[DGD] capability based security?

[bart at wotf.org]

On Thu, 24 Mar 2016 17:44:10 +0100, bart wrote

> Performance for such things has very little to do with performance 
> of the language, and everything with picking smart ways to do the 
> work. The language only matters for what is a smart way to implement 
> those things.

TO qualify this a bit, much of the work for a text search engine can be done
with set operations, and mostly depends on having good metadata. It obviously
does help if you have a good performing way of doing those. In case of LPC,
much of the heavy lifting for this can be handled by the driver, and hence
already runs as 'native' code. LPC isn't unique in this, and almost any
language good at this can be used to build a very well performing text search
engine.

Often choice of language is driven by what developers know and are comfortable
with, and not by what in theory might be the best performing language. It
allows developers to concentrate on good implementations instead of on
struggling with unfamiliar language details.

This all doesn't make LPC the obvious choice for building a text search
engine, except maybe for the small group of people who are extremely familiar
with that language and its possible runtime environments. For those, it can
offer a runtime environment small enough to be able to understand the majority
if not all implementation details, which makes prototyping a very nice
experience, virtually all surprises will be due to your prototype, not due to
the runtime environment. Obviously the same is true for other languages with
similar extremely small runtime environments.

Bart.
--
http://www.flickr.com/photos/mrobjective/
http://www.om-d.org/