ned.freed@innosoft.com
Mon, 24 Jan 2000 11:24:40 -0800 (PST)
> I didn't make my point clearly, for which I apologize.
> I was not commenting on the reliability of fax, but on the dangers of
> relaxing the need to report a problem. Fax lacks many of the
> mechanisms to detect or evaluate the seriousness of a particular
> impairment, to be sure. I also agree that users will have vastly
> different interpretations of any reported problem.
> That should not remove the responsibility to report any difficulty
> encountered during transmission.
My point is that while I agree with you from a standpoint of abstract standards
and protocol development, my experience with this in practice with real world
products doesn't agree at all. Far from liking it when absolutely all
difficulties were reported, our users disliked it so much that our choice was
simple: Either disable reporting of problems below a certain threshhold or else
lose many of our customers.
And if the behavior of most of the FAX machines I use is any indicator, their
makers have reached the same conclusion we did.
> I was reacting to the implied
> suggestion that partial delivery be treated as completed delivery.
Well, the delivery most certainly has been completed to the extent it ever
will be, so I don't have a problem with calling it "completed". I think
you meant to say "complete". If so, please remember that the structure
of delivery reports is such that we have to assign a success/failure indicator.
We simply don't have a choice in the matter. And if we elect to assign a
"fail" indicator to any problem whatsoever, we run smack into all the
problems I've already described.
> Failing a list of possible impairments and combinations of
> impairments, against which we could assign a "pass/fail" parameter,
Even assuming we could reach agreement on such thing, which I doubt, such a
list would have to be so context-sensitive that its size would explode in a
combinatoric fashion. And it would have to be constantly updated to be useful.
In short, while I understand the motiviation and to some extent agree with it,
I don't think such a list is viable.
> I am in favor of full reporting. / mark
Believe me, I'd like to be, but without certain guarantees from the user agents
in regards to information correlation I don't think it is practical to do this.
Ned
This archive was generated by hypermail 2.0b3 on Mon Jan 24 2000 - 21:40:21 IST