From: Luben Tuikov <ltuikov@yahoo.com>
Date: Tue, 26 Sep 2006 18:53:26 GMT
Subject: Re: [PATCH] gitweb: "raw" links to blobs at HEAD
Message-ID: <20060926185326.10590.qmail@web31810.mail.mud.yahoo.com>
In-Reply-To: <7v8xk7rup3.fsf@assigned-by-dhcp.cox.net>

--- Junio C Hamano <junkio@cox.net> wrote:
> This does not feel right.  Go to summary view, pick the tree
> link of a non-head commit from the shortlog list, and the raw
> links on the resulting page all say hb=HEAD.

I see your point.

> Doesn't this make clicking the entry itself (blob view) and
> clicking on the raw link (blob-plain view) inconsistent?

It this exactly this consistency which the patch used to de-couple
it and use the "raw" to point to HEAD.

> The former goes to the blob from that revision while the latter
> always goes to the HEAD.

Yes, I completely understand the conundrum.

Ideally, hb would equal HEAD from the _context_, but
  1) hb and h != HEAD for any "raw" link,
  2) one can set hb=HEAD in the context only when the "blob page"
     can be loaded, which is impossible for binary blobs.

So the fix would be that, unless the user had selected a hb/h
explicitly, set "raw" to HEAD (i.e. by default), but leave the entry
with hb/h in SHA.  This way, one can take a reference to HEAD or
SHA of the head depending on what is needed.  And if they had
selected a hb/h explicitly then both the entry and the "raw"
would point to the SHA.

Such a patch can go on top of those two.

    Luben