[development] Wasting time and effort

Daniel F. Kudwien news at unleashedmind.com
Tue Mar 10 02:51:19 UTC 2009


> As has already been pointed out, we already require an 
> application for CVS access. The picture is being painted like 
> anyone off the street can just start committing to the repo, 
> and that's just not the case.

It is.  And I made the best example myself.  If you ever happen(ed) to be in
the need for a Lightbox on your web site, then you are faced with two
duplicate, I mean, really duplicate modules:  Lightbox2 and jLightbox.  The
latter I am guilty for.  And there is nothing I would regret more today.

jLightbox was created "overnight", because Lightbox2 was incompatible with
something else at that time.  What I did not know was:  There was already a
new branch with a complete rewrite of Lightbox2 in CVS that mostly resulted
in being the same as jLightbox.

Today, I am still trying to find some time to merge some minor changes back
into Lightbox2, and more importantly, write a migration path from jLightbox
to Lightbox2.

This is just one tiny example for the annoying mess of senseless module
duplication on drupal.org.


Causes for duplication:

- Developer did not search.

- Developer did search, but existing module uses different lingo.

- Developer did search, but did not realize that existing module does the
same in an abstracted way.

- Developer was not able to imagine that his contribution would be accepted
in existing module. [<-- jLightbox]

- Maintainer of existing module did neither accept the contribution nor
accept developer as co-maintainer.

- Developer was too lazy to grasp the API of existing module, better write a
new one.

- Developer spent weeks of coding in private, now he must release, even if
it's 100% duplicate (ego).

- Developer wants to attract clients (ego).

- Developer does not know how to patch, just learned how to use CVS from the
handbooks.

Possibly more.


Filtering already published projects will get you the same duplicated
results like now, which you still have to research and possibly test.  The
real issue is, however, that there are many projects that would happily
accept contributions, improvements, extensions, and also new co-maintainers.
But instead of trying hard to let developers join forces, we allow silly
duplication, Drupal users have to suffer from in the end.

sun



More information about the development mailing list