harmony-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Sian January" <sianjanu...@googlemail.com>
Subject Re: [general] proposal : starting a release train
Date Thu, 08 Mar 2007 10:05:40 GMT
Hi Geir,

Just a quick thought about the names.  I like the idea of having milestone
releases (M1, M2 etc).  However if the release is called "Apache Harmony Pre
M1" or "Apache Harmony M1", and then we want to use M1, M2 etc for future
versions like Java 6 and Java 7 how will we tell the difference between the
Java 5 M1 release and the Java 6 M1 release?

Thanks,

Sian



On 07/03/07, Geir Magnusson Jr. <geir@pobox.com> wrote:
>
> This is a new thread, building on the thread that Mikhail started, to
> get away from some of the shoals we encountered.
>
> I'd like to propose we start a formal release schedule :
>
> *  Monthly releases, targeted for the last day of each
>     month. (IOW, avoid the April 1 release)
>
> *  For some set of platforms that we support
>
> *  Regression free (if we support an app or feature in release X,
>     we support it in X+1 unless we agree in a vote to
>     regress).
>
> A lot of good things will come from starting down this road :
>
> 1) We can start putting together the process to decide on "code
> freeze", building and testing, etc.
>
> 2) This will allow us to do JIRA triage that is meaningful - we can
> pre-set the releases in JIRA, and bucket the JIRAs into the release
> buckets
>
> 3) We can chart out what we want to do in Wiki (and then harvest for
> the site) what we hope to accomplish in each release.  Up to now, the
> roadmap was very high-level, and for the most part, we've been
> amazingly good at following it (especially considering it was a SWAG
> done in response to a question I got on a concall once...)  I would
> love to see us chart out the remainder of the API, some sense of VM
> feature (like "switch to GCv5"), and a set of apps we plan to have
> running.
>
> An important element to this will be naming.  I think that as we're
> just figuring this out, we should avoid date-oriented names (like
> "March", "April"...) so we can shift the schedule as need be.  While
> I like the "M1, M2, M3" scheme,  I personally utterly detest the "1.0-
> M1" approach, as it ties your hands horribly.  I think that we should
> avoid prefixing a version number for now, and use version numbers
> when we get the JCK and can plan out a 0.5, 0.6, 0.7... to 1.0 train
>
> So it could be
>
> Apache Harmony Pre M1
> Apache Harmony Pre M2
> Apache Harmony Pre M3
> ...
>
> for now.   This is really common and banal.  Boring....
>
> We're building on the shoulders of giants here, and I think it would
> be nice to honor those that came before us.  Here's a list of great
> people in the history of Java who I'd love to honor, as many are my
> heros :
>
> 1) "Hamilton" (Graham, The Enforcer)
> 2) "Bloch" (Josh, API hero)
> 3) "Gafter" (Neil, API hero)
> 4) "Bracha" (Gilad, VM hero)
> 5) "Click" (Cliff Click, hotspot hero)
> 6) "Gingel" (Rob, Chief Engineer, and guy who grokked OSS)
> 7) "Detlefs" (David, GC hero)
> 8) "Lindholm" (Tim, VM spec guy)
> 9) "Yellin" (Frank, VM spec guy)
>
> Thoughts?  Comments?  Flames?
>
> geir
>
>
>
>


-- 
Unless stated otherwise above:
IBM United Kingdom Limited - Registered in England and Wales with number
741598.
Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message