On Sat, Nov 10, 2007 at 05:23:17PM +1100, Steve Bennett wrote:
On 11/10/07, Jay R. Ashworth jra@baylink.com wrote:
I believe that all of these arguments about what to do and how to do it would be *very* well served by a) defining a list of the corner cases/pinch points and b) surveying the WMF wikis to see on how many pages they *actually* appear.
So:
- Carefully compile a list of stuff we don't want
- Test the validity of that list
- Throw the list out, since we don't need to implement it.
You're being... *purposefully* obtuse? :-)
It would be better to:
- Define a subset of the grammar that we think is useful
- See how many pages correspond to that grammar
- Implement a parser for that grammar.
But that won't make the sale necessary to implement 4: deploy that parser on WMF.
Cheers, -- jra