[PATCH 1 of 6] bundle2: generate check:heads in a independent function

[Matt Mackall]

On Fri, 2015-10-02 at 14:55 -0700, Ryan McElroy wrote:
> On 10/2/2015 2:33 PM, Ryan McElroy wrote:
> 
> No I didn't.
> 
> > # HG changeset patch
> > # User Ryan McElroy <rmcelroy at fb.com>
> > # Date 1443721694 25200
> > #      Thu Oct 01 10:48:14 2015 -0700
> > # Node ID b75c4651b186360298089c0650c2f4eed0046265
> > # Parent  97dc6ab42aad232c73180dee648685c26662230b
> > bundle2: generate check:heads in a independent function
> >
> I don't have any idea where these patches came from. I certainly didn't 
> run any commands recently to send these out. Please disregard.
> 
> Anybody know what's going on here?

The pushgate works by extracting the email address from the first
changeset in a pushed series and uses that as the source address for a
patchbomb. As always with SMTP, there are basically no defenses against
forging email, and pushgate does it on purpose so the author can get
replies. Here, someone in Canada apparently misfired with either a push
-r or a rebase and pushed a bunch of old changes.

I might try to add a sanity check that prevents pushgate from accepting
series with more than one author to prevent accidents like this.

-- 
Mathematics is the supreme nostalgia of our time.