No Frozen Rawhide

[Ben Boeckel]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Paul W. Frields wrote:

> On Wed, Jul 22, 2009 at 06:24:40PM -0400, Bill McGonigle 
wrote:
>> On 07/22/2009 12:39 PM, Bruno Wolff III wrote:
>> > I think the confusion for normal users will be minimal, 
because normal users
>> > won't be looking at the raw repos in any case.
>> 
>> There's a class of users between active developers/QA folk 
and people
>> who only use GUI package managers who are apt to go to a 
download site
>> looking for RPM's.
>> 
>> That said, I think a succinct README ("This is pre-release 
code, there
>> be goblins here") in the right directory that Apache auto-
displays for
>> index listings would be sufficient to warn off those users.  
Otherwise,
>> _somebody_ is going to say, "oh, cool, 12 is out, I missed 
that."
> 
> This would be hard to enforce with mirrors, some of which may 
not use
> those automatic features.  Any user trying to install from 
Rawhide,
> though, receives the famous betanag up front, which basically 
tells
> them "I'm a pre-release, are you sure you want to install me?"
> 

Should we also have warnings in preupgrade? Or is this not how 
it works? I haven't used it myself yet.

Also, should yum repo configs grow a "eatsbabies=1" (or 
warnonenable to be more descriptive, but I'm partial to the 
eating babies or a "here-there-be-dragons" option) for rawhide 
to have yum confirm when using --enablerepo=rawhide on a non-
rawhide install? This may be good anyways since cherry picking 
from rawhide has other problems (such as xz payload this time, 
libssl/gcc/hash upgrades before) that may drag in the world as 
dependencies.

- --Ben
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (GNU/Linux)

iEYEARECAAYFAkpnsiUACgkQiPi+MRHG3qRfIgCfVPhT45wJ4hXdbsciLUl7w8tp
VX0AoJFxiaOAZLfQqlsSF/HeZ+yyH5YM
=N52C
-----END PGP SIGNATURE-----