At 11:40 2001-10-27 -0700, Daniel A. Koepke wrote: >On Fri, 26 Oct 2001, Caniffe wrote: > > > In saying all of that though, I didn't mean to forget OBuild. I was > > conceptualising on an OLC system to include, and first thought was Oasis. > >The system to include--if any--is GenOLC with a thin interface to its >functionality. Then Oasis and Obuild and NoteOLC and some other package >simply become front-ends. The thin interface's primary purpose would be >as a demonstration of writing GenOLC front-ends. > >This is the most inclusive of all choices. No one really wants to write >the code for the low-level OLC routines and there wouldn't be too much >point in rewriting them. Some people want to make their own interface to >the OLC system, and that'd be made that much easier. Some people don't >want to bother doing that, and they can just use the Oasis, Obuild, Note, >etc. front-ends that suit them. (Sure, why not use several?) IMHO: One thing to note about the GenOLC package is that it has really funky functions for mobs. It trods through several functions just to write down the mob file, and if and when you modify this code in any way, it becomes obvious that the only thorn in your side is genmob.c. The flow is hard to follow compared to the other files; which I noticed when I converted it to MySQL support this weekend. The code juggles the file pointer around and calls other functions to write especs and various other things. Should have been kept in one function just like the other OLC-types. It wasn't hard, but it should have been structured like the rest of the code. Just my two cents, /Torgny -- +---------------------------------------------------------------+ | FAQ: http://qsilver.queensu.ca/~fletchra/Circle/list-faq.html | | Archives: http://post.queensu.ca/listserv/wwwarch/circle.html | +---------------------------------------------------------------+
This archive was generated by hypermail 2b30 : 12/06/01 PST