obliterate functionality?

[Matt Mackall]

On Tue, 2008-03-18 at 16:09 +0000, Bruce M Simpson wrote:
> My grounds for advocating "obliterate" in view of those points, are that 
> in the situations outlined above, people can and do make mistakes, and 
> even if "obliterate" is not guaranteed to purge all copies of a "leaked" 
> or otherwise mistakenly committed revision, the accountability is there 
> and you can say you tried to obliterate the file, regardless of the 
> downstream change-set penalties or the fact that someone now has a copy 
> of the file (the "dog ate my homework" argument).

You can do that today in various ways. And the actual history editing
will be relatively painless compared to the resulting headaches of
everyone having to rebase their work, so I'm not particularly bothered
that it looks onerous. It -is- onerous.

And I am not interested in the idea of making it a single simple
command. I am quite certain that it would be painfully misused much more
often than not because most everyone who wants it hasn't thought through
the implications. There is no warning I could put on it that would be
big enough. And as I've pointed out before, the mere existence of such a
command would encourage people to make mistakes.

But if someone wants to publish an obliterate extension that rewrites
subsequent history automatically in one command and put their email
address next to it, they're welcome to.

-- 
Mathematics is the supreme nostalgia of our time.