> We also discussed that a server MAY keep a list of clients that are t=
rying
> to connect in a "queue", and then serve each one one at a time.
This is easy to implement, but the connection attempt will appear to "h=
ang"
indefinitely.
> Then the
> client doesn't receive an error (except if the "queue" is filled). T=
his
> gives the end-user a much happier experience.
I'd be happier to get server-error-service-unavailable (0x0502) with an=
estimate of the the length of the delay indicated in the message. A cl=
ient
could then give a user the choice of canceling, retrying, or queuing lo=
cally
and retrying after delay. At that point the user might decide to try a=
nother
printer, or just queue the job locally (client side) and go on.
> I think that both approaches should be put into the FAQ.
Fine with me.
-Carl
hastings at cp10.es.xerox.com on 02/06/98 07:34:46 PM
Please respond to hastings at cp10.es.xerox.com @ internet
To: ipp at pwg.org @ internet, Carl Kugler/Boulder/IBM at ibmus
cc:
Subject: Re: IPP> IPP > FAQ - How should the server behave?
At 08:10 02/02/1998 PST, Carl Kugler wrote:
>Henrik-
snip...
>> 3. How should a non-spooling IPP-server handle concurrent print-jo=
b
>> requests?
>>Return server-error-service-unavailable (0x0502) to indicate that the
server is
>temporarily unable to handle a request.
>>> -Carl
>>---------------------- Forwarded by Carl Kugler/Boulder/IBM on 02/02/9=
8 08:41
We also discussed that a server MAY keep a list of clients that are try=
ing
to connect in a "queue", and then serve each one one at a time. Then t=
he
client doesn't receive an error (except if the "queue" is filled). Thi=
s
gives the end-user a much happier experience.
I think that both approaches should be put into the FAQ.
Tom
=