On Thu, Jun 05, 2014 at 04:49:58PM +0100, Jeremy Harris wrote:
> On 05/06/14 16:40, Viktor Dukhovni wrote:
> >On Thu, Jun 05, 2014 at 04:17:08PM +0100, Git Commit wrote:
> >
> >>http://git.exim.org/exim.git/commitdiff/315e209ab63a4d198d4592247cabf52b9401844d
> >>Author: Jeremy Harris <jgh146exb@???>
> >>AuthorDate: Thu Jun 5 15:16:29 2014 +0100
> >>Commit: Jeremy Harris <jgh146exb@???>
> >>CommitDate: Thu Jun 5 15:16:29 2014 +0100
> >
> >I think this code is too ad-hoc to pass review. It assumes particular
> >formatting of output for human consumption. The original code that
> >attempted to map the signature algorithm OID to its name was better,
> >and should be repaired not replaced with a mess.
> >
>
> What formatting of output would you like to see?
I was trying to say that tls_cert_signature_algorithm() should if
possible not parse the output of X509_print_ex(). The simplest
solution is to return the OID, rather than trying to convert it to
a human readable string, which is rather compex for some of the
fancier algorithms.
i2a_ASN1_OBJECT() converts ASN.1 objects to the usual dotted notation.
--
Viktor.
