Re: [epp-dev] Re: [eclipse.org-committers] Eclipse project announcement

[Jeff McAffer <jeff@xxxxxxxxx>]

Timothy Webb wrote:
> Jeff,
>
> I appreciate the detailed response.  In short, perhaps the challenge 
> is that it is unclear what has been prototyped and what has been 
> committed to.  As opposed to having this turn into an even more 
> involved discussion, perhaps can we step back and you can clarify a 
> few points?
>
> 1) Is this officially planned for rollout on Eclipse.org's main 
> download area?
The original goal in my mind is to have this available to users from the 
place where they choose packages.  For example as a "customize" link 
beside the packages.  It may also be relevant to trigger the wizard from 
individual project download pages.  That remains to be seen IMHO.
> 2) What is the timeframe that is being targeted? Is this slated for 
> Eclipse 3.5? Or the September refresh for 3.4.1 perhaps?
Like any Eclipse code, when it (whatever it is) provides value and is 
robust enough to be used, it will be deployed.  Personally I would like 
to see it in time for the fall maintenance release.
> 3) When do you expect to ramp up investigation in this area?
Starting to think about that now from my team point of view.  Can't 
speak for others.
> Regarding limitations -- the mere fact that this is being focused on 
> showing off select interesting technologies is by definition a 
> limitation.  In the context of the work being done, it may well be an 
> appropriate limitation but this comes back to my suggestion for 
> stepping back and understanding the goals.
It is a limitation if it prevents other people from doing what they 
want.  To date there is nothing I can think of in any of the design that 
would stop others from choosing completely different implementation 
strategies and moving forward.  I think the challenge here is that you 
are imagining one and only one implementation of the server side code 
and everyone is expected to eat it.  That is not at all what I am 
imagining.  I see people integrating this function into their server 
story whatever it might be in a lightweight way using whatever 
technology they choose with no limitations.  Indeed there is no reason 
why the PHP shopping cart model you mentioned could not be done and 
ultimately feed into (or not) the EPP wizard.  That is, mixing the 
approaches on the same server infrastructure.
> Also, some notes on a Wiki page does not a Waterfall make.  Having a 
> general understanding of the system being investigated / built for 
> before starting with the agile implementation process is certainly in 
> context and there are also plenty of examples of that in the Eclipse 
> space.
Sure and there is nothing wrong with prototyping ideas.  Now we have 
both.  We are sure to be successful. ;-)

Jeff

p.s, heads up, I am unlikely to respond on this thread for a while due 
to other commitments.