Though the idea of bridging is nice, I doubt that it is as simple as just a question of providing syntactic sugar for the "scientist". Any scientist worth calling himself such, can deal with a spelled out lambda as well as with λ. I think the problem is rather that some mathematicians don't care for issues of complexity and implementation, and as you say some software engineers don't understand enough mathematics. That said, a prettyfying solution can make some algorithms look more compact, that perhaps may help in the mental visualization during development.
After all, what is the point of having a unicode supporting system, if we don't make use of all those nice glyphs available to us. :-) On Thu, Jun 6, 2013 at 12:13 PM, Oleg Goldshmidt <[email protected]> wrote: > > > Is this a good idea? Probably absolutely not, but it is quite fun. :-) > > Besides being cool clean fun, I wonder - only semi-jokingly - if it may > be useful to streamline the interface between "algorithmists" and > "developers". In many organizations "scientists" are supposed to produce > algorithms in the form of pseudocode that "developers" are supposed to > translate to working code. Stuff is often lost in translation, the > pseudocode is found to be incomplete or buggy or completely > unimplementable, the developers are found to be mathematicaly illiterate > or worse, and so on. > > So can, for the beneft of companies who employ scientists who cannot > program but know maths and corresponding notation alongside developers > who know the syntax but can't figure out what is wanted of them, the gap > be bridged? Can the "mathematician" write down the algo in pseudocode, > using agreed upon conventions, then "unfontify" it and check that it a) > compiles; b) produces correct output for given inputs thus testing for > bugs ike off-by-one tha can easily crawl into unrunnable pseudocode? If > the pseudocode contains some statements that the "pretty-unfontified" > version's compiler barfs on that likely means that the algo is > incomplete and some operation/function/whatever is un(der)?defined. This > will be flagged before it aggravates the coder's life, etc. > > It's fun to muse, not just to write elisp. Thumbs up for knowing to > enjoy yourselves, regardless of utility. > > -- > Oleg Goldshmidt | [email protected] >
_______________________________________________ Linux-il mailing list [email protected] http://mailman.cs.huji.ac.il/mailman/listinfo/linux-il
