🚀 go-pugleaf

On Sunday, in article <838588137snz.4950@grempc.demon.co.uk>
     Graham@grempc.demon.co.uk "Graham E Kinns" wrote:

GK> John Washington <john@wash.demon.co.uk> wrote:
..
GK> >No I haven't.  That was precisely why I posted it, but I thought
GK> >it was glaringly obvious!

You arrogant sunnavabitch, Washington.  In view of what
follows I'm unqualified to say _anything_ is obvious!

...
GK> >If you mean line breaks (\n) are inserted, I disagree.
...
[GK's correction snipped]

Damn!  My foot with .45" hole in it [waves smoking gun].

Thanks Graham, I (now) agree with you completely.

On Saturday, in article
     <Snews.960727.044016.chris.315@keris.demon.co.uk>
     chris@keris.demon.co.uk "Chris Croughton" wrote:

CC> In article <25734@wash.demon.co.uk>
CC>            john@wash.demon.co.uk "John Washington" wrote:
CC> 
CC> >This is d.i.developers, not d.i.newusers, so I expect we've all
CC> >read the relevant RFCs.
CC> 
CC> Having a bad day, John?  This isn't like you normally...

<panto>  All together, oh yes it is.  :-)

On Saturday, in article
     <Snews.960727.044840.chris.316@keris.demon.co.uk>
     chris@keris.demon.co.uk "Chris Croughton" wrote:

CC> In article <838398667snz@wash.demon.co.uk>
CC>            john@wash.demon.co.uk "John Washington" wrote:
CC> 
CC> >And meanwhile I've confirmed this is indeed the cause of my
CC> >problem.  In fact the 900K mail was a red herring.  Another
CC> >much shorter mail (from jobsite) contained a line of precisely
CC> >257 characters (including the final \n) and the last char
CC> >before the \n was '.'.
CC> 
CC> Can you post your patch, please?  I'm happy with straight diff(1)
..

All I did was move the int no_check_dot = 0; up to the start
of the function, so it's outside the while (as I think someone
else suggested).  Actually I also changed the buffer length
from 256 to 300, just out of curiosity, and added a diagnostic
to tell me when lines over 253 bytes came in (why 253?  why not?).
And I find that I'm getting some mail with lines exceeding 300,
even.  And I also have a diagnostic that does a tprintf every 1000th
line of the incoming mail, so I can get an idea of whether it's
worth waiting a little longer for a long mail, or leave it 'til
next time.  If any of that interests someone I'll mail it.

But, as Graham has pointed out, it still needs a little more than
just moving the decln, since the lines will still be split
with a spurious newline.

..
CC> >always imagined that multiple servers were only handling
CC> >the mail from different punts, and that each punt took
CC> >mail only off the head of its queue.  Not so.  BRILLIANT,
CC> >Demon, thanks.
CC> 
CC> Fascinating, I'd assumed that as well.  Presumably this starts up after
CC> the 15 minute or whatever retry timer?

I've no idea.  I really thrashed my system by putting in an
smtplog of every incoming line while investigating the problem,
so my 900K mail took well over an hour to trickle in.  The
mail behind it started coming in at some point when I'd gone
off for a coffee or two.

..
CC> I can send you a scanned picture of a fiver if you like <g>.  Well, I
CC> could if I had a scanner <g>...

Ta.  It's probably the best offer I'm going to get!

--
John Washington