war story: parallel(1) command

[Eric]

On Thu, 1 Aug 2013, Christopher Browne wrote:

> While I agree that SHA-1 was designed to try to resist collisions, I
> do not think I have quite so much faith as you do as to the permanency
> or *true* impossibility of this.
>
> People thought that the Knapsack cipher was strong; it turned out not
> to be nearly as strong as imagined.  It would be unfortunate if Git
> depended (to the point of falling apart destructively!) on the True
> Impossibility of SHA-1 collisions, only for us to discover, 10 years
> from now, that someone can induce collisions at some moderate
> computational cost, and, thereby, throw patches at Git repositories
> that make them fall over.

I agree.
My argument was not tied specifically to SHA-1.

> But I think I rather go with Cromwell, and be a little paranoid that
> what, at this point, seems "impossible before the heat death of the
> universe" might prove more tractable.

Yes, then the initial mathematical assumptions were wrong.
In that case argue the math, not the probabilities.

--
Eric B.
--
The Toronto Linux Users Group.      Meetings: http://gtalug.org/
TLUG requests: Linux topics, No HTML, wrap text below 80 columns
How to UNSUBSCRIBE: http://gtalug.org/wiki/Mailing_lists