RC>>> Dont know where you come up with this, since I am guilty of
 RC>>> the above configuration..

 ml>> which one? the "blackhole mailer" one or the
"dynamic routing and
 ml>> packing" one? i guess you mean the "blackhole" one?

 RC> I run both a mailer with a BSO and my tosser Squish does my
 RC> routing for me.

that's a totally "blackhole mailer" configuration... sorry, i
guess that the terminology hasn't tended to include the tosser...

 RC>>> if I removed this keyword from SQUISH.CFG

 RC>>> Killintransit

 RC>>> Netmail passing through here would remain in the *.MSG folder,
 RC>>> even after it was packed out and sent to its next hop.

 ml>> right and that is not desired for this operation...

 RC> The above keyword is in place and I have been reading netmail
 RC> on 123/500 and have seen intransit *.msgs appear momentarily..

momentarily... the question is, is this moment long enough for a PING
capable util to run thru the messages and perform the operation on them...

 ml>> now, if (and i'm not saying that it is) squish is one of those that
 ml>> skips the intermediate step of placing intransit netmail in a MSG
 ml>> area, then if (yes, another one) it also has a special config verb
 ml>> like "unpacktomsg" that would cause it to not skip
this step AND if
 ml>> there is a means of having it process inbound only, allow for the
 ml>> execution of other utils, process outbound only, then it may be
 ml>> possible for this to work as desired...

 RC> Squish and Gecho do not qualify as one of those that behave as
 RC> you fear when used with a BSO.

"do not qualify" as in they do unpack netmail in the PKTs to MSG
instead of PKT->PKT?

 ml>> does that all make sense? the keys being

 ml>> 1. all netmail must hit MSG format
 ml>> 2. the tosser not go ahead and also do the outbound processing
 ml>> 3. PING (and other utils) can be run against the MSGs
 ml>> 4. the tosser called for outbound processing (including packing since
 ml>> the intermediate tools have now been run)

 RC> I can do the PING thing accurately with my setup.

and the trace function works, too?

 RC> In fact I have requested Ward to send me his O/T template
 RC> files.

excellent... i had hope to see them in the proposal since the other steps
were there as well...

)\/(ark