Friday, December 16, 2011

On the problems of syntax

This is what I mean about getting stuck having to define a default syntax. Consider all the variable and arithmetic links that will exist in version 3.0. (c--.c>0.AndStuff) For an expert defining what an expressive and simple variable opcode might look like, it would be nice to know the final nested order format for the core ugly set (),.*?$!_ . That lets the next stage syntax designer know when pointer is available, when can he recast the keyword over a sequence, default to a known type set on the keyword; stuff like that.

I look at byte codes, like Java and assembly codes and sql compiled tokens and all that. I can get a good guess, but I would never be able to get a consensus on what a nested order looks like.

Here is how the process is going to work. XML will crash the semantic net, real soon. Then folks from SLQITE3, SPARQL, GOOGLE, iCloud, OraCloud, and FaceCloud will want to look at fixed field, binary format nested order. These folks will want to look and act like, real quick, suddenly. Then sqlite3 becomes a default choice, and something like the engine on the right of my blog will be adopted. So I want it to be vague right now, sort of, well, it will look like this, maybe not exactly. I want the entire industry to believe they can do it right. Stealth marketing.

No comments: