[PATCH 5 of 5] manifestv2: implement slow readdelta() without revdiff

[Matt Mackall]

On Tue, 2015-03-31 at 16:22 +0000, Martin von Zweigbergk wrote:
> On Tue, Mar 31, 2015 at 4:08 AM Matt Mackall <mpm at selenic.com> wrote:
> 
> > On Mon, 2015-03-30 at 17:19 -0700, Martin von Zweigbergk wrote:
> > > # HG changeset patch
> > > # User Martin von Zweigbergk <martinvonz at google.com>
> > > # Date 1427514090 25200
> > > #      Fri Mar 27 20:41:30 2015 -0700
> > > # Node ID 165a99e70157e581ef3328efcc3f5a195d388894
> > > # Parent  816d01895c212b8d313d748f9b8c530b7f92c249
> > > manifestv2: implement slow readdelta() without revdiff
> >
> > These are queued for default, thanks. I'm going to probably want to
> > change the manifestv2 name down the road so don't get too attached to
> > it.
> >
> 
> Yes, I expected some discussion about the name. Do you already know what
> you want to name it? I'm also happy to hear suggestions for the flag that
> says "store as v2, hash as v2" (for BC).  Do you want me to continue with
> the manifestv2 name for now regardless and rename it later? FYI, I have
> patches that add _parsev1 and _parsev2 methods and such coming up.

I'd rather name it after a salient feature of the new format. But I'd
rather do a bulk bikeshedding in Montreal.

-- 
Mathematics is the supreme nostalgia of our time.