Diff and merge of archives - proposal

[jbowtie at amathaine.com]

On Thu, Oct 14, 2010 at 7:46 AM, John Arbash Meinel
<john at arbash-meinel.com> wrote:
>
> Note that if someone is versioning a 'tar.gz' file, often they need the
> exact binary content. As an example, debian packages build from a tar.gz
> file. However they use a sha/md5sum to make sure that the tarball is
> valid. But they use the shasum of the compressed content. Because of
> this, there have been a lot of hacks (pristine-tar), because gzip is not
> deterministic.
>

If they need exact binary content, then I presume they will resolve a
conflict by choosing either the left or right versions, just as you
are forced to do today.

> (At the least, it includes timestamp info in the metadata, but you also
> have lots of flags and tweaks, such that two people probably won't get
> the same content.)
>

Probably there needs to be some way to choose to ignore "unimportant"
differences in the same way you can currently choose to ignore
whitespace-only changes.

> And if they *don't* need the binary blob to be exact, then why do they
> need to version it as a binary blob?
>

There are a *lot* of free software binary formats that are actually
thinly disguised archives - ODF, compressed Inkscape, etc. In the
absence of more specific handlers we should be able to diff and merge
these as archives, which would help with a *lot* of use cases where
the only viable option right now is to either not use version control
at all or to only allow one person at a time to edit the file.