[darcs-users] Why Bitkeeper still wins

[David Brown]

On Tue, 22 Mar 2005 14:42:03 +0100, Peter Busser <busser at m-privacy.de>  
wrote:

> On Tuesday 22 March 2005 11:47, Jamie Webb wrote:
>> On Tue, Mar 22, 2005 at 08:20:48AM +0100, Peter Busser wrote:
>> > And it uses
>> > SHA-1 hashes for everything. SHA-1 is starting to reach the end its
>> > useful life.
>>
>> That depends on your definition of useful.
>
> Things only can get worse for SHA-1, not better. Therefore it is arrogant
> and/or foolish to ignore the advice to (slowly) move away from SHA-1.   
> People
> in the field *are* giving the advice to migrate to more stronger hashes.

SHA-1 hasn't lost any usefulness as a non-cryptographic hash function.   
Systems that use it for identification of file data aren't going to have  
any problem with it.  The only thing that might be needed would be a  
larger hash function.  Even if someone determines how to trivially break  
SHA-1, it is still useful for identification.  True, a malicious use could  
then start generating conflicting files, but it is unlikely to come up in  
real use of a revision control system.

OpenCM also uses SHA-1 (currently, although they are discussing needing  
more bits) to identify each file at each rev.  It is a convenient way to  
refer to an exact version of a file.

Dave