[mythtv] MythTV - Version 1.0
Jay R. Ashworth
jra at baylink.com
Fri Feb 10 02:43:29 UTC 2006
On Thu, Feb 09, 2006 at 01:02:53PM -0500, Sasha Z wrote:
> On 2/8/06, Jay R. Ashworth <jra at baylink.com> wrote:
> > On Wed, Feb 08, 2006 at 10:54:57AM -0800, Ross Campbell wrote:
> > > > Is it time to call it Version 1.0?
> > > >
> > > > Myth is too good to have it start with a zero!
> > > We have to have a 0.5x and 0.99.x series first :)
> > >
> > > and version 0.99 needs to be followed by 0.100!
> > > <peeve>
> > > I hate it when opensource programs go from version .9 to .10 or .99 to
> > > .100 - version numbers should be NUMBERS. New versions should have
> > > HIGHER numbers than the previous versions, or at very least they
> > > should alphanumerically sort as "higher"
> > > </peeve)
> > You're attempting to parse software version numbers as rational
> > numbers.
> > It should become apparent that this is unreasonable at the point at
> > which version numbers grow a third component (0.18.1).
> > Each component of a version number is an independent item, and should
> > be interpreted that way. 0.100, while less common than 0.10, is
> > perfectly valid, and is higher than 0.10 (since each component of the
> > version number is an independent integer, trailing zeros are
> > significant).
> > This is the most common version numbering scheme, and also the sanest.
> What's insane about things like SVN revision numbers where the number
> is an integer that just keeps growing? Why bother separating into
> major and minor revisions? Certain revision numbers could just have a
> "stable" flag set when they are "released" as opposed to "in
They're not 'insane', I'm just not fond of them for releases. Not all
development milestones are intended to be 'releases', where a release,
with a release number, is intended as a contract between people who run
a program and people who support it, to agree that they're both talking
about the same group of components.
Releases, as the user sees them, should have incremental numbers, for
the same reason that Interstate exits should have sequentially
increasing numbers (instead of being numbered after mileage, which is
worse than useless on several different levels).
Expecting a user to understand that 8742, 8768, and 8831 are
"releases" is, imho, unreasonable.
1.2.3 -> 1.2.4, 1.2.3 -> 1.3.0 and 1.2.3 ->2.0.0 are all version number
changes which convey useful information to a user.
Jay R. Ashworth jra at baylink.com
Designer Baylink RFC 2100
Ashworth & Associates The Things I Think '87 e24
St Petersburg FL USA http://baylink.pitas.com +1 727 647 1274
Q: Should I include quotations after my message body?
More information about the mythtv-dev