re: Long item-list fix

From: Raymond J Nawara Jr. (dreck@rahul.net)
Date: 06/20/95


From: "Daniel [Trice] Koepke" <dkoepke@global.california.com>
   On Mon, 19 Jun 1995, Phil Priston wrote:
   > I have given this some thought but you are gonna need to use some 
   > very impressive gramma controlling code, just look at the example you 
   > give...
   > A is turned into Three:
   > 	This is the easy part...
   > Coat becomes Coats:
   > 	Ok this COULD be done...
   > is lying here becomes are lying here...
   > 	All of a sudden tbis gets a little complicated...
   > 	We are in the relms of an  english language parser like those used 
   > in gramma checkers here.... you wanna write one for circle? *grin*

   You are incorrect in this assumption (sorry for being blunt).

   Since for the count, you need to have an int, short int, something that 
   keeps track of how many, it makes it possible, but the idea I used and 
   implemented in my MUD server is plural and plural-long descriptions and,
   in actuality, the "is"->"are" is very simple.

   (count == 1 "is" : "are")

   So, where's the grammar checker?  I fail to see at all where you'd need a 
   grammar checker, considering all your changing is "is" to "are" if 
   there's more than one..

Your making an assumption here too ;) You're assuming all longdescs
are of the form <article> <noun> <is|are> here. This is not the case
for a significant portion of longdescs in the sample dist, and limits
the creativity of the builders.

For the third time, the pluralization in this manner is possible if
you set limits on the structure of the long desc. If you keep the long
desc an unstructured string, it becomes less than trivial.

	Ray



This archive was generated by hypermail 2b30 : 12/18/00 PST