This is the mail archive of the
gnats-devel@sources.redhat.com
mailing list for the GNATS project.
RE: gnatsd output question
- To: Marcus Williams <marcus at quintic dot co dot uk>
- Subject: RE: gnatsd output question
- From: Rick Macdonald <rickm at vsl dot com>
- Date: Wed, 5 Sep 2001 06:59:11 -0600 (MDT)
- cc: GNATS Development Mailing List <gnats-devel at sources dot redhat dot com>
On Wed, 5 Sep 2001, Marcus Williams wrote:
> I'd tend to agree with Rick Macdonald that the lock messages are
> internal to Gnats so dont need to be displayed (unless they fail).
>
> It may be worth doing something like the FTP specs (RFC 0959?) say and
> wrap the three messages inside a multiline result so it becomes (from
> what I remember of the RFC):
>
> subm
> 211 Ok.
> .
> 210- GNATS database is now locked
> 402- Failure reading header
> 210- GNATS database is now unlocked
>
> The trailing dash after the result code indicates a multiline
> continuation. However, this is wrong as well as we're only interested
> in the 402 in the middle so it probably should be
>
> subm
> 211 Ok.
> .
> 402- GNATS database is now locked
> 402- Failure reading header
> 402- GNATS database is now unlocked
>
> Since you can only tell that theres a header error by reading the full
> message, maybe more error codes should be added so then we can
> exchange 402 for whatever the code is for "Failure reading header".
> The multiline reply is then only for interactive/information purposes.
Well, the multiline is better than three separate ones but I still would
vote to not show the lock/unlock at all. If a submission fails because the
internal lock fails, there should still be just one line. Something like
402 - Failure locking GNATS database
Think about a successful submission. Nobody needs to see
210- GNATS database is now locked
210- PR added
210- GNATS database is now unlocked
...RickM...