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

Re: Dropping the .0 on release numbers?

On Tue, Sep 14, 2010 at 01:58:51PM -0400, Michael Gilbert wrote:
> The .0 actually has quite a bit relevance since it signifies a new
> major long-term release.  It also demonstrates stability when used in
> conjunction with the third digit.  6.0.1 seems like a rather minor
> update, which accurately describes stable point updates.  Whereas, 6.1
> seems like a much more experimental update.

If somebody wants to infer such meaning from the version strings we use,
that's up to them, but such an inference really isn't accurate.  When
you only have one "stable" version at a time, and every new stable
version constitutes a major release, the .0 is redudant.  

IMO, when a release identifier consists of just an integer, incrementing
that integer is a pretty strong suggestion that a new major release has
happened.  A .0 suffix isn't necessary for that.

To signify minor releases, some people do things like 5.01, indicating a
smaller change than 6.1.  That'd be one possibility for stable updates.
Or maybe just something like 5r1 as we've done in the past.

> Also, as I mentioned elsewhere, testing could start to get beta-like
> versions numbers, which would be useful for branding snapshots (e.g.
> 6.9.20100912).

Why not just 20100912?  What does "6.9" actually signify, and how is
6.9.20100912 distinguishable from some update to 6.9 in this scheme?

For the record, count me in as supporting integer release numbers.  I've
wanted to see this for some time.


Attachment: signature.asc
Description: Digital signature

Reply to: