Monday, October 19, 2009

The LaTeX Syndrom

Reading a certain paper for my seminar of software engineering, I had nausea when seeing the appalling notation they chose to talk about the soundness of their system. My supervisor told me it was of no importance but still I can't help but think that the notation could be improved a lot. They seem to have chosen to use a formalism in order to pack as much information as possible in one page and I am convinced that they must not have used it much with pen and paper because it would have been much to heavy. This leads me to the hypothetical LaTeX syndrome. I think they needed LaTeX to typeset such a notation and would not be able to do it without LaTeX. I think in the design of a notation, it should be taken as a yardstick that the notation is usable with pen and paper and it must not be a torture to use it for any length of time. I would go one step further and state that I would like those that do not follow this principle to be condemned to use their formalism on paper for as long as necessary for them to repent. I could also ask for a reward for the people that do abide to the principle: they could have the joy of using their formalism and see their problems nicely stated and elegantly solved with it. I am not making a statement against LaTeX here, at least not yet. What I am complaining about is the people that use it without the application of good judgement. As a reader, I think the bad formalism was an important difficulty in seeing the point they used it for. Fortunately for me, they did not use it too much. I must admit that I now avoid LaTeX altogether since most of my efforts are spent on LaTeX itself rather than on the problem at hand. I have started using rich text editors which don't pretend to be smart. The big drawback is that they are harder to use with version control. For that reason, I'm thinking of using plain text but, for some reason, Mac OS does not make that easy. Simon Hudon Written after a bad day on October 19th 2009 Meilen

2 comments:

  1. I think you meant "withOUT" in the second sentence of your second paragraph.

    As a mathematician, I use LaTeX when I want the math equations I use to be nicely presented. I'm not sure to understand how you write your equations with your rich text editor...

    ReplyDelete
  2. Thanks for the correction!

    As for how I do in rich text: I don't really use any particular feature of rich text except being able put something in bold. I use mostly predicate calculus and I don't need many levels of formulae in one expression as you would with fractions or inference rules for example. And for fancy symbols, I found some pretty satisfactory ascii replacements. I use /\ and \/ for conjunction and disjunction, A and E in bold for universal and existential quantifications and my awesome mac keyboard supports ¬, ≠, ≤ and ≥. As for subscript and superscripts, I was already avoiding them before; I replaced subscripted variables by functions if I need to look with variables, otherwise, multi-character variable names are fine with me since I don't use any invisible operators. Finally, for the iterated summation and product, I use the same unambiguous notation as for quantification (which, it is):

    ( op x,y: x > y /\ y > 0: x + y )

    For op of your choice. It has the benefit of delineating clearly the scope of the dummies (x, y) and identifying them non-ambiguously.

    ReplyDelete