command-line sieve client that supports TLS

Andrew Morgan morgan at orst.edu
Tue Nov 14 13:48:34 EST 2006


On Tue, 14 Nov 2006, Phil Pennock wrote:

> On 2006-11-14 at 06:23 +0100, Phil Pennock wrote:
>> I just wrote one.  I'm going to bed without figuring out what I'm doing
>> wrong with GSSAPI
>
> Fixed.  I made assumptions about the zero-data-to-send case.
> Managesieve draft was unhelpful, I'll find the appropriate forum for
> that, but ktrace was my friend in the end.
>
> This also fixes the lack of final server verification in case the auth
> mechanism supports that -- ie, DIGEST-MD5.  So the previously posted
> version should be replaced with this one as it's a security risk to not
> be able to verify the server; STARTTLS should have done that for us, but
> a man-in-the-middle attack could have dropped that.
>
> Other changes are that some junk has been stripped out, a bug-fix in an
> error condition, and the password prompt goes to stderr to allow stdout
> to be captured by a script.
>
> Oh, and version numbers come from a Subversion repository -- this isn't
> really the 63rd version ;^) since the first commit of this script was
> v62.
>
> If anyone is seriously considering using this, please let me know and
> I'll clean up the action stuff and let scripts be supplied, just as for
> sieveshell, instead of the simple one-shot action stuff here, which was
> provided for embedded control where expect-like interaction isn't used;
> I decided to write for a simple API.
>
>> It's attached, and also on my website in case it gets mangled somewhere
>> in the mail path between me and you.
>
>   <URL:http://people.spodhuis.org/~pdp/software/sieve_connect-v63>
>
> # 14117 octets long
> MD5(sieve_connect-v63) =d8b1eb0f7e763d7fc661393da60e23bd
> SHA1(sieve_connect-v63) =03cc48babe3f13ef829a09a5322a276a494533eb
> RIPEMD160(sieve_connect-v63) =dddbbb77594653051d2f20a1627e9096fa3cc3c7

It would be great if this could be included with the Cyrus distribution. 
In the meantime, perhaps you could post it to the Cyrus Wiki?

Thanks,
 	Andy


More information about the Info-cyrus mailing list