Re: git copies of cvs modules available

[Holger Weiss <lists%jhweiss.de@localhost>]

* Michal Suchanek <hramrach%centrum.cz@localhost> [2009-10-29 09:54]:
> 2009/10/29 Arnaud Lacombe <lacombar%gmail.com@localhost>:
> > On Tue, Oct 27, 2009 at 9:10 PM, Greg A. Woods <woods%planix.ca@localhost> 
> > wrote:
> >> ..., humans usually just want to see if the IDs are the same or not,
> >> perhaps with some hint as to their relationship in time if they're not.
> >
> > well, as an heavy git user, what I want to know is if the source I'm
> > looking at are the same as the one who has an issue with the source.
> > With a SHA1, no matter on which repo I'll look at, I will have this
> > guarantee (minus an improbable collision). CVS doesn't provide this
> > basic guarantee that 2 same filename with the same CVS number check
> > out at different time are the same (data corruption or malicious
> > change). That is a serious flaw.
>
> The hash git gives you is changeset hash (or something like that).
> This changes every time you reapply the changeset to a different repo
> or a different branch of the same repo and has nothing to do with file
> hashes.

Git stores the contents of files as blob objects which are referenced
via their SHA1 hash in just the same way as commit objects. For $Id$
expansion, the SHA1 of the blob object is used:

| When the attribute ident is set for a path, git replaces $Id$ in
| the blob object with $Id:, followed by the 40-character hexadecimal
| blob object name, followed by a dollar sign $ upon checkout.

[ gitattributes(5) ]

Holger