Daniel Fagerstrom wrote:
First, let me say I don't completely disagree with deprecating some
blocks or components, although it may appear that way from my comments
below. But this is not a simple issue.
We are chosen as committers as induviduals and not as representants
for our companies. From a community stand point I would say that it is
time to deprecate the SQLTransformer. As a representative for my
company I would rather say: no way, we have tons of code that depend
on it. It is a complicated question, but I don't think that the answer
is: I need it at my work so the rest of you should support it.
This is a really good point but I'm not sure I'd come to the same
conclusion. Personally, I know I'd never want to use the SQLTransformer
for any of the projects I work on. But then, if I needed to create a
really simple website it might be the quickest way to do it. I tend to
try to look at it from what I think the general Cocoon customer base
wants - and that is imprecise and tricky. In the case of the
SQLTransformer, if we remove it and tell our customers that in order to
upgrade they a) have to maintain the old transformer themself, or b)
have to write a whole bunch of Java code are rearchitect their
application, then I'd be very reluctant to remove the component.
Frankly, I think that is why XSP is still around - some folks still find
it the easiest way to get the job done despite our recommendations that
there is a better way.
I think that a better answer is that in such a case it would better
become my and my companies responsibility to support it. It could
happen inhouse, at SF or in some legacy area in our repository.
I don't think that we should let our community be choked under the
weight of historical functionallity because we are afraid of upsetting
someone.
Well, I don't think "choked" would be the appropriate description. Many
of the blocks are just ignored for long periods of time. Then someone
applies a few patches and they get ignored again. With downloadable
binary blocks I don't see that as a problem.
You have certainly heard that the key to success is to focus and be
really good at doing one thing. Do we have a focus? Are we really good
at one thing?
Take a look at the community. Most of us are focused on one or a few
things. As a whole we focus on a lot of things. I think that that is OK.
The hard thing is to prioritize, it is easy to add but it is hard to
remove. But it is not impossible take a good look at the Maven plugin
matrix that Jorg pointed to:
http://docs.codehaus.org/display/MAVEN/Maven+Plugin+Matrix. They have
removed things and assigned different priorities to different plugins,
are we strong enough to chose?
It is hard for me to correlate the matrix with your point. Maven 1
plugins were written in Jelly. Maven 2 plugins are Java. They all had to
be re-written from scratch. This isn't a case of them "dumping" plugins.
It is a case of them not having reimplemented some of them yet. It
makes sense that they would prioritize that.
/Daniel
Ralph
