river-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Mark Brouwer <mark.brou...@cheiron.org>
Subject Re: Short term plan forward... (proposal)
Date Wed, 23 May 2007 10:28:32 GMT
Hi Dan,

Dan Creswell wrote:

> I can imagine that a release with some bug fixes mightn't get too much
> contention but would take longer as we thrash around unit tests etc
> (which I think are essential).

Setting things up will take its time, such as getting the JIRA issues
in, test environment, etc., etc. During that process it is my opinion
other tasks can be done in parallel.

Although my opinion is a minority in this case, which is fine, but still
even Jim's proposal leaves some room for solving issues so I'm going to
enter the twilight zone and want a bit more done than what people can
download and is already available since October 2005.

I think the process of picking issues, doing code review and working
together is something we should 'train' even for the first release. It
won't be hard to pick already a few simple issues for which there are
currently no direct tests, or you won't/can't write unit tests and that
are normally just done by code review. Such as there are:

RIVER-5, RIVER-7, RIVER-8, RIVER-9, RIVER-18, RIVER-24, RIVER-25 (I
expect some people to differ from opinion whether these are trivial but
that is OK). There would have been more although they are related to
files that contain larger fixes as well. The issues are already listed,
the fixes are already in the field, so I really can't see the problem of
a bit of exercise on the code and us working together.

> If we get into RFE's, from my own past discussions in this area, I know
> I'm going to have a lot of things to say and a lot of stuff to
> challenge/thrash out.  Basically my view is that will delay the first
> release far too long.

I agree, but I expect each of us can make a judgment about when it is an
opportune moment to make these public (rather sooner than later I would
say) and for which phase to schedule it for discussion and
incorporation. And misjudgments are a good way to learn to work in this
project, so a few of them are even necessary I guess :-)

My problem so far is that getting a release out of the door as a sign of
life seems to be the most important aspect instead of us (all) working
together. To me the release represents a gesture to the world of our
good intentions (which can be important although I'm not so very
sensitive to that) but has no advantage over October 2005 (legally it
is even an incubator release) while at the same time it is putting some
people in the wait mode and that I have a 'slight problem' with.

I believe the questions of Bob are still unanswered so probably it is
best to have people who want to have a release out of the door ASAP
provide an answer for them and make a task of the things to be done for
the first release and then have a vote about it. I don't know whether a
vote is normal in this case but no doubt our mentors or somebody more
acquainted with the process here will help us out with that.

I hope my language is not too harsh. I'm Dutch so that means I've been
brought up in a consensus society where people
go-with-the-flow-although-screaming-and-kicking so probably I support
what the majority ends up with if they let me do 3 issues ;-).
-- 
Mark





Mime
View raw message