Re: [PATCH] pack-objects: name pack files after trailer hash

From: Junio C Hamano <gitster@pobox.com>
Date: Dec 16, 2013, 19:33 UTC
Message-ID: <xmqqzjo0oako.fsf@gitster.dls.corp.google.com>
In-Reply-To: <20131216190445.GB29324@sigill.intra.peff.net>

Jeff King <peff@peff.net> writes:

> I was tempted to explicitly say something like "this is
> opaque and meaningless to you, don't rely on it", but I don't know that
> there is any need.

Thanks.

When we did the original naming, it was envisioned that we may use the name for fsck to make sure that the pack contains what it contains in the name, but it never materialized. The most prominent and useful characteristic of the new naming scheme is that two packfiles with the same name must be identical, and we may want to start using it some time later once everybody repacked their packs with the updated pack-objects.

But until that time comes, some packs in existing repositories will hash to their names while others do not, so spelling out how the new names are derived without saying older pack-objects used to name their output differently may add more confusion than it is worth.

>  	<base-name> to determine the name of the created file.
>  	When this option is used, the two files are written in
>  	<base-name>-<SHA-1>.{pack,idx} files.  <SHA-1> is a hash
> +	of the bytes of the packfile, and is written to the standard

"hash of the bytes of the packfile" tempts one to do

    $ sha1sum .git/objects/pack/pack-*.pack

but that is not what we expect. I wonder if there are better ways to phrase it (or alternatively perhaps we want to make that expectation hold by updating our code to hash)?