Re: [Monotone-devel] Monotone Security

[Daniel Carrera]

Peter Stirling wrote:
> Presumably the issue is not when it's only one revision. If you are 
> intent on malice then you could add 10,000,000 revisions, at which
> point calling kill_rev_locally will be rather more effort, especially
> if the revisions are signed by a (compromised or otherwise) key which
> had some revisions that you don't want to delete.

Ok. So the encumberance attack becomes something more like a DOS attack. 
Could you not use selectors to delete everything more recent than a 
certain date that was signed by the compromised key?

Daniel.

PS: For those who missed my first email, here's the link to the article 
I wrote:

http://daniel.carrera.name/Monotone_Security/