incubator-general mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Nicola Ken Barozzi <>
Subject Re: Some sparse nots about changes to the incubation process
Date Tue, 08 Jul 2003 21:37:00 GMT

Paul Hammant wrote, On 08/07/2003 17.44:

> Nicola,
>>>What is the problem you are trying to solve here? 
>>There have been discussions on the PMC lists; mainly the problem is 
>>about having a project look like and act like a normal Apache project, 
>>while in fact it's not. It's *becoming* an Apache Project, but it might 
>>as well fail. This has to be made very clear for the developers that 
>>help it and for our users.
> OK. Here's the deal. Incubator projects are not listed immediately off the main site,
like jakarta
> ones are. There is already some clarity with that design :-)

Could help, not sure. The main goal of points 4,6 and 7 is to ensure 
that the project and the users know that's in in incubation, not a final 

>>I'm not advocating changes to the current projects. I'm advocating a 
>>different way of handling the next one, XMLBeans.
> One depot will get mightily noisy dude. Especially for new-and-refactoring projects.

As I told Steven some mails back, I understand the point. Hence I agree 
with the incubator-projectname CVS module name and separate mailing 
list(s) if the project is big enough (as I gather this is).

Here is a rewrite of those points:


4. Making the incubating project have it's own CVS module with the final 
destination name is a terrible idea. It gives the incubating project and 
the new users the idea that the project is already a full-apache 
project. This is the same problem that had arised on Avalon, where 
projects that were really sandbox code started appearing in the main 
excalibur CVS, making users use them as they were properly released, 
which was not true.

All incubating projects therefore should reside in a module that is 
names incubator-projectname.
I'd also propose that all committers of the sponsoring PMC can have 
access to it, as well as all the committers from other incubating 
projects (so they can eventually lend a hand in need).


6. Making a separate mailing list right from the start that is in the 
final destination address is again IMHO not a good idea. I would propose 
that all incubating projects start de-facto on the ML if they have low traffic, and 
eventually migrate to In case the 
project has already sustained traffic, we will create MLs of type 

7. The webpage of the incubatong project should have the incubator logo 
and the project one. The final PMC may link to it, and have their logo 
somewhere else on the page, but not as part of the header.

7-b. The final PMC project should not link to the incubating project 
page, but only to the incubator main page.

>>Use what you prefer for your project, there is no rule to which 
>>build-document-whatever system you use; just don'├Čt impose it on others.
>>(Forrest is much more advanced for Documentation BTW)
> Knew that. But clover, pmd, simian, checkstyle and other reports are magic :-)

They are not Maven, are they? ;-P

> When will that Forrest plugin for Maven be ready ?

ASA Leo gets actively working on it.

Nicola Ken Barozzi         
             - verba volant, scripta manent -
    (discussions get forgotten, just code remains)

To unsubscribe, e-mail:
For additional commands, e-mail:

View raw message