On Oct 18, 2006, at 8:04 PM, Mads Toftum wrote:
On Wed, Oct 18, 2006 at 07:10:37PM +0200, Mads Toftum wrote:
On the surface everything appears fine to the point of graduating
minus
the proving that you can do a release. But given past experiences
with
graduating large projects too fast,
Do note that harmony has been set up for a different kind of
incubation from day 0 as part of all the lessons learned from
previous big-project incubations.
I'd be much in favor of keeping
Harmony in incubation for another 6 months - it worries me
slightly that
you seem to be in haste to move on almost immediately after all
requirements look ok, not giving it a bit of time to prove that
this is
not just a temporary state.
All requirements looked ok (goes check his harmony-thoughts.txt...) 5
months ago. We didn't move on then, just to give it another few
months to make sure we were in a stable state.
Also there's the whole issue of Sun and
opensourcing java - sure, it shouldn't affect Harmony directly,
but on
the other hand I see a risk there (not great, but they may
surprise us).
At apachecon in austin last week, there was an "open source java
UnBOF" with harmony people, sun people, a whole range of ASF members
there, including most of the board, and other people. I think
afterwards there were various mixed opinions on various things
(inevitable with that kind of crowd), but I know I (and others) came
away with that feeling reassured that the whole sun opensources java
thing wouldn't negatively affect harmony.
In any case, I don't think that's any kind of concern for the incubator.
Just my EUR.02s worth on non-binding -1.
Note that, as an ASF member, you can ask to be on the incubator PMC,
and then your vote would be binding. I don't think it matters much --
non-binding votes tend to be taken seriously as well.
Gah, that went out a bit too fast. I missed saying that I'm biased by
thinking that the project should never have happened in the first
place.
Not so much that there shouldn't be an a free implementation of java -
but that a project like this is bound to grow to a size where it
doesn't
fit with the rest of the foundation. Implementing a whole language is
big enough that there's plenty of foundations out there doing nothing
but a language implementation (perl, python, ....) - I think that
would
be the best for a project like Harmony as well.
Fair enough. I would say that this is sort of a foundation-level
concern, but in terms of how incubation works right now, I think we
shouldn't factor that into the "graduate or not" decision. If
"harmony" ever becomes something that doesn't fit the rest of the
foundation (I just don't see that right now), I think that should be
addressed then, independently of whether it is incubating or not.
The other bit is the speed at which it happened - I'd expected Harmony
to be in the incubator for years and now it is aiming for graduation
less than 18 months later without having even neared the first
release.
Yeah, harmony development is going at an amazing pace. I had never
expected it to be where it is now this soon. Yet everything seems to
be in place for that development pace to continue.
So, that's what my reluctance to letting Harmony graduate is about -
feel free to ignore it.
It is not ignored.
LSD
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
