>Jay Martin wrote:
>Is IPP expected to support checkpointing to the point where the
>client will resume submission of document at *precisely* the point
>where the transmission failed?
>I wouldn't think so. Rather, the client would resume transmission
>at the *start* of the document that failed.
>Is this true?
There are several places you could restart:
1) The entire job (not adequate for large production
2) The document that failed (also not adequate for
large production jobs. Although there was some
argument that large jobs could be broken up into
multiple documents, I doubt that we want IPP to
have that kind of impact on existing production
applications. At least I'm not willing to go and tell
my customers that they have to rewrite their apps.)
3) The last Send Document - this doesn't seem
too difficult, especially with sequence numbers
to identify each segment. This also seems like a
rational boundary to build checkpoints on.
4) bytes within a Send Document - much harder to do,
requires some dialogue between the client and the
server to synchronize data.
5) You could also restart on a page boundary, but
this is really much harder to do.