Hi Laura,
The situation you describe concerning the Winsock error remains unchanged
from as I have previously described it. If your BrailleNote stalled, for
instance, after the download of a particular message and, as an example,
the server gives up on you, you get this error at the next moment your
BrailleNote makes a send request for the next command. The point at which
the BrailleNote appears to do its disk operations is immediately following
the mail or partial mail download when the message is new, as described in
my algorithm.
The question of how and when the BrailleNote does message duplication
detection is addressed in the same lengthy message. It is unfortunately
not reliable detection - it uses exclusively a message field called the
Message-ID, which is intended to uniquely identify a particular instance
of the particular message, but which is an optional field and hence is
not, but usually is, found necessarily in any given mail. If RFC 2821's
suggestion for a mail system to add a Message-ID field where one is absent
is not followed, then messages, including those from the BrailleNote which
does not generate its own Message-ID, will be downloaded again, whatever
the circumstances. This is not related to the Winsock error in any way,
except perhaps that since mail is only ever written on disk when found to
be new, it seems more likely that such an error will occur when
downloading new mail - new mail which you have not downloaded before or
which has no Message-ID field (in the latter case, a duplicate results).
I will finally add that my original test mails, whose sizes I still have,
were all under 5 kilobytes in size - hence, as you know, the regardless
duplicate download. Since they were all generated on my BrailleNote, and
since the Message-ID is not added by my mail system for (in my opinion)
good reason, the BrailleNote would download all messages from all
BrailleNote users, regardless of size (an awful lot of my mailbox's
contents) any number of times I retrieved mail. It was only my good
fortune that the statistics of the server showed, a week later, an unusual
pattern of data transfers that were excessive on the BrailleNote, compared
to downloads made by my PC for approximately the same number of messages
with approximately the same number of bytes (these would, in fact, be the
mail I'd downloaded on my BrailleNote but not deleted in the morning that
I would then fetch again later on my PC in the evening). This of course
lead to a conclusion which I wrongly gave previously.
Few! Well, are we all clear now? I hope you got something from this,
because without burrowing into the unnecessarily boring primitives of
TCP/IP and the SMTP protocol (did I really just say that? <Grin>) I'm
afraid that's the best I can do for you for now.
Cheers,
Sabahattin
--
Thought for the day:
Advertising (n): the science of arresting the human
intelligence for long enough to get money from it.
-- Stephen Leacock.
Sabahattin Gucukoglu
Phone: +44 20 88008915
Mobile: +44 7986 053399
http://www.sabahattin-gucukoglu.com/
Email/MSN: <[EMAIL PROTECTED]>
