an Incubator case study

[Leo Simons <le...@apache.org>]

Hi gang,

The avalon project is currently merging an externally developed 
codebase, "MerlinDeveloper", with an internally developed codebase that 
has the same goals. The codebase is about 60 files. We found some 
licensing issues when we started incubation (referenced LGPL code, wrong 
license headers, etc etc), but these have all been resolved quickly by 
the projects developer once we identified them.

I'm one of the mentors of the project, and as a 'hobby' I've pretty much 
made a detailed record of how much time it takes. So far I have spent:

* 0 hours convincing the incubator that this project should be
     incubated;
* 1 hour convincing the avalon pmc that the project should be incubated;
* 3 hours doing a code inspection (which would have likely taken me
     2 hours longer, btw, hadn't there been forms and process
     descriptions to use in incubator cvs, cause I would have had to find
     out what to look for all by myself);
* <1 hour filling out a status form, sending a status report, running
     project acceptance votes for two PMCs;
* 1 hour figuring out parts of the process as we go ("no, we don't
     really need a mailing list"), but this is also part of my role as
     incubator pmc member, so I'm not sure if I should book it as "time
     spent incubating";
* 4 hours explaining to various parties what we are doing, that it is
     not bureaucratic, that it is not all a big formal mess, that this
     thing really is required, etc etc.;
* I'll probably spend a few more minutes tallying vote results and
     notifying two PMCs of the results.

 From this it is easy to see that if it weren't for all the resentment 
by various parties towards the incubation process, it seems likely that 
the import of the project would have actually been more efficient, and 
easier, than if the import process would have been managed by the avalon 
project all by itself, because avalon benefits from the expertise and 
knowledge of the incubator project.

I'll also note that 'defending' that something should be done and spent 
time on is something that I really don't enjoy at all, so those 4 hours 
spent count double IMHO.

Do with this data what you please. The conclusion I personally draw from 
it is "we should stop complaining about process, and JFDI".

-- 
cheers,

- Leo Simons


---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org