[antlr-interest] ST 4.0 planning

[Austin Hastings]

Brent,

>> pretend you have a SQL database as the ONLY thing available 
>> to the template engine
>>     
>
> I don't know ST well enough to comment on its implementation, but it this is
> an intriguing idea. Wouldn't you need expression support, in addition, to
> allow the manipulation/combination of that data?
>   


I don't know - that's why it was a thought experiment. The same thing 
could be said for xpath (except that would make ST look a lot like every 
other XSLT processor out there).
 
The point is that if you have a single format for data, then maybe it's 
a little easier to do template work. So the program fires up sqllite, 
and creates tables and populates them, and one part can overwrite the 
stuff done by another part, and that's okay. And when it's all ready, 
the template engine runs and the template gets to make SQL calls.
 
The same thing would be true for Xpath. The program builds a tree, of 
whatever shape, and then the template looks things up.

The point is not "I think you should include SQL in the templates." The 
point is "there should be a common data source." Some kind of blackboard 
model (global data, single instance) makes sense to me considering the 
nature of the problem. But it has to support some kind of internal 
structuring. That's what the experiment is for.

=Austin