[plt-dev] setup-plt parallelization progress

On Jan 11, Jay McCarthy wrote:
> 
> We're looking at smarter ways that just random collection orders to
> make it better. We have lots of ideas but some are more expedient than
> others given the reality of the information available from the
> compilation manager. If you have anything you think we haven't thought
> of, we'd be happy to know. Also, right now we don't do anything to
> parallelize the doc building or indexing.

Whatever you end up with, it would be a really good idea to make it
deterministic somehow.  If the first instance starts the
sub-instances, then you could at least do the random splitting in the
first and pass it as arguments to the subs, using some known random
seed to do the random choosing.

(There shouldn't be any order issues, but when there are for some
reason, trying to recreate the bug will be a nightmare.  This has
happened in the past, and as a result I made setup-plt always sort the
directories and/or files.)

-- 
          ((lambda (x) (x x)) (lambda (x) (x x)))          Eli Barzilay:
                    http://barzilay.org/                   Maze is Life!