[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Summary[2]: dpkg and alpha/beta versioning



Jason Gunthorpe writes:
 > 
 > On 7 Jul 1998, Adam P. Harris wrote:
 > 
 > > 
 > > Yann, following your summary of alpha/beta versioning, it seems to me
 > > that a best practiced has arised and reached some consensus.  I wasn't
 > > following too closely, but I thought I saw this happening.  Did it?

It *seems* that most discussions now turn around the `~' solution, and
real comparison with the "sub-epoch" scheme have been done - at least
not in the list...

 > > If so, could you just take the best practice, and, if it's something
 > > requiring dpkg changes, raise discussion in dpkg-dev@lists.debian.org.
 > > If the best practices (hope hope) does *not* require dpkg changes, how
 > > about submitting a bug against the pacakging manual to document the
 > > best practice?
 > > 
 > > Then I could feel like we've accomplished something, instead of just
 > > being a debating society.

Yes.

 > 
 > Here is a new version compare function that takes into account the ~
 > feature. My tests were:
 > 
 > # ver1 ver2 result
 > # -ve = < +ve = >
 > 1.1~pre 1.1 -1
 > 1.1~1pre 1.1~pre -1
 > 1.1~1pre 1.1 -1
 > 1.1~2 1.1~3 1
 > 1.foo~boo 1.fooboo -1
 > 1.foo~1boo 1.foo~boo -1 
[...]
 > The exact syntax it expects is ~xx where xx is an optional digit string. ~
 > alone acts like a single character and compares less than any other
 > character including eos. ~xx acts like a negative digit and compares as
 > you would expect to other digits, ~xx also compares less than eos.

Hm, it seems you give a "negative meaning" to the `~' char, as Giuliano P
Procida suggested, ie:

	1.1~2 << 1.1~1 << 1.1

However, I still fail to find a use for it as such: if the "1.1~1*"
series are pre-versions of "1.1*", then the "1.1~2*" are pre-versions
of the pre-versions ?  That does not make much sense to me...  OTOH,
maybe you want to use this for updates to stable ?  Even then I'm not
sure we need that, as we have a know number of dists to update, and we
can write: 1.1-9~1bo << 1.1-9~2hamm << 1.1-9 - I don't think we are
still updating rex, are we ?

However, I think having the following could prove useful, and may be
more intuitive:

	1.1~1 << 1.1~2 << 1.1

My use for the latter would be to cope with the (hypothetical)
ordering I already mentionned:

	1.1~pre << 1.1~1beta << 1.1

Am I missing something here ?
-- 
Yann Dirson    <ydirson@mygale.org> | Stop making M$-Bill richer & richer,
isp-email:   <ydirson@a2points.com> |     support Debian GNU/Linux:
debian-email:   <dirson@debian.org> |         more powerful, more stable !
http://www.mygale.org/~ydirson/     | Check <http://www.debian.org/>


--  
To UNSUBSCRIBE, email to debian-policy-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org


Reply to: