On Mon, 16 Dec 2002, William Adams wrote: >Interesting post / problem, and reminiscent of Pages Corp. Pages program > for NeXTStep.
I am not familiar with this program, but I guess proves that it is impossible to invent anything completely new :). > It's my understanding that some such concepts are being considered (were > considered?) in LaTeX 3. That would be interesting, if such a beast should surface it might be possible to add on Lyx support. > Peter Wilson's Memoir Class shows that one can have a wonderfully > flexible documentclass w/ a lot of options w/ just LaTeX2e though, so it > should be possible to my mind to just write a documentclass w/ all (or > most) of the desired options, and just access them through a dialogue > box as is done now. This is a possibility, I have never looked at the memoir class before, I'll look into it. However it would not give the integration of .cls and bst together. Also I'm in favor of keeping the LyX GUI as simple as possible. > Which would be specific / limited to LyX. Also a huge programming task > 'cause TeX is a bit more flexible, hence intractable than say PostScript. It need not be limited to LyX, if there are people in the LaTeX community which want to help it should be possible to keep the extra Lyx specific parts as a extra "option". On the huge-ness of the programming task I would say that we should start small and flexible and add modules as we go. I think I have no illusions about the task, but I also know that larger programming tasks have been done. I think it is down to if we can be clever enough to make it work easily. The GUI would need to be clear and precise. The internals need to be clever enough to preferably read a clear text file and form the GUI from the file, and understand when two options is mutual exclusive and put the right text into the different output files (In the right order). And just maybe know when LyX needs a reconfigure and LaTeX need a texhash. Difficult? Yes; sweet? Yes, at least I think so... Ingar
