[Monotone-devel] Re: user friendly revision identifies

[Emile Snyder]

On Fri, 2004-12-03 at 14:44, graydon hoare wrote:

> I guess there's a fundamental issue here I don't address enough: from my 
> perspective, bugs in implementation are *the most prevalent* form of 
> failure this system has to put up with, so when someone discusses a 
> scheme involving any form of external identifiers (rather than hashes) 

I didn't mean to suggest that monotone should stop using hashes as the
identifiers internally; I think that's a great design.  I was thinking
more of another type of selector I guess, and didn't use the right
terminology.  But I see now the issue with trying to uniquely specify a
database.

> my first thought is to consider how much more (or less) code will be 
> involved. usually it's more: code to manage extra tables and extra 
> relationships between these tables and existing ones. the code to 
> implement hashing is actually quite small and easily tested.

Less code is always better ;)  But obviously we have to have some code,
so I guess it's just the question of what's useful enough to be worth
the weight.  So far, as I've been playing with using monotone I've had a
hard time with some of the less desirable features of hashes as user
interface objects.  But I probably haven't given myself enough time to
adjust from a CVS mindset, or to get used to the existing selectors.

thanks,
-emile

+----------------------------------------------------------------------
We're standing there pounding a dead parrot on the counter, and the mgt.
response is to frantically swap in new counters to see if that fixes the
problem. -- Peter Gutmann 
+----------------------------------------------------------------------

signature.asc
Description: This is a digitally signed message part