[development] LinksDB vs. Links

[Syscrusher]

On Friday 28 July 2006 10:40, Matthew Jenkins wrote:
> and there's nothing wrong with a little diversity - different ways of 
> looking at the same problems, etc...

For the record, I have absolutely no problem with others building whatever
link-handling modules they wish, using or not using my API as they prefer. I'm
a firm believer in the Open Source philosophy of letting the user community
filter for the best code by voting with their feet. I created Links to scratch
my own itch, because there was nothing else that did exactly what I wanted, and
am sharing it with anyone else who can use it. There's no reason why we have to
have just one links solution, any more than there is any reason why we can only
have one theme engine or one photo album manager.

The main philosophy behind Links is to have each unique URL stored only *once*
in the system, and therefore dead links detection and related admin functions
only have to visit each unique URL *once* per iteration. Point solutions are
leaner and cleaner if you only need to use URLs in one way on your site (e.g.,
just a Yahoo-like directory and nothing else). Links API is built to handle the
larger-scale situation, and there may indeed be a case to be made for a point
solution for the entry-level user who just wants a place to put links and
doesn't need to have a scalable management backend for them later. :-)

Different strokes for different folks, and all that. Maybe when Links and
LinksDB are both done (I don't know the release status of the latter) we can
collaborate on a point-by-point feature comparison to help admins choose the
right one for their site.

Scott (Syscrusher)

-- 
-------------------------------------------------------------------------------
Syscrusher (Scott Courtney)          Drupal page:   http://drupal.org/user/9184
syscrusher at 4th dot com            Home page:     http://4th.com/