Re: [Gnu-arch-users] Re: darcs vs tla

[Karel Gardas]

On Mon, 15 Nov 2004, Dustin Sallings wrote:

>
> On Nov 15, 2004, at 3:45, Karel Gardas wrote:
>
> > Darcs' cons:
> >
> > - darcs does not support patch signing (showstopper for me)
>
>       That's not *exactly* true.  darcs does support having signed patches
> on send when using the following flags:

You are right! I've completely omited this since in comparison with tla's
support it looks quite weak. i.e. tla support signing/verification for
every transport AFAIK, at least I'm using sftp/http and
signing/verification works well for them. So I should rather rewrite my
original statement to:

- darcs does not support patch signing/verification except simple
  email-based transport method (showstopper for me)

>       While the signature isn't recorded in the patch itself or your working
> tree (although the patch name includes the sha1), it can be validated
> if you keep the source of the patch (i.e. an email archive if you're
> using email for transport).

Yes, I will probably need to go to darcs mailing list to find out why
signatures are not part of patch and why it does not verify automatically
when they are presented (while using apply command), i.e. currently it is
quite easy to apply unverified patches just by mistake IMHO.

Anyway thanks for the hint!

Karel
--
Karel Gardas                  address@hidden
ObjectSecurity Ltd.           http://www.objectsecurity.com