[consulting] Costs of forking

[Jeff Eaton]

> So, you're saying people have to communicate in order to work  
> together? ;)
> 
> Costs of forking tends to mean core, not contrib modules. But, I  
> don't see a ton of traffic on lists, forum, or elsewhere with people  
> saying "I tried to submit a change but module owner X wouldn't let  
> me". In the long run (and the not very long run...) it is lower cost  
> to just submit a patch with your change as an issue.


This seems to be the best approach in most cases, though sometimes the
module developer chooses not to apply a patch. There are quite a few
instances of this -- patches to weblinks.module, for example, were
rejected because links.module was the preferred development path for a
year or so. Only now is it really ready for prime time.

That's not to say that there is an obligation by maintainers to apply
patches, but oftentimes custom functionality is of value to too small a
subset of people to make it into a module (or core, naturally). In these
cases, the 'issue+patch' can still serve as a way of archiving the
patch.

-Jeff