Come on Demon, admit that email is broken
David Rule
thing in 12 hours. The MOTD is just laughable as a couple of earlier
threads have proven.
Are Demon pretending the problem isn't there so that they don't have to
admit to their customers that they aren't able to deliver the service
that their customers are PAYING for?
FACT: It is broken.
I really don't care why or who is to blame, I just want Demon to
say "yes, alright it's broken, we're working on it, here's an ETA."
BTW, if you're bored and need entertaining, just ring the helpdesk and
listen to the prerecorded excuse given to discourage you from
complaining to a support person.
David
(A customer of 7 years standing)
Sent via Deja.com http://www.deja.com/
Before you buy.
bobmc...@my-deja.com
Judging by the volume of emails I am receiving on my account your
statement is most decidely FALSE!!!
The volume may be lower than normal today, but the status does give an
update on this situation, and Demon is not the only ISP having email
problems, some ISP's were totally out the game with emails for a couple
of days recently.
Tony Evans
>BROKEN!!
>
>Judging by the volume of emails I am receiving on my account your
>statement is most decidely FALSE!!!
No, his statement is true. Demon's mail system is broken. It's
failing to deliver any of the mail I know I've been sent today, other
than 10 token mails it thought I might like.
One mail took over 12 hours to go from clicknames.net to the demon
system late yesterday.
Just because it's managing to deliver your mail doesn't mean it's
working. When it delivers all the mail, then it's working.
Tony Evans
--
Master of all I survey (at the moment, empty pizza boxes)
Gemmell Mania : http://www.gemmell-mania.org.uk | IRC : Blade
UK LRP Web Ring : http://www.darkstorm.org/uklrp | ICQ : 170850
got a UK related LRP site? Join the ring today ...
John Malone
>BROKEN!!
>
>Judging by the volume of emails I am receiving on my account your
>statement is most decidely FALSE!!!
>
In Message-ID: <8hrkm1$h8r$1...@nnrp1.deja.com> David Rule
<da...@therules.net> wrote
>
>
>> When is the email problem going to be fixed? I haven't received a
>> thing in 12 hours. The MOTD is just laughable as a couple of earlier
>> threads have proven.
It's a wee bit ill from here. Maybe you're just lucky (bobmcintyre).
I collect mail in Linux mode via SMTP and nothing came through today.
When I switched to W95 and Turnpike POP3 this evening it was OK.
I have noticed often recently that one or two mails get stuck on the
punts, finger me@post... shows it just sitting there.
BTW
I was talking to a guy on the helpdesk today and he pronounced "Linux"
correctly !!!
Maybe there's hope :)
--
jo...@malone1.demon.co.uk | Lurk and Learn
Tony Evans
>Are Demon pretending the problem isn't there so that they don't have to
>admit to their customers that they aren't able to deliver the service
>that their customers are PAYING for?
The helpdesk have just confirmed to me that engineers are currently
looking at the mail system, and that it's still got problems.
I asked them to update status@gate since it's now wrong, and around 6
hours old.
Mail from my yahoo account, send within the last hour still isn't
arriving.
I feel sorry for the call centre, they must be getting a large number
of calls from people because of the incorrect nature of status@gate,
that could be avoided.
Oh, what did you say? 0845 number? Who gets the cash for that one?
Tony Evans
--
A man with a watch knows what time it is. A man with two watches is never sure.
Gemmell Mania : http://www.gemmell-mania.org.uk | IRC : Blade
Got some Fantasy reviews on-line? Why not | ICQ : 170850
join the Fantasy Book Review Webring! http://www.darkstorm.org/fbrr
David Rule
know of plenty, including ones I've sent myself. Stuff sent to
@resanal.demon.co.uk is getting through but slowly.
Does Demon have something against 3rd party mail redirection services?
David
On Fri, 09 Jun 2000 22:49:19 +0100, Tony Evans <to...@darkstorm.co.uk>
wrote:
>In demon.service, bobmc...@my-deja.com wrote:
>
>>BROKEN!!
>>
>>Judging by the volume of emails I am receiving on my account your
>>statement is most decidely FALSE!!!
>
David Rule
what I was after more than anything, an admission. I get upset when
MOTD or status@gate say that things are fixed and indeed a recorded
message at the helpdesk say they're fixed when clearly they aren't.
Things go wrong, I accept that, just be honest about them please
Demon!
:-)
David
On Fri, 09 Jun 2000 23:50:11 +0100, Tony Evans <to...@darkstorm.co.uk>
wrote:
>The helpdesk have just confirmed to me that engineers are currently
Tony Evans
>I've not received a single email forwarded from netnames.co.uk and I
>know of plenty, including ones I've sent myself. Stuff sent to
>@resanal.demon.co.uk is getting through but slowly.
I was worried about that initially, but a very quick response from
sup...@clicknames.net suggested it wasn't a problem with them, and
some more test mails indicated that mail appeared to be getting
through randomly.
It appears to have dried up totally now however - which is annoying -
I'm awaiting some rather vital information for the first time in
weeks.
Typical, normally e-mail is junk and arrives on time, now it's vital
and it's just not working :(
Tony Evans
--
I'm not a complete idiot--several parts are missing.
Gemmell Mania : http://www.gemmell-mania.org.uk | IRC : Blade
Like David Gemmell? Why not join his mailing list | ICQ : 170850
mailto:GemmellFanta...@onelist.com
Tony Evans
>I collect mail in Linux mode via SMTP and nothing came through today.
>When I switched to W95 and Turnpike POP3 this evening it was OK.
Hmm, there's nothing at all waiting for me at Demon's end, regardless
of how I would like to collect/receive it.
O.Dear
>
>Judging by the volume of emails I am receiving on my account your
>statement is most decidely FALSE!!!
Well I'm using two accounts to access the same mail at the moment (kind
of a transition thing). My Freeserve account has received about 50 mails
in the past four hours, Demon have managed to deliver about 30 of these.
--
O.Dear
Tony Evans
>I asked them to update status@gate since it's now wrong, and around 6
>hours old.
Still no update :(
Tony Evans
--
I like my steak so rare, a good vet could save it.
Gemmell Mania : http://www.gemmell-mania.org.uk | IRC : Blade
Homepages : http://www.darkstorm.org | ICQ : 170850
Currently Reading : The Shannow Trilogy - David Gemmell
Tony Evans
>Are Demon pretending the problem isn't there so that they don't have to
>admit to their customers that they aren't able to deliver the service
>that their customers are PAYING for?
Latest comment from the helpdesk,
1. yes, sorry, we will update status@gate honest
2. yes, sorry, we might update the recorded message on the helpdesk
3. we had a DOS attack, we have no idea when things will be working
Tony Pay
to...@darkstorm.co.uk (Tony Evans) wrote:
> 3. we had a DOS attack, we have no idea when things will be working
As a fellow sufferer, but an ignorant yet curious one, could you tell me
what a DOS attack is?
Tony
--
_________ Tony Pay
|ony:-) 79 Southmoor Rd To...@stsm.demon.co.uk
| |ay Oxford OX2 6RE www.gmn.com/artists/welcome.asp
tel/fax 01865 553339
... Macho Law forbids me from admitting I'm wrong.
Graham Murray
> As a fellow sufferer, but an ignorant yet curious one, could you tell me
> what a DOS attack is?
Denial Of Service
- - -
Roger Jones
news:20000610....@stsm.demon.co.uk...
> As a fellow sufferer, but an ignorant yet curious one, could you tell me
> what a DOS attack is?
Denial of Service
--
Regards
Roger
Richard Tibbetts
>In article <1312F6213A36...@news.darkstorm.co.uk>,
> to...@darkstorm.co.uk (Tony Evans) wrote:
>
>> 3. we had a DOS attack, we have no idea when things will be working
>
>As a fellow sufferer, but an ignorant yet curious one, could you tell me
>what a DOS attack is?
Denial Of Service.
--
Richard Tibbetts
http://www.primepeace.ltd.uk/
Tony Evans
>In article <1312F6213A36...@news.darkstorm.co.uk>,
> to...@darkstorm.co.uk (Tony Evans) wrote:
>
>> 3. we had a DOS attack, we have no idea when things will be working
>
>As a fellow sufferer, but an ignorant yet curious one, could you tell me
>what a DOS attack is?
Denial Of Service
An attack which in some way attempts to prevent a machine/group of
machines from doing what it is they do.
In this case for example, people might be sending many hundreds of
thousands of e-mails into demon's system, or just sending it loads of
useless tcp/ip traffic, or any other number of things. The end result
is that the mail servers are so busy serving the 'attack' they can't
service real requests.
I don't want Demon to magically 'fix it', I just want them to keep us
up-to-date with what's going on :(
Tony Evans
--
Lord, grant me the serenity to accept the things I cannot change, the courage to change the things I can, and the wisdon to hide the bodies of the people I had to kill because they pissed me off- Anon.
Toby Groves
<to...@darkstorm.co.uk> writes
>No, his statement is true. Demon's mail system is broken. It's
>failing to deliver any of the mail I know I've been sent today, other
>than 10 token mails it thought I might like.
>
>One mail took over 12 hours to go from clicknames.net to the demon
>system late yesterday.
From the Clicknames status page:
"09/06/00 : 16:48 An issue with Demon's mail service has meant
Clicknames mail servers have been holding mail for Demon customers. It
is our understanding Demon have resolved their problems, but queued mail
may take a while to catch up. Demon expect all mail to be caught up by
tomorrow."
--
Toby
Glenys Pople
<to...@darkstorm.co.uk> writes
>I don't want Demon to magically 'fix it', I just want them to keep us
>up-to-date with what's going on :(
I agree. I would by now have expected 150+ more emails than I have
received over the last day. Most would be from mailing lists but I
expected two private emails that were important and urgent.
I, too, would like to know what is going on.
--
Glenys
Tony Evans
>In article <1312F613472B...@news.darkstorm.co.uk>, Tony Evans
><to...@darkstorm.co.uk> writes
>>No, his statement is true. Demon's mail system is broken. It's
>>failing to deliver any of the mail I know I've been sent today, other
>>than 10 token mails it thought I might like.
>>
>>One mail took over 12 hours to go from clicknames.net to the demon
>>system late yesterday.
>
>From the Clicknames status page:
>
>"09/06/00 : 16:48 An issue with Demon's mail service has meant
>Clicknames mail servers have been holding mail for Demon customers. It
>is our understanding Demon have resolved their problems, but queued mail
>may take a while to catch up. Demon expect all mail to be caught up by
>tomorrow."
I know, that went up after I mailed sup...@clicknames.net to ask if
the problem was at their end. The 'Demon expect ...' bit is becoming
increasingly silly considering the service is still broken.
Tony Evans
--
Pizza: Nature's perfect food.
Gemmell Mania : http://www.gemmell-mania.org.uk | IRC : Blade
Tony Evans
>In article <1312F62192D...@news.darkstorm.co.uk>, Tony Evans
><to...@darkstorm.co.uk> writes
>
>>I don't want Demon to magically 'fix it', I just want them to keep us
>>up-to-date with what's going on :(
>
>I agree. I would by now have expected 150+ more emails than I have
>received over the last day. Most would be from mailing lists but I
>expected two private emails that were important and urgent.
Yep, egroups is now reporting the bounces from my demon address (I am
still subscribed to those lists using tagman.dcu, rather than my
darkstorm.co.uk domain, so there's no forwarding issues involved).
However, the most painful thing was waiting over 6 hours to find out
if my web pages had been migrated correctly in the states and what the
new password for the account was :(
>I, too, would like to know what is going on.
Yup, status@gate still wrong, even though two different support folk
said they would have it updated.
Tony Evans
--
Fame: Chiefly a matter of dying at the right moment.
Are you a member of an LRP club or group? | IRC : Blade
Want a couple of web pages and an e-mail address | ICQ : 170850
for free? http://www.darkstorm.org/lrpclubs/offer.htm
David Rule
still status@gate suggests there's not really much of a problem. No
comment from any of the Demon techs in this newsgroup either. Any
conspiracy theorists out there like to start speculating? :-)
David
On Sat, 10 Jun 2000 12:17:03 +0100, Tony Evans <to...@darkstorm.co.uk>
wrote:
>In demon.service, Glenys Pople <g...@howfen.demon.co.uk> wrote:
Ben Smithurst
> BROKEN!!
>
> Judging by the volume of emails I am receiving on my account your
> statement is most decidely FALSE!!!
No, it's not. Maybe it's not completely broken, but it's near enough at the
moment to give that statement justification.
ben@scientia:/var/log/exim/OLD$ zgrep -c "<= .* H=punt-...mail.demon.net" \
main.*.gz
main.01.gz:732
main.02.gz:726
main.03.gz:565
main.04.gz:744
main.05.gz:356
main.06.gz:398
main.07.gz:641
main.08.gz:571
main.09.gz:528
main.10.gz:507
main.11.gz:765
main.12.gz:451
main.13.gz:339
main.14.gz:499
ben@scientia:/var/log/exim/OLD$ grep -c "<= .* H=punt-...mail.demon.net" \
main.00
100
go figure. scientia.dcu normally receive 500-700 messages on weekdays,
it only got 100 yesterday. I'm glad Demon have at least acknowledged
the problem, but I'd like to see more updates, specifically what the
delay in fixing it is. More importantly, I'd like to see if fixed.
(obviously :-)
--
Ben Smithurst / b...@scientia.demon.co.uk / PGP: 0x99392F7D
David Rule
Shame it was only from the Demon helpdesk and not any of the stuff
that's been queued up for the last day and a half.
Anyway, the line I am being given is that there is a "backlog." I'll
say! Have replied asking for an estimate as to when the "backlog" is
expected to be cleared so that should I still not be receiving my
email by then, I can escalate the problem further.
Will keep you posted.
David
Tony Pay
only to be expected? Do others find that?
Tony
--
_________ Tony Pay
|ony:-) 79 Southmoor Rd To...@stsm.demon.co.uk
| |ay Oxford OX2 6RE www.gmn.com/artists/welcome.asp
tel/fax 01865 553339
... Drive C: Error, (A)bort (R)etry (I)gnore (K)ick (S)cream
Denis McMahon
wrote:
>In article <1312F6213A36...@news.darkstorm.co.uk>,
> to...@darkstorm.co.uk (Tony Evans) wrote:
>
>> 3. we had a DOS attack, we have no idea when things will be working
>
>As a fellow sufferer, but an ignorant yet curious one, could you tell me
>what a DOS attack is?
Denial of Service
Some one has (deliberately) sent so much mail towards the demon
mailservers that they can't handle the load.
Rgds
Denis
--
I'm a happy demon customer. You may not be. If you're that
upset with Demon as an ISP, just shut up and find yourself
another ISP. You did? Then sod off to your own ISP support
group, and stop whining in here!
Julia Jones
<To...@stsm.demon.co.uk> writes
>I find that sent email isn't reaching the outside world either. Is that
>only to be expected? Do others find that?
Don't know whether it's reaching the outside world, but some of it isn't
getting back to me. I'm getting fairly reasonable levels of mailing list
traffic, but I've noticed some glitches since yesterday, including one
or two of mine not returning after several hours, when it's normally no
longer than ten minutes.
Another Demonite on one of said mailing lists reported that not only had
her onelist traffic dried up, she could no longer access the onelist
website. Presumably she's still getting that list (run off
metva.com.au).
--
Julia Jones
"Don't philosophise with me, you electronic moron!"
The Turing test - as interpreted by Kerr Avon.
Pedt Scragg
>36 hours since the last email got through from netnames.co.uk and
>still status@gate suggests there's not really much of a problem. No
>comment from any of the Demon techs in this newsgroup either. Any
>conspiracy theorists out there like to start speculating? :-)
>
various places:
Received: from tele-punt-22.mail.demon.net ([194.217.242.7])
by pedt.demon.co.uk with SMTP id <xRK6QDAv...@pedt.demon.co.uk>
for <x...@pedt.demon.co.uk> ; Sat, 10 Jun 2000 13:28:31 +0100
Received: from punt-2.mail.demon.net by mailstore for
x...@pedt.demon.co.uk
id 960638749:20:06923:1; Sat, 10 Jun 2000 12:05:49 GMT
Received: from zash.lupine.org ([205.186.156.18]) by punt-
2.mail.demon.net
id aa2007220; 10 Jun 2000 12:04 GMT
Received: (qmail 25618 invoked by uid 40001); 10 Jun 2000 11:56:25 -0000
is one I got here though the one I sent myself via netnames is still
pending after 4 hours :-( though I have one sent 8 hours ago.
--
Pedt Scragg http://signpost-design.co.uk/ ODP Editor: http://dmoz.org/
Signpost Web Design, Wrecsam, North Wales Member of FSB
NIC Handle: PS4032
Tony Pay
jajones...@nospam.demon.co.uk (Julia Jones) wrote:
> In article <20000610....@stsm.demon.co.uk>, Tony Pay
> <To...@stsm.demon.co.uk> writes
>
> > I find that sent email isn't reaching the outside world either. Is
> > that only to be expected? Do others find that?
>
> Don't know whether it's reaching the outside world, but some of it
> isn't getting back to me. I'm getting fairly reasonable levels of
> mailing list traffic, but I've noticed some glitches since yesterday,
> including one or two of mine not returning after several hours, when
> it's normally no longer than ten minutes.
I subscribe to a mailing list that has a website of recent posts. I can
access the website, and have posted to it without (for 4 hours) effect.
Usually around 30 secs.
I wonder what else (important) hasn't got through. No way of telling.
Tony
--
_________ Tony Pay
|ony:-) 79 Southmoor Rd To...@stsm.demon.co.uk
| |ay Oxford OX2 6RE www.gmn.com/artists/welcome.asp
tel/fax 01865 553339
.... Clones are people two.
.
Richard Clayton
McMahon <de...@pickaxe.demon.co.uk> writes
>On Sat, 10 Jun 2000 09:39:13 +0100, To...@stsm.demon.co.uk (Tony Pay)
>wrote:
>
>>In article <1312F6213A36...@news.darkstorm.co.uk>,
>> to...@darkstorm.co.uk (Tony Evans) wrote:
>>
>>> 3. we had a DOS attack, we have no idea when things will be working
>>
>>As a fellow sufferer, but an ignorant yet curious one, could you tell me
>>what a DOS attack is?
>
>Denial of Service
yes
>Some one has (deliberately) sent so much mail towards the demon
>mailservers that they can't handle the load.
There are several significant difficulties in discussing DOS attacks at
all. This particularly applies to giving details of severity, or the
time periods over which they are experienced.
It is undesirable to explain exactly what effect an attack had at the
receiving end, how it was detected, and what steps were taken to deal
with it. This is because that gives information to the attackers who may
decide to adjust their strategy in the future (assuming that no-one has
yet arrived at their front door and confiscated their computer).
AIUI, most of the current problem is backlogs of email - as people have
noted, there was rather less delivered yesterday than usual. The mail
will have queued on remote machines and delivery will be retried (this
is totally standard procedure). This means that more email is being
delivered today than is usual - and it has taken rather longer to arrive
than is usual.
--
richard writing to inform and not as company policy
fewer than 10 MPs still need adopting: http://www.stand.org.uk/
"Assembly of Japanese bicycle require great peace of mind" quoted in ZAMM
Julian Barker
Richard Clayton <ric...@turnpike.com> wrote
I can accept that, but it does not explain why I normally get 75+ mails
a day, sometimes much more. Since Thursday evening I have had less than
thirty. If the backlog is being cleared, shouldn't I see some of these
held messages? Instead I am getting a dribble of new messages.
--
Julian Barker
There is a coherent plan in the universe,
though I don't know what it is a plan for.
- Fred Hoyle
Graham Murray
> The mail will have queued on remote machines and delivery will be
> retried (this is totally standard procedure). This means that more
> email is being delivered today than is usual - and it has taken
> rather longer to arrive than is usual.
Please could you send some it my way?
I subscribe to a number of normally quite high volume mailing lists
(eg linux kernel) and have received *NO* mails from them at all
today. This morning I even tried sending a 'help' message to one of
the majordomos but even that has not yet elicited a response.
Richard Clayton
<jul...@rodent.demon.co.uk> writes
>Richard Clayton <ric...@turnpike.com> wrote
>>AIUI, most of the current problem is backlogs of email - as people have
>>noted, there was rather less delivered yesterday than usual. The mail
>>will have queued on remote machines and delivery will be retried (this
>>is totally standard procedure). This means that more email is being
>>delivered today than is usual - and it has taken rather longer to arrive
>>than is usual.
>>
>
>
>I can accept that, but it does not explain why I normally get 75+ mails
>a day, sometimes much more.
I don't think you meant quite that :) but you're correct it doesn't.
>Since Thursday evening I have had less than
>thirty.
ie: the email has not yet reached Demon
>If the backlog is being cleared, shouldn't I see some of these
>held messages? Instead I am getting a dribble of new messages.
After experiencing a problem some mail servers will "back off" for a
considerable time. This is a Good Thing since it avoids the wasting of
resources. Other mail servers will return the email to the sender. This
is a Bad Thing because they should be considerably more patient.
Does all your "missing" email originate at a single place ?
Mike Buckley
>thirty. If the backlog is being cleared, shouldn't I see some of these
>held messages? Instead I am getting a dribble of new messages.
>
That's exactly what I'm seeing, if all these problems can be attributed
to a mail backlog it's certainly manifesting itself in an odd way. A
colleague of mine has received an email today which was sent in 1996,
I've seen no email from the Hasbro game site and nothing off a onelist
mailing list, yet mail sent today (except onelist) is being delivered
within a couple of hours.
Most odd.
--
Mike Buckley
Brian {Hamilton Kelly}
jo...@malone1.demon.co.uk "John Malone" writes:
> In article <960729...@dsl.co.uk>, Brian {Hamilton Kelly}
> <b...@dsl.co.uk> writes
> >What do you understand as the "correct" pronunciation of Linux?
> >
>
> Here we go - pronunciation war - makes a change from the usual eh ?
:-)
All the same, YOU started it!
> >Many people (even Linux afficionados) seem to use the pronunciation
> >LINN-ucks,
>
> That is the one I was referring and is IMHO correct for English
> speakers.
Yet it has no philological nor phonemic justification; look at other
words in English that *do* have the LINN- pronunciation, such as linnet.
(OTOH, I'll grant you linen as an exception, although originally that
/was/ pronounced LINE-en). It's as bad IMHO, as those teenagers that
persist in referring to the tradenames Ellesse and Nike as ell-eese and
nike (instead of el-ESS and nike-y); IOW, ignorance of how "foreign"
words are pronounced.
> > whilst others favour LINE-ucks (since this seems to reflect
> >our, anglicized, pronunciation of the name Linus).
>
> First choice for most people new to the word in my experience.
And the sensible one.
> >However, Torvalds himself says that it should be modelled after his own
> >name, but with his Finnish pronunciation of the latter, this comes out as
> >LEEN-ucks.
>
> Yup - but he's Finnish. Strictly speaking that should be the correct one
> out of respect to the creator of the thing and is probably appropriate
> for a mainland European but as a UK citizen I'll call it LINN-ucks.
How do you pronounce the name of Linus in the Peanuts strip (and how is
it pronounced in the TV cartoon made thereof)? The name goes back to the
Romans.
--
Brian {Hamilton Kelly} b...@dsl.co.uk
"We have gone from a world of concentrated knowledge and wisdom to one of
distributed ignorance. And we know and understand less while being incr-
easingly capable." Prof. Peter Cochrane, BT Labs
Sandra Bond
<jo...@malone1.demon.co.uk> writes
>Richard, Clive and Malcolm have been quite helpful with their technical
>explanations but that is only useful to those of us who subscribe to
>demon.service.
I really don't think it takes too many brain cells for someone
confronted with an ongoing problem that's evidently confined to Demon to
subscribe to d.s expressly to find out what's going on. That's what I've
done.
For the record... no egroups mail for some 3 days plus direct to ho-
street, though oddly enough the one egroups list where I subscribe from
a forwarding account pointing at ho-street seems to be unaffected. My
mailing list hosted at aol had nothing for a while, then a sudden burst
of nearly 100 messages at once, then nothing again for 48 hours save a
plaintive email from another British demonite asking me if I'd had
anything as she hadn't either.
It's my opinion that Demon is more than adequate for my needs when it
works, and this hiccup isn't going to make me leave in a huff unless it
is *really* protracted.
But I too am puzzled that the DOS attack appears to be ongoing for some
3 days or more. Is this usual for such an attack?
Sandra
--
"After I die, I shall return to Earth as gatekeeper of a bordello,
and I won't let any of you -- not a one of you -- enter."
(-- Arturo Toscanini)
=====sandra bond: san...@ho-street.demon.co.uk=====
Tony Lynch
<$news$@snuffy.in2home.co.uk> writes
>That's why analogies don't really work as a method of explication.
>The people who don't think in analogies just ignore them, and the
>people who do think in analogies start discussing how well this
>particular analogy fits. Ideal for Usenet, in other words.
>
>
Awkward, pedestrian things, analogies.. metaphors, that's what's needed.
--
Tony
Tempora mutantur nos et mutamur in illis:
Quomodo? fit semper tempore pejor homo.
robbie irvine
Kelly}) wrote a message in demon.service that I am replying to:
>> I guess many systems have given up and bounced mail back to the origin. Not
>> good news for those who are trying to run a business as it gives the
>> impression that your email address is invalid or amateur.
>Whereas the reality there is that those responsible (if that is not too
>strong a word) for running the MTAs on which the mail is resting are the
>amateurs if they have them set up to return to sender in any shorter
>timescale than 72 hours. Personally, I always configure an MTA to
>continue trying for 120 hours, minimum.
depends wither the mta is on a LL or a dailup - a short bounce on
dialup would be preferable.
I'm still seeing variable time on mail - the wife is seeing bugger
all, but then most of her mail is mailing lists from egroups and the
like.
It's hard to tell if my work is affected as the normal recipients have
gone home already.
robbie
robbie
--
rob...@arrakis.nu <*>
want to know about uk.* ? try www.usenet.org.uk
ukvoting webpages http://www.ukvoting.org.uk/
Clive D.W. Feather
<pe...@signpost-design.co.uk> writes
>For outgoing, post.dcu is *still* changing IP address during the same
>connection from here.
If the phrase "multiple A record" doesn't mean anything to you, then I
would suggest you find out about it. But you always gave me the
impression of having that sort of knowledge.
Put simply, it's *supposed* to change. Why do you think it's a problem ?
--
Clive D.W. Feather | Internet Expert | Work: <cl...@demon.net>
Tel: +44 20 8371 1138 | Demon Internet | Home: <cl...@davros.org>
Fax: +44 20 8371 1037 | Thus plc | Web: <http://www.davros.org>
Written on my laptop; please observe the Reply-To address
Clive D.W. Feather
<pe...@signpost-design.co.uk> writes
>It doesn't help when the first connection is to *.242.55 or .56 or .57
>which either takes a bloody long time to resolve or never resolves
Do you actually mean "doesn't resolve" ? Or do you mean "doesn't
respond" ?
>Helpdesk and the backline
>didn't seem to think that it should change during a session.
Curious. I wonder why they think that ? Did they say ?
Clive D.W. Feather
<al...@whirlnet.co.uk> writes
>>>Why are punt-11 and punt-21 *still* down?
Even if you were right, why do you care ?
>Here is my analysis:
>
>punt-10: responds almost instantly
>punt-11: A one-or-two minute wait yields a connection on port 25, but no
> banner is displayed for about two or three minutes.
>punt-20: responds almost instantly
>punt-21: will not respond on port 25 at all. Over five minute wait yeilds
> nothing.
So long as one of those four responds, mail will get through. That's why
there are redundant systems in place.
>Please explain what I am seeing, what is the difference between those two,
>why should one half be busy and the other not?
You are seeing a side-effect of the work being done to resolve the
problem. For the reasons Richard already gave, I'm not going to go into
details.
Clive D.W. Feather
>That's why I asked, in the next sentence, "Is that only to be expected?"
Sorry. Sent mail *should* have been unaffected by the problems that I'm
aware of.
>> How do you know it isn't getting out ? Because of the lack of replies
>> to your account ?
>No, of course not.
Then my next sentence doesn't apply.
[But "of course" does *not* apply to some of the people who have posted
similar queries in the past.]
>As I posted previously, it's because I subscribe to a mailing list that
>maintains a website of recent posts, on which mine hasn't appeared for 7
>hours, instead of the usual few seconds.
Okay, I didn't see that message.
I would need more details to be able to investigate further.
>> [My use of the service is showing no problems with outgoing mail and
>> incoming mail arriving, sometimes delayed.]
>I'm happy for you.
That wasn't meant to be snide, just to show that things are working at
least to some extent. It is *not* a case of "no mail's going in or out
and we're keeping that a secret".
Pedt Scragg
>In article <0nabdLCq...@pedt.demon.co.uk>, Pedt Scragg
><pe...@signpost-design.co.uk> writes
>
>>For outgoing, post.dcu is *still* changing IP address during the same
>>connection from here.
>
>called "post" -- they share the load
It doesn't help when the first connection is to *.242.55 or .56 or .57
which either takes a bloody long time to resolve or never resolves and
mail sits in my out tray for 5 minutes or so. Helpdesk and the backline
didn't seem to think that it should change during a session.
Gives up every time or takes ages to resolve to .55, .56, .57 when it
reaches
[194.159.36.68] finch-service-2-162.router.demon.net
>
>> Reported direct to Demon a while ago but still happening:
>
>excellent!
>
If you mean excellent it is still happening then I wish the first
attempts did not go via .55 .56 or .57 as urgent email sits there
stubbornly until a traceroute shows that I'm getting a different
part of the cluster - usually about 5 minutes down the line. Although
I switched back to using post.dcu to do tests I have now gone back to
using a specific IP address as advised by the helpdesk backline.
If you mean excellent it was reported direct to Demon then, for
something that seemed obscure, a direct report was preferable than
a whinge in ds. Updated report gone off today.
TP5.01 on Green from 01978 357xxx
Ben Smithurst
> On Sun, 11 Jun 2000 17:45:23 +0100, b...@scientia.demon.co.uk (Ben
> Smithurst) wrote:
>
> <snip>
>
>> Well, I just tried that, and nothing happened. Perhaps Demon should
>> change the MX records to point to our own hosts as their servers are
>> obviously broken (I know they won't though).
>
> I very much hope that they would not do this !
>
> Are you aware of the implications of this ?
Yes.
> Do you dial up to your Demon account on a 24/7 basis ?
> Does everyone using Demon do this ?
No, no. But for a temporary measure, I'd be willing to stay online for
a few hours waiting for MTAs to retry if it meant getting this backlog
cleared. The suggestion was only half serious anyway.
> I very much doubt it, if the MX records were changed to use our hosts
> then the mails would bounce if
>
> [1] Our machines were not connected when the mails were trying to be
> delivered.
>
> [2] We didn`t have an SMTP server listening on port 25, Although many
> people probably use turnpike are they all collecting via SMTP ?
I see Dave Holland attempted to educate you here. These are both
*temporary* errors. The only thing which will cause the remote MTA to
permanently give up is a 5xx code[0], if you know of a way for a remote
MTA to get this code when the destination host is unreachable, I'd like
to hear about it. As the sending host won't get further than connect(),
I think it's unlikely.
[0] unless I've forgotten something.
Ben Smithurst
> I'm still seeing variable time on mail - the wife is seeing bugger
> all, but then most of her mail is mailing lists from egroups and the
> like.
>
> It's hard to tell if my work is affected as the normal recipients have
> gone home already.
I seem to be getting a bit more mail today, things are *slightly*
improving it seems. I guess I'll have to start checking all my mailing
list subscriptions and resubscribing if necessary (must be nearly four
days now, when MTAs might just give up on some messages), still, I doubt
we'll see an apology from Demon for the wasted time this will cause.
Graham Murray
> Even if you were right, why do you care ?
Because some systems (eg the vger.rutgers.edu mailing list expanders)
always send mail to the same punt. They seem to have resolved it once
and cached the result. I know that this is not optimum behaviour but
we have no control over the systems used by people attempting to send
mail to us.
Richard Clayton
street.demon.co.uk> writes
>But I too am puzzled that the DOS attack appears to be ongoing for some
>3 days or more. Is this usual for such an attack?
no - for the simple reason that tracing is usually only possible whilst
the attack is ongoing - hence stopping is a wise move if you wish to
avoid being caught!
Andy
<ric...@turnpike.com> writes>>What would it have cost Demon to send a short batch e-mail to it's own
>>customers explaining briefly about mail servers backing off a while then
>>retrying later. That could be understood by anyone and would be
>>appreciated by most.
>
>I really don't think that shipping more email through the system is an
>entirely desirable thing to do.
>
:)
Richard, are you able [1] to tell us if the problems are now internal to
demon or external - ie, are you now delivering immediately all the mail
you are receiving?
[1] and if so will you please do it here and now, he said trying to
phrase the question so as to preclude "yes" as the answer.
--
Andy
For Austrian philately <URL: http://www.kitzbuhel.demon.co.uk/austamps/>
For Lupus <URL: http://www.kitzbuhel.demon.co.uk/lupus/>
For my other interests <URL: http://www.kitzbuhel.demon.co.uk/>
michael lefevre
news:slrn8k6i6e....@muir-et5.staff.demon.net...
> On Sun, 11 Jun 2000 07:41:31 +0100, Roger Jones
<{news200004}@roga.demon.co.uk> wrote:
>
> >I guess many systems have given up and bounced mail back to the origin.
Not
> >good news for those who are trying to run a business as it gives the
> >impression that your email address is invalid or amateur.
>
many MTAs (including demon's) will send warning messages back to the sender
if the mail is delayed by more than 24 hours (some will warn after only 4
hours)...
i know the difference between a failure and a warning... many people don't,
and mere warnings give a bad impression of the service...
> >I know for certain that at least 2 of my clients sent me important emails
on
> >Friday that have not yet appeared, and now I think it is unlikely that
they
> >will.
> >
> >It is perhaps a shame that we do not have a service level agreement with
> >demon which involves significant refunds when parts of the service are
not
> >workling.
>
> Do you believe the UK Post Office would give you any guarantee for
> delivery of mail posted overseas ?
i don't really see how that's relevant... we're talking about mail received,
not mail sent out... demon could offer a guarantee that the demon email
services would be available for some percentage of the time, and offer a
rebate if that level of service wasn't achieved... not that i believe that
demon would do that...
and if Demon is striving to obtain a service as reliable as the UK post
office, i'm worried...
michael
michael lefevre
news:AFHGZpGc...@romana.davros.org...
> In article <20000611....@stsm.demon.co.uk>, Tony Pay
> <To...@stsm.demon.co.uk> writes
> >>>I find that sent email isn't reaching the outside world either.
> >> Sent mail runs through different systems.
> >That's why I asked, in the next sentence, "Is that only to be expected?"
>
> Sorry. Sent mail *should* have been unaffected by the problems that I'm
> aware of.
>
> >> How do you know it isn't getting out ? Because of the lack of replies
> >> to your account ?
> >No, of course not.
>
> Then my next sentence doesn't apply.
>
> [But "of course" does *not* apply to some of the people who have posted
> similar queries in the past.]
>
> >As I posted previously, it's because I subscribe to a mailing list that
> >maintains a website of recent posts, on which mine hasn't appeared for 7
> >hours, instead of the usual few seconds.
>
> Okay, I didn't see that message.
>
> I would need more details to be able to investigate further.
i wonder if that was egroups... i received messages from the demon mailers
on saturday telling me that the outgoing messages i sent on friday had been
delayed by 24 hours... those messages arrived yesterday...
header (bits replaced by the egroups website with ... and bits by be with
REMOVED...)
From ml@m... Sun Jun 11 17:20:20 2000
Return-Path: <REM...@michaellefevre.com>
Received: (qmail 27088 invoked from network); 12 Jun 2000 00:20:19 -0000
Received: from unknown (10.1.10.142) by m2.onelist.org with QMQP; 12 Jun
2000 00:20:19 -0000
Received: from unknown (HELO finch-post-12.mail.demon.net) (194.217.242.41)
by mta3 with SMTP; 12 Jun 2000 00:20:19 -0000
Received: from alzdisint.demon.co.uk ([194.222.184.244] helo=michael) by
finch-post-12.mail.demon.net with smtp (Exim 2.12 #1) id 130QJL-000Dtu-0C
for REM...@egroups.com; Fri, 9 Jun 2000 15:03:03 +0000
Message-ID: <025e01bfd223$f0fa7f60$0501a8c0@a...
To: REM...@egroups.com
References: 70.e3642.26724aee@a... <000901bfd215$5975c5e0$fd066cd5@gonzo
Subject: Re: [REMOVED] fuckin furious!
Date: Fri, 9 Jun 2000 16:01:47 +0100
MIME-Version: 1.0
Content-Type: multipart/alternative;
boundary="----=_NextPart_000_0258_01BFD22C.04145AA0"
X-Priority: 3
X-MSMail-Priority: Normal
X-Mailer: Microsoft Outlook Express 5.00.2919.6600
X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2919.6600
X-eGroups-From: "michael lefevre" <REM...@michaellefevre.com>
From: "michael lefevre" ml@m...
> >> [My use of the service is showing no problems with outgoing mail and
> >> incoming mail arriving, sometimes delayed.]
> >I'm happy for you.
>
> That wasn't meant to be snide, just to show that things are working at
> least to some extent. It is *not* a case of "no mail's going in or out
> and we're keeping that a secret".
well both status@gate and your helpdesk seem to be under the impression that
everything is working fine now...
i just tried a few times, and not all of the attempted connections to the
punts worked... this together with your comment of "to some extent", would
appear to contradict the 'official' view that everything is 100% fine now...
michael
Ben Smithurst
> Because some systems (eg the vger.rutgers.edu mailing list expanders)
> always send mail to the same punt. They seem to have resolved it once
> and cached the result. I know that this is not optimum behaviour but
> we have no control over the systems used by people attempting to send
> mail to us.
That's broken behaviour. Did you ask the admin to fix their broken
server? (Assuming he/she/it knows how, if they're running such a broken
system in the first place I won't assume they do.)
Pedt Scragg
>In article <UPeTXuD4...@pedt.demon.co.uk>, Pedt Scragg
><pe...@signpost-design.co.uk> writes
>>It doesn't help when the first connection is to *.242.55 or .56 or .57
>>which either takes a bloody long time to resolve or never resolves
>
>respond" ?
Sorry, 'doesn't respond' is correct. It seems to me that the nigger in
the woodpile is when I reach
[194.159.36.68] finch-service-2-162.router.demon.net
I can see DNS etc. but cannot normally reach the *.242.55, .56, .57 bits
of post.dcu
>
>>Helpdesk and the backline
>>didn't seem to think that it should change during a session.
>
>Curious. I wonder why they think that ? Did they say ?
>
They didn't say why they thought that.
Actually, I would like to thank those who dealt with my original queries
both on the helpdesk and the backline who went to a number of lengths to
try and find the problem which obviously seems to be a small minority of
customers. Interestingly the first helpdesk bod I spoke to also couldn't
get past the router listed above.
s/small/extremely small
or we'd be seeing all sorts of complaints here ;-)
Additional info has just been emailed to the gentleman who responded to
the report I made.
John Malone
>>explanations but that is only useful to those of us who subscribe to
>>demon.service.
>
>I really don't think it takes too many brain cells for someone
>confronted with an ongoing problem that's evidently confined to Demon to
>subscribe to d.s expressly to find out what's going on. That's what I've
>done.
>
This may surprise you but there are some people out there that do not
even know what Usenet is ;)
Some of them pay Demon TAM and they deserve an explanation for their
lack of e-mail.
I think it's unreasonable to expect people to subscribe to d.s to find
an answer and perfectly reasonable to expect Demon to provide up to date
situation reports.
Having said that you did the right thing by looking here for an answer
but you knew where to look didn't you ?
>
>It's my opinion that Demon is more than adequate for my needs when it
>works, and this hiccup isn't going to make me leave in a huff unless it
>is *really* protracted.
>
Demon is fine for me and I intend to stay.
I can't speak for the customers who have been left in the dark.
--
jo...@malone1.demon.co.uk | Lurk and Learn
Tony Pay
b...@dsl.co.uk (Brian {Hamilton Kelly}) wrote:
> In article <zYbH6VAK...@malone1.demon.co.uk>
> jo...@malone1.demon.co.uk "John Malone" writes:
> > Here we go - pronunciation war - makes a change from the usual eh ?
>
> :-)
OK, how do you pronounce TeX?
Or is that such a chestnut that it's not worth bringing up?
Tony
--
_________ Tony Pay
|ony:-) 79 Southmoor Rd To...@stsm.demon.co.uk
| |ay Oxford OX2 6RE www.gmn.com/artists/welcome.asp
tel/fax 01865 553339
... Some Do, Some Don't, Some Will and Some Won't.
Martin
street.demon.co.uk> writes
>In article <skmAlWC6...@malone1.demon.co.uk>, John Malone
><jo...@malone1.demon.co.uk> writes
>>Richard, Clive and Malcolm have been quite helpful with their technical
>>explanations but that is only useful to those of us who subscribe to
>>demon.service.
>
>I really don't think it takes too many brain cells for someone
>confronted with an ongoing problem that's evidently confined to Demon to
>subscribe to d.s expressly to find out what's going on. That's what I've
>done.
>
It's the only way to find out.
Oh and thanks to all those who have been working hard on the problem.
And thanks to all those who have been working very hard, distancing
themselves (problem what problem?), and issuing minimal information.
.....Martin
Richard Clayton
michael lefevre <michae...@michaellefevre.com> writes
>well both status@gate and your helpdesk seem to be under the impression that
>everything is working fine now...
>
>i just tried a few times, and not all of the attempted connections to the
>punts worked...
there is still a lot of email flowing through the system
was the failure to connect at the TCP/IP level or a timeout waiting for
an SMTP banner ?
>this together with your comment of "to some extent", would
>appear to contradict the 'official' view that everything is 100% fine now...
--
Tony Evans
>On Sun, 11 Jun 2000 17:45:23 +0100, b...@scientia.demon.co.uk (Ben
>Smithurst) wrote:
><snip>
>
>>Well, I just tried that, and nothing happened. Perhaps Demon should
>>change the MX records to point to our own hosts as their servers are
>>obviously broken (I know they won't though).
>I very much hope that they would not do this !
I wish they would give us the option.
>Are you aware of the implications of this ?
Yes, I think he is.
>Do you dial up to your Demon account on a 24/7 basis ?
You wouldn't need to.
>Does everyone using Demon do this ?
Probably very few.
>I very much doubt it, if the MX records were changed to use our hosts
>then the mails would bounce if
>[1] Our machines were not connected when the mails were trying to be
>delivered.
No they wouldn't, they would be held until the MTA could make a
connection, or until a number of days had passed and then they would
be bounced. Or, they would be delivered to the demon mail machines if
they were listed in alternative MX records.
>[2] We didn`t have an SMTP server listening on port 25, Although many
>people probably use turnpike are they all collecting via SMTP ?
I think you'll find the people who want MX records pointing at their
machines know how to receive mail properly at them.
>I doubt these very much, so please exercise a little logic and
>knowledge before posting :)
You should follow your own advice.
Tony Evans
--
I like my steak so rare, a good vet could save it.
Gemmell Mania : http://www.gemmell-mania.org.uk | IRC : Blade
JAGW - Just Another Gathering Website at | ICQ : 170850
http://www.darkstorm.org/gathering - lots of photos!
Andy
<anto...@antonymous.freeserve.co.uk> writes
>In article <394499E2...@hogsmeade.co.uk>, Sarah Thornhill
><sar...@hogsmeade.co.uk> writes
>>
>>
>>The corporate mind, like God, moves in mysterious ways :o)
>
>Metaphysically, one must suppose.
>
>Exercise for the week.. attempt to imagine, envisage, whatever, "the
>corporate mind".. great fun.
For some value of Fun..
35 years of experience enables me to state that the following will be
found true by any employee of a company with over 99 employees:
1. Ascertain all the available data relevant to a decision you await
from Higher Management
2. Analyse it logically
3. Draw the only possible conclusion
4. You may now proceed on the basis that, with 0.95 probability, that is
NOT what they will decide.
Stephen Agar
>In article <dXbBHJAA...@meurglys.co.uk>, Stephen Agar
><ste...@merglys.co.uk> wrote:
>> In article <8i0nlb$efl$1...@teasel.co.uk>, Ellen Mizzell
>> <$news$@snuffy.in2home.co.uk> apparently said...
>>>In article <vLyNQXAi...@meurglys.co.uk>, Stephen Agar
>>><ste...@merglys.co.uk> wrote:
>>>
>>>I already made the same mistake and got corrected -- Malcolm's
>>>analogy is about mail posted overseas to the UK, not mail posted
>>>to an overseas destination.
>>>
>>
>> Same thing applies. Most foreign PO's will give compensation if you keep
>> proof of posting.
>
>But to fit in with MM's analogy, they'd have to compensate the
>intended recipient, not the person who posted it.
But of course the analogy doesn't work at all because with a postal
service only the sender pays to send, but with the Internet (usually)
the sender pays to send and the recipient pays to receive. Now that's a
good idea - maybe we should go back 150 years and start charging
addressees. :-)
>
>> Still, a bit off topic - I just can't resist the bait
>> every couple of weeks when someone uses a Post Office analogy...
>>
>
>That's why analogies don't really work as a method of explication.
>The people who don't think in analogies just ignore them, and the
>people who do think in analogies start discussing how well this
>particular analogy fits. Ideal for Usenet, in other words.
>
>
Yep.
--
Stephen Agar, Brighton, UK
http://www.diplomacy-archive.com
Mike Pellatt
<michae...@michaellefevre.com> wrote:
>"Malcolm Muir" <mal...@demon.net> wrote in message
>news:slrn8k6i6e....@muir-et5.staff.demon.net...
>> On Sun, 11 Jun 2000 07:41:31 +0100, Roger Jones
><{news200004}@roga.demon.co.uk> wrote:
>>
>> >I guess many systems have given up and bounced mail back to the origin.
>Not
>> >good news for those who are trying to run a business as it gives the
>> >impression that your email address is invalid or amateur.
>>
>> Most MTAs will not do that for at least 4 days.
>
>many MTAs (including demon's) will send warning messages back to the sender
>if the mail is delayed by more than 24 hours (some will warn after only 4
>hours)...
>
>i know the difference between a failure and a warning... many people don't,
>and mere warnings give a bad impression of the service...
Why ?? I'd say they give a _good_ impression. They say:
"Hi, pal. Welcome to the Internet. I know you expect email to be
delivered almost instantaneously. Sorry, but for some reason or
other I couldn't do that right now. You may choose to describe
that reason as "The internet isn't working". I'm a persistent bugger, so
I'll keep trying to deliver it for a week for you, before I give
up and send it back to you."
>and if Demon is striving to obtain a service as reliable as the UK post
>office, i'm worried...
The service of the Post Office is, bar theft and matches dropped
into post boxes, entirely within their control. Delivery of
email to Demon is not like that. It's pretty amazing it's as
good as it is. It's also depressing how poor the PO is, but that
seems to be about getting/keeping decent staff who don't prefer
to toss mailbags into the river rather than putting the contents
through the right letterboxes.
--
Mike Pellatt
John Malone
<b...@dsl.co.uk> writes>> In article <960729...@dsl.co.uk>, Brian {Hamilton Kelly}
>> <b...@dsl.co.uk> writes
>> >What do you understand as the "correct" pronunciation of Linux?
>> >
>>
>> Here we go - pronunciation war - makes a change from the usual eh ?
>
>:-)
>
>
Guilty - fair cop.
>> >Many people (even Linux afficionados) seem to use the pronunciation
>> >LINN-ucks,
>>
>> That is the one I was referring and is IMHO correct for English
>> speakers.
>
>Yet it has no philological nor phonemic justification; look at other
>words in English that *do* have the LINN- pronunciation, such as linnet.
>(OTOH, I'll grant you linen as an exception, although originally that
>/was/ pronounced LINE-en). It's as bad IMHO, as those teenagers that
>persist in referring to the tradenames Ellesse and Nike as ell-eese and
>nike (instead of el-ESS and nike-y); IOW, ignorance of how "foreign"
>words are pronounced.
>
I'm wearing a pair of Ellesse (el-ESS) trainers right now.
I wondered why the young 'uns were giving me funny looks.
>
>How do you pronounce the name of Linus in the Peanuts strip (and how is
>it pronounced in the TV cartoon made thereof)? The name goes back to the
>Romans.
>
I must have lived a sheltered life. I've never seen a complete episode
of Peanuts but anyway, I would pronounce it Lie-nuss.
Anyone wanna argue :)
John Malone
<To...@stsm.demon.co.uk> writes
>> jo...@malone1.demon.co.uk "John Malone" writes:
>
>> > Here we go - pronunciation war - makes a change from the usual eh ?
>>
>> :-)
>
>
>Or is that such a chestnut that it's not worth bringing up?
>
That's liable to elicit a tetchy response.
Tony Lynch
<an...@kitzbuhel.demon.co.uk> writes
>In article <KX57AKAN...@antonymous.freeserve.co.uk>, Tony Lynch
><anto...@antonymous.freeserve.co.uk> writes
>>In article <394499E2...@hogsmeade.co.uk>, Sarah Thornhill
>><sar...@hogsmeade.co.uk> writes
>>>
>>>
>>>The corporate mind, like God, moves in mysterious ways :o)
>>
>>Metaphysically, one must suppose.
>>
>>Exercise for the week.. attempt to imagine, envisage, whatever, "the
>>corporate mind".. great fun.
>
>For some value of Fun..
>
>35 years of experience enables me to state that the following will be
>found true by any employee of a company with over 99 employees:
>
>1. Ascertain all the available data relevant to a decision you await
>from Higher Management
>
>2. Analyse it logically
>
>3. Draw the only possible conclusion
>
>4. You may now proceed on the basis that, with 0.95 probability, that is
>NOT what they will decide.
Splendid.. corporate decisions, then, having much the same status as
Divine Mystery, I suppose.
Would explain a lot..
--
Tony
Tempora mutantur nos et mutamur in illis:
Quomodo? fit semper tempore pejor homo.
Alan Ford
>On 11 Jun 2000 22:47:55 GMT, Alan Ford
><al...@whirlnet.co.uk> wrote:
>
>>Indeed. Three years ago we would have already had several demon.announce
>>postings keeping us informed of what the problems are, what is being
>>done to fix them, advice of what we should do, when things will be getting
>>back to normal, etc. Up-to-the minute, important, useful information.
>
>I think your rose-coloured spectacles need removing.
Actually they're more dirt-coloured, I haven't cleaned them in a couple
of days :(
>Or your
>"nostalgia ain't what it used to be" program needs to be kill -9'ed.
>
>Things have _always_ been bad, certianly since the 10,000 subscribers
>barrier was broken [1], with the odd bit of excellent communication that
>led us to hope that the philosophy had changed. Sadly, those hopes
>were always dashed.
Well yes, I know we have infinite wants, we always want more information.
But things definitely WERE better three years ago. If we had an online
demon.announce archive you could remind youself of the many updates
received about the news SNAFU. I was content during that period. Why?
Because I was being informed.
--
Alan Ford * al...@whirlnet.co.uk * http://www.whirlnet.co.uk/
PGP Key: 0x8F807D7D - email p...@whirlnet.co.uk or see keyservers
Demon Newsgroups Info + FAQs: http://www.whirlnet.co.uk/demon/
!! YOUR PRIVACY IS AT RISK. TELL YOUR MP *NOW*! http://www.stand.org.uk/ !!
Alan Ford
>Do you dial up to your Demon account on a 24/7 basis ?
No and no.
>I very much doubt it, if the MX records were changed to use our hosts
>then the mails would bounce if
>
>[1] Our machines were not connected when the mails were trying to be
>delivered.
No, mail will be queued until we connect and it can connect to us. As
long as: a) they are using a proper MTA; and b) they try to send mail
when we are conencted, which, if we are irregular, "quick burst" users,
is unlikely.
>[2] We didn`t have an SMTP server listening on port 25, Although many
>people probably use turnpike are they all collecting via SMTP ?
If the punts are put as second priority machines, proper MTAs will send
to them if they can't reach ourselves as top priority.
However it is worth realising that a large proportion of the world's
MTAs are actually crap and this wouldn't work in practice, there are
too many problems that could occur.
It's fine the way it is, as long as the punts don't break agian :(
It would be even better if we could have our five minute kicks back
again.
Alan Ford
<cl...@on-the-train.demon.co.uk> wrote:
>In article <slrn8k82n...@whirlnet.demon.co.uk>, Alan Ford
><al...@whirlnet.co.uk> writes
>>>>Why are punt-11 and punt-21 *still* down?
>
>Even if you were right, why do you care ?
1) I am curious.
2) If machines are broken it is a matter of principle.
3) Some MTAs are crap and only talk to certain punts.
>>Please explain what I am seeing, what is the difference between those two,
>>why should one half be busy and the other not?
>
>You are seeing a side-effect of the work being done to resolve the
>problem. For the reasons Richard already gave, I'm not going to go into
>details.
Thank you. That's really all I wanted to hear, as I realise no more could
be given. It's a *side-effect* -- I thought things seemed a bit odd for
a simple DoS attack.
Alan Ford
>OK, how do you pronounce TeX?
>
>Or is that such a chestnut that it's not worth bringing up?
Quite a chestnut, really :)
However there is only one *correct* pronunciation, "tekh" (like "loch").
Peter Gradwell
Grinter) wrote:
>Tony Evans <to...@darkstorm.co.uk> writes:
>> It would explain to me why clicknames.net seems to be holding onto
>> everything for Demon rather than letting it go in dribs and drabs.
>
>Well, now we've seen a post from Peter indicating that they're using
>qmail.
>
>It implements a per-message delay mechanism, but if a server is
>failing to respond altogether (indicating some form unreachability)
>then it will add it to its time-out list and won't keep trying it for
>every message in the queue or every new message.
indeed, but if you run
/var/qmail/bin/qmail-tcpok
it clears the dns/delay cache and you can then do
killall -HUP qmail-send
and rerun the queue... not that i've done that a few times :)
peter
--
peter at gradwell dot com. online @ http://www.gradwell.com/
Ben Smithurst
> OK, how do you pronounce TeX?
Like the "Tex" in Texas, I know it's wrong but I use the word in
speaking so little that I don't really care.
How do you pronounce the name of that popular editor, "vi"? I pronounce
that wrong too, but that's more because I learnt the word long before I
ever learnt the correct pronunciation. (It's "vee eye" for those who
don't know, not just a single syllable rhyming with "bye" as I've always
used.)
Brian {Hamilton Kelly}
to...@stsm.demon.co.uk "Tony Pay" writes:
> In article <960764...@dsl.co.uk>,
> b...@dsl.co.uk (Brian {Hamilton Kelly}) wrote:
>
> > In article <zYbH6VAK...@malone1.demon.co.uk>
> > jo...@malone1.demon.co.uk "John Malone" writes:
>
> > > Here we go - pronunciation war - makes a change from the usual eh ?
> >
> > :-)
>
> OK, how do you pronounce TeX?
>
> Or is that such a chestnut that it's not worth bringing up?
I would have thought that, given my use of braces to indicate \bgroup
\egroup for my double-barrelled unhyphenated surname, that you can be
assured that I *do* know how to pronounce it.
(Oh, and given that it's also been my e-mail address for 15 years at
work, too)
--
Brian {Hamilton Kelly} b...@dsl.co.uk
"We have gone from a world of concentrated knowledge and wisdom to one of
distributed ignorance. And we know and understand less while being incr-
easingly capable." Prof. Peter Cochrane, BT Labs
Andy
>It's as bad IMHO, as those teenagers that
>persist in referring to the tradenames Ellesse and Nike as ell-eese and
>nike (instead of el-ESS and nike-y); IOW, ignorance of how "foreign"
>words are pronounced.
>
John Lynch
Buckley <Mike.B...@cotspig.co.uk> writes
>In article <jx$TJ5Amd...@rodent.demon.co.uk>, Julian Barker
><jul...@rodent.demon.co.uk> writes
>>I can accept that, but it does not explain why I normally get 75+ mails
>>a day, sometimes much more. Since Thursday evening I have had less than
>>thirty. If the backlog is being cleared, shouldn't I see some of these
>>held messages? Instead I am getting a dribble of new messages.
>>
>
>That's exactly what I'm seeing, if all these problems can be attributed
>to a mail backlog it's certainly manifesting itself in an odd way. A
>colleague of mine has received an email today which was sent in 1996,
>I've seen no email from the Hasbro game site and nothing off a onelist
>mailing list, yet mail sent today (except onelist) is being delivered
>within a couple of hours.
>
>Most odd.
And isn't it terribly convenient that mailbombs (that _is_ what we're
talking about, isn't it?) mustn't be talked about? Demon's mail service
went for a s**t. They _say_ they were mailbombed. I don't know that's
true. You don't know it's true. And this, remember, is the company that
-- since they were bought by Scottish Power -- has used deceit as a
regular tool of customer communications
--
John Lynch
Malcolm Muir
>And isn't it terribly convenient that mailbombs (that _is_ what we're
>talking about, isn't it?) mustn't be talked about? Demon's mail service
>went for a s**t. They _say_ they were mailbombed. I don't know that's
>true. You don't know it's true. And this, remember, is the company that
>-- since they were bought by Scottish Power -- has used deceit as a
>regular tool of customer communications
For the record since you are making false accusations. At no point
have we claimed we were mailbombed.
--
Malcolm S. Muir Demon Internet
Sunderland 322 Regents Park Road
England London N3 2QQ
Graham Murray
> Not so: if the punts remained as lower priority (i.e. higher number in
> the MX record) then the sending MTA *should* attempt to deliver to them
> instead of to our own hosts.
Or maybe better (for the situation this weekend) would be to have the
Demon host as a lower priority MX record than the punts. That way
correctly behaving MTAs would attempt to deliver to the punts first
and only try the host if that failed.
Once Demon offer ADSL, I would hope that they would make the
customer's system the highest priority MX record, with the punts as
backup for when the customer is not online.
Stefan Davids
Ben> How do you pronounce the name of that popular editor, "vi"? I
Ben> pronounce that wrong too, but that's more because I learnt the
Ben> word long before I ever learnt the correct pronunciation. (It's
Ben> "vee eye" for those who don't know, not just a single syllable
Ben> rhyming with "bye" as I've always used.)
Interesting. I've always pronounced it 'vee eye' but I'd become
convinced I had mislearnt (as is common when you see things written
down long before you ever hear someone say it) because everyone else
I've subsequently heard say it says 'vi' to rhyme with 'bye'.
Stefan
robbie irvine
Smithurst) wrote a message in demon.service that I am replying to:
>I seem to be getting a bit more mail today, things are *slightly*
>improving it seems. I guess I'll have to start checking all my mailing
>list subscriptions and resubscribing if necessary (must be nearly four
>days now, when MTAs might just give up on some messages), still, I doubt
>we'll see an apology from Demon for the wasted time this will cause.
Well the wife got 200odd messages rejected by egroups (she's not
happy) I've cobbled together a script for her to retrieve them from
the list archives. I lost some mail to it timing out, but my mirrored
setup at work grabbed them ok so....
I would be interested in why it happened/who/what did it.
Not a good couple of days.
robbie
--
rob...@arrakis.nu <*>
want to know about uk.* ? try www.usenet.org.uk
ukvoting webpages http://www.ukvoting.org.uk/
James Coupe
> However, Torvalds himself says that it should be modelled after his own
> name, but with his Finnish pronunciation of the latter, this comes out as
> LEEN-ucks.
>
> I've never heard an Anglophone use this pronunciation.
Actually, I have heard it used. But mostly only when the person
concerned was being annoying.
--
James Coupe | PGP Key 0x5D623D5D
"I know my ex-boyfriend lies. Oh, he does it every time. It's just his
permanent disguise, yeah, yeah, but he's drop dead gorgeous."
michael lefevre
news:slrn8kadc2....@lurch.pellatt.co.uk...
i have no problem with that sort of warning being sent... my point was it
gives a bad impression of the recipient's mail service...
the sender's service is saying "i'm trying to get this message through, but
i can't because the recipients mail servers are stuffed, i'll keep trying
until someone over there fixes the problem"... this is a useful thing to be
doing (all the better if the sender has control over it (we like DSN))...
however, it gives a bad impression of the recipient's service...
michael
James Coupe
> Continuing that... Some will say that the port is run by a useless team, and
> change their business to a different port.
And if the problem is that someone is maliciously sending hundreds of
small boats to clog up the port.....?
michael lefevre
news:HG1SnCE$7RR5...@turnpike.com...
> In article <960824796.21836.0...@news.demon.co.uk>,
> michael lefevre <michae...@michaellefevre.com> writes
>
> >well both status@gate and your helpdesk seem to be under the impression
that
> >everything is working fine now...
> >
> >i just tried a few times, and not all of the attempted connections to the
> >punts worked...
>
> there is still a lot of email flowing through the system
>
> was the failure to connect at the TCP/IP level or a timeout waiting for
> an SMTP banner ?
tcp connected, and the connection timed out before the smtp banner
appeared... not exactly sure what the timeout is... around a minute i
think... on the connections that worked, the banner appeared pretty much
immediately...
i just tried a few connections and it worked fine... :)
michael
Mark Evans
> In article <zYbH6VAK...@malone1.demon.co.uk>
> jo...@malone1.demon.co.uk "John Malone" writes:
>> In article <960729...@dsl.co.uk>, Brian {Hamilton Kelly}
>> <b...@dsl.co.uk> writes
>> >What do you understand as the "correct" pronunciation of Linux?
>> >
>>
>> Here we go - pronunciation war - makes a change from the usual eh ?
>> >Many people (even Linux afficionados) seem to use the pronunciation
>> >LINN-ucks,
>>
>> That is the one I was referring and is IMHO correct for English
>> speakers.
> Yet it has no philological nor phonemic justification; look at other
> words in English that *do* have the LINN- pronunciation, such as linnet.
> (OTOH, I'll grant you linen as an exception, although originally that
> /was/ pronounced LINE-en). It's as bad IMHO, as those teenagers that
> persist in referring to the tradenames Ellesse and Nike as ell-eese and
> nike (instead of el-ESS and nike-y); IOW, ignorance of how "foreign"
> words are pronounced.
An older example is the famous Roman city burried by a volcanic
eruption. Latin argably isn't a "foreign langauge", since it is
one of the root languages for modern English.
Mark Evans
> news:wFLKZyCq...@romana.davros.org...
>> In article <jx$TJ5Amd...@rodent.demon.co.uk>, Julian Barker
>> <jul...@rodent.demon.co.uk> writes
>> >I can accept that, but it does not explain why I normally get 75+ mails
>> >a day, sometimes much more. Since Thursday evening I have had less than
>> >thirty. If the backlog is being cleared, shouldn't I see some of these
>> >held messages? Instead I am getting a dribble of new messages.
>>
>>
>> When a sender fails to contact the recipient, they are supposed to wait
>> before trying. Most systems wait longer each time that they tried. After
>> a while they are waiting several hours between tries.
>>
> I guess many systems have given up and bounced mail back to the origin. Not
> good news for those who are trying to run a business as it gives the
> impression that your email address is invalid or amateur.
When in fact it is the MTA which is bouncing the email which is "amateur".
Any decent MTA should attempt to deliver for several days.
Also Demon having off site (e.g. the other side of the North Sea) MX machines
wouldn't be a bad idea either.
> I know for certain that at least 2 of my clients sent me important emails on
> Friday that have not yet appeared, and now I think it is unlikely that they
> will.
> It is perhaps a shame that we do not have a service level agreement with
> demon which involves significant refunds when parts of the service are not
> workling.
Even if there were Demon surely can't accept responsiblity for the results
of other people running misconfigured software.
Mark Evans
> On Sun, 11 Jun 2000 07:41:31 +0100, Roger Jones <{news200004}@roga.demon.co.uk> wrote:
>>I guess many systems have given up and bounced mail back to the origin. Not
>>good news for those who are trying to run a business as it gives the
>>impression that your email address is invalid or amateur.
> Most MTAs will not do that for at least 4 days.
Correctly written ones. It wouldn't surprise me in the slightest if those
involved with a certain soon not to be software company can't wait even
that long.
Mark Evans
> On 11 Jun 2000 10:56:17 +0100, Andrew Gierth
> <and...@erlenstar.demon.co.uk> wrote:
>>I have mail still queued on one of my machines since _Friday_, and
>>right now I'm getting:
> Yesterday I had a mail that was sent late on _Thursday_ finally get
> through. There are probably other similar ones like that, but that
> was the one I spotted.
>>$ telnet punt-1.mail.demon.net 25
>>Trying 194.217.242.35...
>>telnet: connect to address 194.217.242.35: Operation timed out
>>telnet: Unable to connect to remote host: Operation timed out
>>$ telnet punt-2.mail.demon.net 25
>>Trying 194.217.242.6...
>>telnet: connect to address 194.217.242.6: Operation timed out
>>telnet: Unable to connect to remote host: Operation timed out
>>
>>(note that both IP addresses of each punt were tried, hence the two
>>failure messages)
> I have tried that several times both yesterday and today, and yesterday
> I could always get ONE punt talking, today I could get TWO (punt-10 and
> punt-20).
> Nevertheless that is still pretty crap, no wonder mail is taking agest
> to get through if the sending MTAs have a one-in-four chance of getting
> a working punt :(
Except that the way the software should work is to try all MX records
(in pref order) and all IP addresses per MX record. (N.B. since MX
records exist the MTA certainly should not be using any A records.)
Thus (pun unintentional) under the circumstances you describe an
MTA should virtually always find a punt.
Mark Evans
> In demon.service, al...@whirlnet.co.uk (Alan Ford) wrote:
>>Nevertheless that is still pretty crap, no wonder mail is taking agest
>>to get through if the sending MTAs have a one-in-four chance of getting
>>a working punt :(
> And, for some large systems like clicknames.net and egroups.com, I
> would expect the problem to be worse.
> Loads of mail backed up and destined for punt*.demon.co.uk. MTA tries
> again for one of the mails, finds it can't connect (high chance), and
> so delays all the rest as well.
If the MTA's really are using this algorithm then it will only make
things worst. Unless the queues for punt*.demon.co.uk expire at the
same time as the DNS TTL. (Which is pointless since until TTL expires
all relevent DNS data will be cached locally anyway.)
An MTA really should do a DNS lookup ever time it attempts to deliver
an email.
"Optimising" SMTP mail is a very non trivial task. Especially since
it is perfectly possible for the DNS data to change whilst an item
is in a queue.
Mark Evans
> In article <slrn8k6i6e....@muir-et5.staff.demon.net>, Malcolm
> Muir <mal...@demon.net> writes
>>On Sun, 11 Jun 2000 07:41:31 +0100, Roger Jones <{news200004}@roga.demon.co.
>>uk> wrote:
>>
>>>I guess many systems have given up and bounced mail back to the origin. Not
>>>good news for those who are trying to run a business as it gives the
>>>impression that your email address is invalid or amateur.
>>
>>Most MTAs will not do that for at least 4 days.
>>
> Looks like some is bouncing - following received yesterday:
[snip]
> The MSN ListBot Team
^^^
No surprises here then :)
Mark Evans
> In article <slrn8k6i6e....@muir-et5.staff.demon.net>, Malcolm
> Muir <mal...@demon.net> writes
> [snippage]
>>Do you believe the UK Post Office would give you any guarantee for
>>delivery of mail posted overseas ?
> If a friend sent me a package from a different country, and he received
> it back to himself with a message saying that the British Post Office
> refused to accept it for delivery, then I might start getting a little
> miffed.
> Isn't that what demon are currently doing?
Nope, it's what other MTA's are doing.
Effectivly you are getting the equivalent of the other post office
saying. "Your package missed the flight to the UK and we aren't
prepared to wait for the next one (regardless of what the standards
say.)"
It's one of these situations where the fault isn't where it appears
to be.
Mark Evans
> In article <$5T1+KAl...@gtblakey.demon.co.uk>, Geoff
> <ge...@gtblakey.demon.co.uk> writes
>>In article <slrn8k6i6e....@muir-et5.staff.demon.net>, Malcolm
>>Muir <mal...@demon.net> writes
>>
>>[snippage]
>>>Do you believe the UK Post Office would give you any guarantee for
>>>delivery of mail posted overseas ?
>>
>>If a friend sent me a package from a different country, and he received
>>it back to himself with a message saying that the British Post Office
>>refused to accept it for delivery, then I might start getting a little
>>miffed.
>>
>>Isn't that what demon are currently doing?
> no -- a reasonable analogy (and all analogies with the Internet usually
> break down sooner rather than later) is that the foreign ship is trying
> to dock in Liverpool so that the packages can be unloaded. The port is
> rather busy and so the ship needs to wait until a berth is ready.
Maybe it would be closer to say in this case "The dockers are on strike
but negotiations are going well so they should be back to work soon".
> Most ships will hang around in the roads, but some will sail home and
(There being an accepted minimum time for the ships to wait...)
> report that Britain is "closed".
Mark Evans
> Here is my analysis:
> punt-10: responds almost instantly
> punt-11: A one-or-two minute wait yields a connection on port 25, but no
> banner is displayed for about two or three minutes.
> punt-20: responds almost instantly
> punt-21: will not respond on port 25 at all. Over five minute wait yeilds
> nothing.
In which case what Demon really should do is ensure that no MX records
are pointing to punt-11 or punt-21. Or failing that take down anything
on port 25. Either will ensure that MTA's are not stuck waiting to
deliver.
Mark Evans
> In article <slrn8k82n...@whirlnet.demon.co.uk>, Alan Ford
> <al...@whirlnet.co.uk> writes
>>>>Why are punt-11 and punt-21 *still* down?
> Even if you were right, why do you care ?
>>Here is my analysis:
>>
>>punt-10: responds almost instantly
>>punt-11: A one-or-two minute wait yields a connection on port 25, but no
>> banner is displayed for about two or three minutes.
>>punt-20: responds almost instantly
>>punt-21: will not respond on port 25 at all. Over five minute wait yeilds
>> nothing.
> So long as one of those four responds, mail will get through. That's why
> there are redundant systems in place.
However it is likly to get through rather quicker if any of the 4 which
don't work immediatly refuse TCP connection attempts. :)
Mark Evans
> "Clive D.W. Feather" <cl...@on-the-train.demon.co.uk> writes:
>> Even if you were right, why do you care ?
> Because some systems (eg the vger.rutgers.edu mailing list expanders)
> always send mail to the same punt. They seem to have resolved it once
> and cached the result. I know that this is not optimum behaviour but
> we have no control over the systems used by people attempting to send
> mail to us.
"Not optimum" is a rather mild way to put it. An MTA should always
try all MX records and it certainly shouldn't cache DNS data in excess
of the TTL. Indeed it probably has no business caching DNS at all,
running a copy of BIND on the same machine will do a perfectly good
job of caching DNS data.
IMHO whilst there might be valid claims about Demon doing things
wrong with their email system. Failing to cope with other people's
gross misconfigurations certainly isn't something to blame Demon for.
Mark Evans
> Graham Murray wrote:
>> Because some systems (eg the vger.rutgers.edu mailing list expanders)
>> always send mail to the same punt. They seem to have resolved it once
>> and cached the result. I know that this is not optimum behaviour but
>> we have no control over the systems used by people attempting to send
>> mail to us.
> That's broken behaviour. Did you ask the admin to fix their broken
> server? (Assuming he/she/it knows how, if they're running such a broken
> system in the first place I won't assume they do.)
Might also be an idea to try and locate the author of the software
concerned (assuming it isn't the product of a large company) and
point them in the direction of the relevent RFCs.
Such behaviour sounds more like a fundermental misunderstanding
of the DNS than misconfiguing otherwise working software.
Mark Evans
> "Malcolm Muir" <mal...@demon.net> wrote in message
> news:slrn8k6i6e....@muir-et5.staff.demon.net...
>>
>> >I guess many systems have given up and bounced mail back to the origin.
> Not
>> >good news for those who are trying to run a business as it gives the
>> >impression that your email address is invalid or amateur.
>>
>> Most MTAs will not do that for at least 4 days.
> many MTAs (including demon's) will send warning messages back to the sender
> if the mail is delayed by more than 24 hours (some will warn after only 4
> hours)...
IMHO 24 is still a bit eager. Unless the support for connectivity draws
no distinction between weekdays and weekends. 4 is utterly daft except
where the same support draws no distinction between day and night.
> i know the difference between a failure and a warning... many people don't,
> and mere warnings give a bad impression of the service...
Especially if they are too frequent.
Mark Evans
> On Mon, 12 Jun 2000 16:58:01 +0100, michael lefevre
> <michae...@michaellefevre.com> wrote:
>>
>>and mere warnings give a bad impression of the service...
> Why ?? I'd say they give a _good_ impression. They say:
> "Hi, pal. Welcome to the Internet. I know you expect email to be
> delivered almost instantaneously. Sorry, but for some reason or
> other I couldn't do that right now. You may choose to describe
> that reason as "The internet isn't working". I'm a persistent bugger, so
> I'll keep trying to deliver it for a week for you, before I give
> up and send it back to you."
Is this your custom version? The ones I have experienced are considerably
more "techie" in language.
Mark Evans
> robbie irvine wrote:
>> I'm still seeing variable time on mail - the wife is seeing bugger
>> all, but then most of her mail is mailing lists from egroups and the
>> like.
>>
>> It's hard to tell if my work is affected as the normal recipients have
>> gone home already.
> I seem to be getting a bit more mail today, things are *slightly*
> improving it seems. I guess I'll have to start checking all my mailing
> list subscriptions and resubscribing if necessary (must be nearly four
> days now, when MTAs might just give up on some messages), still, I doubt
> we'll see an apology from Demon for the wasted time this will cause.
Except that it isn't Demon's fault if other people have their
MTA's configured for such short timeouts.
It's another example of "they don't build them like they used to".
Mark Evans
> In article <slrn8k54q...@strontium.scientia.demon.co.uk>, Ben
> Smithurst <b...@scientia.demon.co.uk> writes
>>I don't know what's going on with my FreeBSD lists though, a few things
>>are getting through, so the above idea can't apply to that. I guess
>>I'll have to ask their postmaster if nothing happens soon.
> The server in question had about 10,000 messages for the 'demon.co.uk'
> domain queued up five minutes ago. 381 of these were for me :(
> You can always try 'telnet ..... 25, HELO, ETRN host.demon.co.uk' etc.
AFAIK the punts don't support ETRN.
Mark Evans
> Andy Didcott wrote:
>> Do you dial up to your Demon account on a 24/7 basis ?
>> Does everyone using Demon do this ?
> No, no. But for a temporary measure, I'd be willing to stay online for
> a few hours waiting for MTAs to retry if it meant getting this backlog
> cleared. The suggestion was only half serious anyway.
Also it dosn't help with some of the broken MTA behaviour people
have mentioned.
A possible non temporary setting would be for the highest preference
MX to be the host itself and the punts to be lower preference...
>> I very much doubt it, if the MX records were changed to use our hosts
>> then the mails would bounce if
>>
>> [1] Our machines were not connected when the mails were trying to be
>> delivered.
>>
>> [2] We didn`t have an SMTP server listening on port 25, Although many
>> people probably use turnpike are they all collecting via SMTP ?
> I see Dave Holland attempted to educate you here. These are both
> *temporary* errors. The only thing which will cause the remote MTA to
Also they are both conditions which immediatly return an error, so the
MTA is not stuck in limbo. Which is something which appears to be
happening with some of the punts.
> permanently give up is a 5xx code[0], if you know of a way for a remote
> MTA to get this code when the destination host is unreachable, I'd like
> to hear about it. As the sending host won't get further than connect(),
> I think it's unlikely.
> [0] unless I've forgotten something.
You've forgotten that the writer of an MTA may lack a clue and never
bother to read an rfc :)
Mark Evans
> and...@iname.com "Andy Didcott" writes:
>> I very much doubt it, if the MX records were changed to use our hosts
>> then the mails would bounce if
>>
>> [1] Our machines were not connected when the mails were trying to be
>> delivered.
> Not so: if the punts remained as lower priority (i.e. higher number in
> the MX record) then the sending MTA *should* attempt to deliver to them
> instead of to our own hosts.
> Sadly, however, there are a lot of broken MTAs out there in the real
Do you mean a lot of broken MTAs or many instances of a few broken
MTAs? The former is rather more worrying than the latter...
> world, that *only* attempt delivery to the MX relay with the highest
> priority. In the circumstances, therefore, what you say would be true
I appear to have missed the RFC which says you can do this :)
> *for those MTAs*; OTOH, would one want to receive e-mail from someone who
> runs an MTA so incompetently in the first place?
Unfortunate if it's their ISP or BOFH who is to blame, rather than the
end user...
>> [2] We didn`t have an SMTP server listening on port 25, Although many
>> people probably use turnpike are they all collecting via SMTP ?
> Once again, if a remote MTA attempted to connect to your smtpd when you
> were online, but not actually running one, then the connection would be
> refused --- at which point the MTA *should* attempt to use a different MX
> relay. Once again, the same problems of brain-dead MTAs arises.
Except that the MTA is simply copying the brain-deadedness of the
programmer. The way MX records work is clearly documented (including
special cases such as what to do if several records have the same
pref or there are no MX records, but there is an A record). Its not
even that hard to code, a sorted linked list works perfectly well.
Mark Evans
> Mark Knight wrote:
>> In article <slrn8k54q...@strontium.scientia.demon.co.uk>, Ben
>> Smithurst <b...@scientia.demon.co.uk> writes
>>
>>> I don't know what's going on with my FreeBSD lists though, a few things
>>> are getting through, so the above idea can't apply to that. I guess
>>> I'll have to ask their postmaster if nothing happens soon.
>>
>> The server in question had about 10,000 messages for the 'demon.co.uk'
>> domain queued up five minutes ago. 381 of these were for me :(
>>
>> You can always try 'telnet ..... 25, HELO, ETRN host.demon.co.uk' etc.
> Well, I just tried that, and nothing happened. Perhaps Demon should
> change the MX records to point to our own hosts as their servers are
> obviously broken (I know they won't though).
They'd probably raise objections about it not working where the
machine wasn't running anything on port 25 or had everthing apart
from the punts firewalled.
Also it still wouldn't help where the MTAs attempting to send
are broken.