Brion Vibber wrote:
Sites like Wiktionary or Wikinews are not really suitable for a wiki to begin with. I think it's unfortunate that people have so casually thrown "Wiki" on the front of the name and slapped up a new project wiki without any actual plan for making it work.
Wikipedia is *very well* suited to a wiki model because an encyclopedia is largely unstructured prose text, with cross-references between articles and some light indexing. That's exactly what a wiki is, and our wiki software is strongly geared to that.
Wiktionary requires structured data which is *completely* unlike what the wiki model provides. The project's been limping along with unstructured text for years and is going to have a hard time converting all that when the appropriate software is finally written for it (a project which is soon to get underway, I'm told, with a grant for development which Erik Moeller will be working on).
This really depends on the vision that one applies to a dictionary. If the focus is on the host language of a particular Wiktionary the present software can be made to do, but I can see where it will not be adequate for Gerard's vision of a super-dictionary. The super-dictionary has a much stronger focus on translation issues, and that can tolerate somewhat less semantic debate than might be the case in a particular language. Once the super-dictionary has a working model it will probably be easier to co-ordinate between that and the various language specific projects.
Ec