🚀 go-pugleaf

I was unable to read this message, as I automatically kill all 
messages longer than 200 lines.

However, from the Subject line this appears exactly like a problem I 
had, reported here some weeks ago, but with no response. Essentially, 
in a dial-up system, Mercury/32 would dial and then crash with 4 
successive messages. Sometimes a few messages would be delivered.
Once this happened, it would recur every time.

It was associated with particular messages. I think the common factor 
was large numbers of Cc: addresses. When it firts happened, I moved 
all the .MDD files to a separate directory where Mercury/32 couldn't 
find them; next session worked fine. I then moved half of them back,
and moved them around in a binary chop way until I identified the 
message causing the problem (if you have 1024 potential suspects, you 
can identify the culprit in just 10 tries this way). In my case, all 
messages from this particular sender caused trouble, and no others; 
so when the problem recurred I knew what message to seek and destroy.

I sent David Harris a copy of the culprit messages, so the problem 
may be solved in the latest Mercury/32 (2.16); having worked round 
the problem, I haven't yet tested whether it is actually solved.

Best wishes,
-- 
Michael Salem