[RFC] history editing vs history presentation

[Robert Collins]

On Wed, 2009-06-10 at 18:28 -0700, Maritza Mendez wrote:
> Points well taken.  Your idea of ”shadow history” matches my second
> thought quite well.  Namely no history is ever lost.  History edits
> are purely additive and history which has been ”overwritten” is simply
> marked as superseded.  This works for all forms of history, not just
> commit messages.
> 
> On the other hand . I reject arguments based on the size of the repo.
> It is perfectly acceptable for extreme use cases to carry matching
> consequences.  If that becomes unbearable, then just as you pointed
> out in a different context, the user can branch to a new location,
> leaving behind the "shadow" history.

Actually, branching would copy the shadow history too for a number of
reasons. Disconnecting permanent links in history is much harder than
removing external annotations. I wouldn't have raised size as a concern
if it wasn't :). I assert that users need the ability to choose what
degree of auditing and rollback they need - and I think we're agreed
there.

-Rob
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 197 bytes
Desc: This is a digitally signed message part
Url : https://lists.ubuntu.com/archives/bazaar/attachments/20090611/e757c98b/attachment.pgp