Mailinglist Archive: opensuse-project (539 mails)

< Previous Next >
[opensuse-project] Re: First Survey on openSUSE Version naming is open now
  • From: Jim Henderson <hendersj@xxxxxxxxx>
  • Date: Tue, 15 Mar 2011 20:51:24 +0000 (UTC)
  • Message-id: <ilojgc$n54$2@dough.gmane.org>
On Tue, 15 Mar 2011 20:37:23 +0100, Per Jessen wrote:

Jim Henderson wrote:

On Tue, 15 Mar 2011 20:02:13 +0100, Per Jessen wrote:

Jim Henderson wrote:

On Tue, 15 Mar 2011 11:29:34 -0400, Chuck Payne wrote:

Me too, though I wish we just keep numbering, it is a part of name,
like me change my last name from Payne to Smith.

Same here, I don't see a compelling reason to change, maybe just a
compelling reason to define what constitutes a major version number
increase.

One compelling reason to change is that our current numbering scheme
doesn't match what we do - i.e. we have no major.minor release scheme.

One might argue that we have a scheme, we don't have something that
differentiates between major and minor releases. ie, we have a scheme
(major.minor), but it lacks definitions of what constitute a major or
minor release.

But that might just be semantics. :-)

Important nonetheless.

Andreas started out by saying it quite clearly "openSUSE does not have a
major and minor numbering, even if it seems so. There is right now no
difference in any way between what we would do for openSUSE 11.4 or
12.0".

Fair point.

To me, that is a significant discrepancy that we have to fix, and there
are two ways of doing that:

1) introduce planning of major releases of openSUSE. 2) stop using
major.minor version numbers of openSUSE.

I have sofar been advocating 1), but as the powers that be do not seem
overly inclined to taking that path, 2) is the alternative. I would
much prefer "openSUSE 27" with no implied meaning over "openSUSE 13.4"
with an implied ".4" meaning of no value.

Arguably, though, there'd be no difference between 27 and 28, why not go
from 20 to 30?

When it comes down to it, release versioning is more of a marketing
exercise than it is a project management/product management exercise
(versioning is important to identify changes in SVN, for example, but
those revision numbers only have meaning in that they are a measure of
timed events - commits - that show a progression).

So in your example, the difference really is that the decimal point is in
the "wrong" place.

But maybe more logically, 12.0 is the release that comes after 11.9
(mathematically, that's certainly true - versioning systems that insist
on 11.10 being > 11.9 and 11.1 != 11.10 notwithstanding - I've always
considered that to be a special kind of nonsense personally, though I
understand why it exists).

Since major.minor in marketing terms implies small changes in each
".minor" increase and large changes for each 'major' digit increase, that
does become a marketing discussion.

So from the standpoint of a marketing discussion, I agree, major.minor
doesn't make much sense unless the user can actually discern the meaning
of a particular change in version number. Otherwise it's arbitrary.

Jim
--
Jim Henderson
Please keep on-topic replies on the list so everyone benefits

--
To unsubscribe, e-mail: opensuse-project+unsubscribe@xxxxxxxxxxxx
For additional commands, e-mail: opensuse-project+help@xxxxxxxxxxxx

< Previous Next >
Follow Ups