[darcs-devel] darcs patch: Documentation only - clarify meaning of --from and --a...

David Roundy droundy at darcs.net
Mon Dec 11 15:33:21 PST 2006


On Mon, Dec 11, 2006 at 11:10:53PM +0000, Kirsten Chevalier wrote:
> > I haven't checked the code, but I'm pretty sure that --from *does* set
> > the From address in the mail sent--or at least is intended to, and that
> > that is all that it affects.
> 
> No, I don't think so. The program used to send email is configurable by
> the user. I don't know exactly which email clients do and don't allow
> email to be sent from arbitrary From: addresses, but not all of them
> do. Furthermore, allowing that is Considered Harmful. Sure, if you're
> using Unix you probably have sendmail and then you can set the From: to
> whatever you want (which probably isn't exactly right either, but that's
> a good enough oversimplification). If you are less fortunate, then the
> thing that sends email by default is Outlook, and it's probably not
> configured in a way that allows you to set the From:. And I hate to admit
> it, but that's probably a good thing.

I disagree on whether it's a good thing for a mailer to restrict the from
address--since the protocol does no such thing, it does nothing to harm
malware mailers or spammers, but can have the effect of locking in users to
a single email address, which rarely is desirable.  But in any case, I
think our conclusions are at least overlapping.  It should be documented
what --from is there for, and that it may or may not work, depending on the
mailer used.

If the MAPI protocol doesn't allow setting the from address, then we should
definitely remove the --from option when configured with MAPI.

> So, in my opinion it would actually just be best to get rid of --from or
> at least only allow it when sendmail exists, or something. For those of
> us who are unfortunate, --from just won't work, so it's misleading to
> have it as an option.

But for those of us who are fortunate, --from may be close to being a
necessity, particularly on computers that don't accept email (but may not
have knowledge of a valid return mail address to use).
-- 
David Roundy
Department of Physics
Oregon State University




More information about the darcs-devel mailing list