Let's say that a line monitor somewhere in the world decides to hold a file
for a long period of time (eg until the next weekend), or to split it into
small chunks on which he furthermore runs a complex encoding program that
reduces the chunks to 1/3rd of their original size, but then 1 damaged bit
(base RSCS V2 *grin*) means the file is turned into a jelly-fish. Let's say
that the postmaster asks LISTSERV to send notification to the file owners
about this, just in case they might think that their file has been lost and
they'd wish to resend it.
The questions is: should we inform the SENDER, the RECIPIENT, or BOTH?
1. S&R are human people. You ought to inform the sender, because otherwise he
might think the file has been lost and might resend it. You ought to inform
the recipient, because otherwise he might get anxious and that's not good
for his peace-maker. But that's less important, the sender might inform him
if he starts asking "you sure you DID send that darned file??" This is
quite common. Score: S=1/0, R=0.5/0, F=1 (copy to Sender: pro/con, id for
recipient, Frequency).
2. S=server, R=human. Then you must only inform R, who might get anxious and
order a second copy, but not S, who might get upset and tell you to try
HELP, yeah, you might be more lucky with a HELP command. This case is quite
common (although servers rarely send large files, MAILERs tend to). S=0/1
R=1/0 F=1
3. S=human, R=server. For the same reason as above R should not be informed,
and S should. But S is supposed to be somehow intelligent (idiots don't
send files to servers, esp not large ones) and will check the links for his
file. And that case is pretty rare anyway, so S=0.5/0 R=0/1 F=0.25
4. S&R are servers who won't get upset about network delays. S=R=0/1, and
since this is VERY rare, F=0.1.
Final scores:
S=1 Pros 1.125 Cons 1.1
R=1 Pros 1.5 Cons 0.225
Do you agree with this analysis, and what would you decide about sending the
notification to the sender or not?
Eric
|
