I don't want to tempt fate but Tony's original message that I received was
only 300K or so - big enough and attachments are not a good idea as
Alison has pointed out.
I have just checked my maillog file and this is the only sizeable mail I
received all Saturday - certainly from Tony. There are no effective
limits on the various accounts via which my mail arrives - certainly not
as low as 22MB anyway.
So not everyone received the 22MB file! Don't beat yourself up Tony!
I suspect like Ted that it is Hillingdon NHS at fault. Perhaps their
members should be contacted and temporarily suspended until their server
behaves correctly. Such errors have already been noticed.
Best wishes
John
John Logsdon "Try to make things as simple
Quantex Research Ltd, Manchester UK as possible but not simpler"
[log in to unmask] [log in to unmask]
+44(0)161 445 4951/G:+44(0)7717758675 www.quantex-research.com
On Mon, 13 Nov 2006, Ted Harding wrote:
> On 13-Nov-06 Tony Greenfield wrote:
> >
> > I'm told that my message with the embarrassing 22MB has
> > now gone out for the third time.
> >
> > I am totally bewildered by this.
> >
> > I sent it only once.
> >
> > So: please tell me, how can this happen?
> >
> > Has a gorgon taken control of my computer?
> >
> > Tony
>
> Probably not -- though I can only offer a semi-guess (based on
> possibly relevant previous experience). [Last-minute edit: see
> at end for the almost certain resolution of the question]
>
> First, many people (such as certainly me and apparently Alison)
> will not have received the message on their email servers for a
> reason such as the following.
>
> In my case, this is because my account on nessie.mcc.ac.uk has
> a disk quota ("soft" usage limit, which I'm allowed to exceed
> for a limited time) of 10MB (and anyway I've used about half
> of this on stored files) and a disk limit ("hard" usage limit,
> which I'm never allowed to exceed) of 20 MB. So Tony's 22 MB
> will have been blocked. From what Alison wrote, the situation
> is probably similar for her.
>
> So far so good. The next question is: what happens to a mail
> which is blocked by someone's mail server, for this or any
> other reason?
>
> In most cases, perhaps a simple error message is sent back to
> the JISC mail server that is attempting to distribute the
> message. Provided this registers as a standard SMTP error
> during the sending process, that is probably the end of the
> matter.
>
> In other cases, however, perhaps the receiving mail server
> (misconfigured) sends the offending message back to the
> (or an) originating address, as if from the intended recipient.
> This could result in the message slipping back into the
> RadStats broadcasting system via the back door, with the
> effect that it gets sent out again, as a result of the
> JISCmail server itself not behaving as it should.
>
> Now of course that should not happen. But I have myself
> encountered problems due to misbehaviour of the JISCmail
> server. In particular this has happended when I have
> received an email from someone which I thought would be
> of interest to RadStats, so have forwarded it to the list
> (with some interpolated comment by myself).
>
> When one forwards an email, usually the main headers
> ("From:", "Subject", "Date:") are included in the *body*
> (i.e. the main text) of the email, along with any amendments
> that the forwarded may make.
>
> However, what happens with RadStats is that the JISCmail
> server reads the "From:" line in the body of the mail as
> if it were a "From:" header in the Headers section of the
> message, over-riding the true "From:" header that it will
> already have read. The result, in the cases where I have
> been bitten by this myself, is that the person who sent
> the message to me is erroneously identified as the sender
> to RadStats (whereas it was really me), and so receives
> a bounce on the grounds that they are not subscribed to
> RadStats.
>
> So my "theory" is that some people have received Tony's
> email more than once because the JISCmail server has been
> erroneously re-distributing mails bounced back to it by
> recipient servers which rejected the 22 MB mail in the
> first place, but which were themselves perhaps misconfigured
> so as to bounce it in the wrong way.
>
> I've just looked at the RadStats archives, and can see the
> three copies of Tony's messages. In each case, I see that
> there is a "CONFIDENTIALITY" footer apparently from the
> "Hillingdon PCT". My memory is that there has been trouble
> from that quarter (Hillngdon) before -- messages being
> incorrectly bounced back to the list, and incorrectly handled
> by the list when bounced back.
>
> I suspect that the problem at the JISCmail end is due to
> the same reason as I described above for forwarded messages:
> JISCmail's server reads a "From:" line in the body of the
> error mail as if it were the "From:" header, and therefore
> treats it as a valid message to be sent to the list.
>
> When I had that problem previously (about a year ago),
> the JISCmail admins were informed of, and acknowledged,
> the problem; but it doesn't seem to have been fixed.
>
> Thank heavens something seems to prevent this degenerating
> into a full-blown mail loop: JISCmail sends to Hillingdon
> who send it back quasi-to RadStats so it is sent back to
> Hillingdon who ...
>
> At 22 MB a kick this could saturate all available storage
> in short order!
>
> STOP PRESS:
> In fact, I've now just done a search on "Hillingdon" in the
> archives and found a message from myself about this very
> problem, on 15 Feb 2006,, after it had happened to John Bibby.
> I wrote:
>
> ==================================================================
> > [John Bibby]:
> > PS: Also I don't know who put the "all the information contained in
> > this message is legally privileged" stuff on - is this what happens
> > when 'Cuba' is in the title?
>
> Dunno about the £->L thing, but don't worry about the "legally
> privileged" thing, John.
>
> That's from Hillingdon PCT at hillingdon.nhs.uk which has a habit
> of bouncing messages back to the list. This happened recently to
> one of John Logsdon's messages amd to one of Robert Moore's (30 Jan).
> Inspection of the headers shows that they are delivered from
> the list to Hillingdon, and then returned from Hillingdon to the
> list as if from the original sender. Meanwhile, Hillingdon has
> inserted the legalese garbology.
>
> All best wishes,
> Ted.
> ==================================================================
>
> See
> http://www.jiscmail.ac.uk/cgi-bin/
> webadmin?A2=ind0603&L=RADSTATS&P=R480&I=-3
>
> [all one line] for the message from John Bibby with the
> "CONFIDENTIALITY" which he subsequently queried, and
>
> http://www.jiscmail.ac.uk/cgi-bin/
> webadmin?A2=ind0602&L=RADSTATS&P=R5104&I=-3
>
> for my response (quoted above) to his query.
>
> So don't worry, Tony -- I retract my earlier diffidence about
> my theory, and now feel pretty sure it's down to Hillingdon
> Primary Care Trust (aided and abetted by JISCmail).
>
> All best wishes,
> Ted.
>
> --------------------------------------------------------------------
> E-Mail: (Ted Harding) <[log in to unmask]>
> Fax-to-email: +44 (0)870 094 0861
> Date: 13-Nov-06 Time: 18:25:47
> ------------------------------ XFMail ------------------------------
>
> ******************************************************
> Please note that if you press the 'Reply' button your
> message will go only to the sender of this message.
> If you want to reply to the whole list, use your mailer's
> 'Reply-to-All' button to send your message automatically
> to [log in to unmask]
> Disclaimer: The messages sent to this list are the views of the sender and cannot be assumed to be representative of the range of views held by subscribers to the Radical Statistics Group. To find out more about Radical Statistics and its aims and activities and read current and past issues of our newsletter you are invited to visit our web site www.radstats.org.uk.
> *******************************************************
>
******************************************************
Please note that if you press the 'Reply' button your
message will go only to the sender of this message.
If you want to reply to the whole list, use your mailer's
'Reply-to-All' button to send your message automatically
to [log in to unmask]
Disclaimer: The messages sent to this list are the views of the sender and cannot be assumed to be representative of the range of views held by subscribers to the Radical Statistics Group. To find out more about Radical Statistics and its aims and activities and read current and past issues of our newsletter you are invited to visit our web site www.radstats.org.uk.
*******************************************************
|