.  If we do supply
patch files then we probably should use CVS to control the differences and
always supply the diff file from the 1.1.1.1 version.

But, if we use CVS then we also have to sync up with a newly released version
and that in itself can be nightmareish.  Let's put it up for a vote on this
list.  The question is concerning source distribution and whether it should
contain the original (pristine) source with a diff file for patching it to
build with MinGW or whether it should contain the modified source with no diff
file ready to build with MinGW?

Choose one:

[ ] Original (pristine) source with diff
[ ] Modified source with no diff

----------------------------------------

Tally:

1 for Original (pristine) source with diff
1 for Modified source with no diff

Cheers,


=====
Earnie Boyd
mailto:ear...@ya...

---         <http://earniesystems.safeshopper.com>         ---
--- Cygwin: POSIX on Windows <http://gw32.freeyellow.com/> ---
---   Minimalist GNU for Windows <http://www.mingw.org/>   ---

__________________________________________________
Do You Yahoo!?
Thousands of Stores.  Millions of Products.  All in one Place.
http://shopping.yahoo.com/