authentication - Verifying Mercurial Changesets are from who they say they're from -

I am checking to use Mercurial in corporate environments. I plan to be central to a webserver (IIS) hosted by Using the treasure, developers who push locally or after testing the changes in their teams will push.

I have been configured to authenticate users against IIS Active Directory, but there is a hole in it, while I can apply it that can push, I do not apply it They can sign their changes as themselves. For example, a basic "committed" scenario is given:

Users are committed to their local repository

User puts their changes in the central repository

In step 1, the user provides the user name for their .hgrc file or whatever) for their local repository but in reality there is no way to apply it to their "real" user name is.

In step 2, the user is allowed to push credentials to "real" IIS, but in the history of their changes, they will appear with the username given in step 1. It seems that Bob has used "Alice" as his user name in step 1, then he is sure that Elise has been convicted for any change in his buggy.

Is there any way to ensure that these usernames get match (by hook or something) during push? Or alternatively, some other way to ensure a proper level of authenticity in the transformation?

Edit: On the further consideration, I think I do not really want to apply it to name line up; If Bob and Alice are collaborating in a separate repo, Bob will eventually be able to push all his changes forward, not just his own. What I really want to do is to make sure that if it comes down, then I can tell which makes a change in a more definite way than what the username was used.

I am thinking the answer is, but I still do not think that I got the full picture.

I finally found, which essentially says that my options are essentially to everyone's GPG Along with signing for changes, or installing from the side of a pulldown that pushes the user towards the central repository.

Ry4an also told (essentially duplicates) confirms with some good answers that I found elsewhere.