Hi Paul,

On 04 Sep 17 18:46, Paul Quinn wrote to Wilfred van Velzen:
  about: "Trialing FMail/lnx status check-in":

 PQ> FMail/lnx is wonderful, Wilfred.  Thank you for letting me try it out.

Thank you for testing it! 

 PQ> What was what I call 'Phase 1' testing ceased for me at about midday on
 PQ> Saturday.  Test packets originated from -this- node as received from my
 PQ> current peer linked Fidonet systems.  They were shoved otherwise-untouched
 PQ> to my test system via binkD, and ingested by FMail.

 PQ> With FMail, I have enjoyed seeing _full_ inter-zone SEENBY & PATH data in
 PQ> messages.  These items form the singular function that's caused me to try
 PQ> FMail.

 PQ> FMail was under rigorous supervision for about three weeks of 24/7
 PQ> operation.  I think there's about three Mb of logfiles with nary a glitch
 PQ> reported, IIRC.

(I had to look up "nary" )

 PQ> It did turn up that weirdness with message sorting that I was later to
 PQ> find was subject to what Uncle Google & Wikipedia tells me is the
 PQ> 'Observer effect'.  After I reworked the data originating from FastEcho,
 PQ> whose packet size limits dated from the POTS days, all was well.

 PQ> Interestingly, a couple of AreaMgr requests were replied to but the
 PQ> responses were not sent.  I guess that could be the case with potential
 PQ> PING responses as well.  Is that normal?  (On my main node [~/384] I would
 PQ> have a BATch file check for any *.Msg files in the netmail path, since the
 PQ> primary netmail area is usually empty, and would attempt a PACK if any
 PQ> such mails existed.)  How do you recommend they be completed by FMail if
 PQ> it is the only configured netmail area?

Do these mails have the 'Crash' flag set? Crash mail needs to be packed with
"special" command line options:

  .../fmail pack '*' '-c'

Where '*' can be replaced by any fidonet address, or match pattern, if you want
 to just pack netmail for specific destinations. E.g. '3:640/*'.

And make sure to put the pattern between single quotes, otherwise your shell
might expand it to filenames!

 PQ> The 'Phase 2' test occurred later on Saturday and lasted for about 10
 PQ> seconds or less.    It's premise was simply continuing function with a
 PQ> pre-existing messagebase, from that morning's backup from this node.  I
 PQ> did allow the *.Msg, HMB areas & dupe ID database to remain as they were.
 PQ> FMail cruised through the test without even a grunt.

 PQ> On Sunday I was surprised and pleased by the 'badmail' capability to
 PQ> detect old mails... it found one!  (It was a phurphy from region 18 IIRC,
 PQ> and was a re-badged old mail from NZ.)  Did yours?  ;-)

There was some regurgitated echomail originating from 1:15/0, partly in wrong
areas. But none in my bad area, because none were older then my 60 day setting
for old mail... And none in a NZ_* area...?

 PQ> Now 'Phase 3'.  I'm thinking of doing some sort of export SCANning of
 PQ> outbound mails.  I think I need something further from you?  Do you have
 PQ> something of a 32-bit flavour, please?  

The fmail and ftools I gave you are 32-bit, so I don't understand?

Wilfred.

--- FMail-W32 2.0.1.4
 * Origin:  point@work  (2:280/464.112)