To me this whole notion of which format to use, if it should be WYSIWYG, and what kind of markup to use is somewhat dependent on that usage scenarios one consider.
I too would like to see something which moves beyond plain strings. As the computer scientist I am, would also like it to be something uniform across usage. However - there is difference in how we I view a comment and how I view a book. In a method comment I would like to be able to add a little bold, and perhaps a few links to elsewhere. In a class method I would like to be to also do lists, perhaps a heading or two (which I could do with bold perhaps). A package comment I start to want to include diagrams and images. In a full book or paper, well, I want it all there (aka level of Latex or Word). Pillar is no worse than 99% of the other markup's out there, so fine with me. What I tried in my little Pillar/Nautilus combo was to show how one can make a morph which shows pillar markup as a rendered text (as opposed to string), but editing that text is done by editing the pillar markup. I believe someone with more experience than I can make one that works more smoothly. Also, I believe the same trick can be used of inlined method comments. So, my point is, that Pillar and other markup languages were intended to be for wiki, and web pages, not for books. That is something at the level of class comments or package comments. -- View this message in context: http://forum.world.st/Improving-the-documentation-model-tp4820814p4821286.html Sent from the Pharo Smalltalk Developers mailing list archive at Nabble.com.
