Patch: Offline transfer mode

[Steve Traugott]

On Wed, Mar 23, 2005 at 05:41:40PM -0800, Wayne Davison wrote:
> On Wed, Mar 23, 2005 at 03:16:17PM -0800, Steve Traugott wrote:
> > And then writefd() would need to be toggled from there to write to the
> > correct fd...
> 
> On the receiving side, everything that is read from the sender is
> already sent to the batch file, so I think the only thing necessary is
> to have the code call discard_receive_data() instead of writing out the
> data to a temp file.

Whups -- yes, I'd forgotten that readfd() takes care of things for us.

> > Yep, this would all be much simpler if I just went ahead and
> > refactored writefd().
> 
> I'm thinking that's not necessary.  It should be enough to have the
> sender just write the index/update data directly to the batch fd without
> needing to change the routines in io.c (except to have a way to get the
> value of the batch fd).

Do you mean just set f_out to batch_fd at the right time in send_files,
then set it back before protocol shutdown?  I don't know why I didn't
see that before.  Looks like I'd still need to play with
write_batch_monitor_out to get writefd() to call write() instead of
letting it stray into writefd_unbuffered() or the iobuf stuff...

Or were you thinking of somehow avoiding writefd() altogether?

> (I'll comment on some of the other parts of your email later.)

Uh oh.  ;-)

Steve
-- 
Stephen G. Traugott  (KG6HDQ)
UNIX/Linux Infrastructure Architect, TerraLuna LLC
stevegt at TerraLuna.Org 
http://www.stevegt.com -- http://Infrastructures.Org
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
Url : http://lists.samba.org/archive/rsync/attachments/20050323/3b74dc74/attachment.bin