Windis & KA9Q
John Hein
bodge).
However, as of today, I seem to be suffering from a corrupted Date last
News Received.
I'm wondering if another bodge is required.
--
[ John Hein GM1YME | Phaggots do it on the phone! ]
[ johnd...@drink.demon.co.uk | Sine Pretio Loquimini Omnibus ]
[ johnd...@cix.compulink.co.uk| ]
[ Telephone: +44 131 558 1279 |http://www.scotsgay.co.uk/people/john.html]
[ TeleFax: +44 131 539 2999 | 44 B5/6 f+ t- w+ d g++ k- s++! r-- p ]
[ Lambda BBS: +44 131 556 6316 | S8/9 b g- l y- z/ n o++ x-- a+ u- v- j++ ]
BMG support
> Subject: Windis & KA9Q
>
> I have been using this setup for a number of years (along with the Y2K
> bodge).
That anyone is still using Windis or KA9Q is jaw dropping
BUT BOTH !!!!!!!!!!!!!
AAAAAAAAAmazing...............
John - do tell why you have stayed with them?
You can't still be running a 1991 PC so why use 1991
power software?
--
Roger
Dave Scott
johnd...@drink.demon.co.uk "John Hein" wrote:
> I have been using this setup for a number of years (along with the Y2K
> bodge).
>
> However, as of today, I seem to be suffering from a corrupted Date last
> News Received.
>
> I'm wondering if another bodge is required.
[copy sent by email in case news not getting through]
You don't say what kind of corruption, but I'm betting it's your system
clock being incorrectly reset, 1 hour slow?
From your headers :-
> Date: Mon, 01 Apr 02 18:35:54 GMT
Should of course be "BST", or far better "+0100 (BST)".
Snews requires an environment variable "TZ" to be set, i.e.
TZ=GMT0BST-1,M3.5.0/01:00,M10.5.0/02:00
WinDIS OTOH is set via it's config. tabs, under "time", "timezone"
setting. This may be set to "GMT0BST1" at a guess? That's meant to apply
GMT/BST changeover automatically on the correct dates, but as it works
to USAian changeover rules gets it wrong on certain years. The only sure
way is to either manually switch between "GMT0" and "BST-1" as
appropriate, or see :-
http://www.roborat.demon.co.uk/programs/djsutils.htm#TZSETTER
Also noting your headers :-
> X-Newsreader: Demon Internet Simple News v1.30
That version is itself non-Y2K compliant. Consider upgrading to 1.31, or
even better the washington/roborat version, see :-
http://www.roborat.demon.co.uk/programs/snews.htm
Or of course I may have the wrong end of the stick here...
--
()---()
(0 0) Dave Scott e-mail: da...@roborat.demon.co.uk
----m-( )-m-------------------------------------------------------------
=o=
Raj Rijhwani
<VA.0000006...@businessmailgateway.com>
sup...@BusinessMailGateway.com "BMG support" wrote:
> That anyone is still using Windis or KA9Q is jaw dropping
Why? It works. Which is more than can be said for most of the bloated
proprietary garbage released since...
> You can't still be running a 1991 PC so why use 1991
> power software?
What makes you so sure?
--
Raj Rijhwani (umtsb5/16) | This is the voice of the Mysterons...
r...@rijhwani.org | ... We know that you can hear us Earthmen
http://www.rijhwani.org/raj/ | "Lieutenant Green: Launch all Angels!"
BMG support
> From: r...@rijhwani.org (Raj Rijhwani)
> Newsgroups: demon.ip.support.winsock
> Why? It works. Which is more than can be said for most of the bloated
> proprietary garbage released since...
Actually it doesn't work WINDIS has a killer bug in it when it handles
email. KA9Q is a dos program which required DOS based TCP stack to
run. And Windis is 16bit ugh....
I remember John's posts on CIX way back before there was a Demon.
AH happy days.
Look I am not biased against old appropriate software, the news reader
that it use (VA) started around that time. (Called Wigwam then).
But it has been (just about) kept up to date.
That the Windis and Ka9q combination is amazing :)
--
Roger
Dr John Stockton
news:demon.ip.support.winsock, Dave Scott <da...@roborat.demon.co.uk>
wrote at Tue, 2 Apr 2002 00:39:02 :-
>
>Snews requires an environment variable "TZ" to be set, i.e.
>TZ=GMT0BST-1,M3.5.0/01:00,M10.5.0/02:00
Thanks - just what I needed, in a timely manner. Not for Snews; but for
the Gawk I got from Garbo, which is preset to California.
--
© John Stockton, Surrey, UK. j...@merlyn.demon.co.uk Turnpike v4.00 MIME. ©
Web <URL:http://www.merlyn.demon.co.uk/> - w. FAQish topics, links, acronyms
PAS, EXE in <URL:http://www.merlyn.demon.co.uk/programs/> - see 00index.txt.
Dates - miscdate.htm moredate.htm js-dates.htm pas-time.htm critdate.htm &c.
John McCabe
<sup...@BusinessMailGateway.com> wrote:
>Actually it doesn't work WINDIS has a killer bug in it when it handles
>email.
Explain please.
> And Windis is 16bit ugh....
I think you'll find WinDIS is 32 bit, and has been since it's last
release in around 1995.
John Hein
sup...@BusinessMailGateway.com "BMG support" writes:
>
>> Subject: Windis & KA9Q
>
>>
>> I have been using this setup for a number of years (along with the Y2K
>> bodge).
>
>That anyone is still using Windis or KA9Q is jaw dropping
>
>BUT BOTH !!!!!!!!!!!!!
>
>AAAAAAAAAmazing...............
>
>John - do tell why you have stayed with them?
Because they do what I want them to.
What's the point in updating when I have no need to?
>
>You can't still be running a 1991 PC so why use 1991
>power software?
>
I still run a BBC B and Z80 2nd Processor.
>
>--
>Roger
John Hein
sup...@BusinessMailGateway.com "BMG support" writes:
>In article <20020402....@rijhwani.org>, Raj Rijhwani wrote:
>> From: r...@rijhwani.org (Raj Rijhwani)
>> Newsgroups: demon.ip.support.winsock
>
>> Why? It works. Which is more than can be said for most of the bloated
>> proprietary garbage released since...
>
>Actually it doesn't work WINDIS has a killer bug in it when it handles
>email.
What killer bug is this?
KA9Q is a dos program which required DOS based TCP stack to
>run. And Windis is 16bit ugh....
>
>I remember John's posts on CIX way back before there was a Demon.
>AH happy days.
>
>Look I am not biased against old appropriate software, the news reader
>that it use (VA) started around that time. (Called Wigwam then).
>But it has been (just about) kept up to date.
>
>That the Windis and Ka9q combination is amazing :)
>
>--
>Roger
>
>
--
BMG support
> >Actually it doesn't work WINDIS has a killer bug in it when it handles
> >email.
>
> What killer bug is this?
> >Actually it doesn't work WINDIS has a killer bug in it when it handles
> >email.
>
> Explain please.
>
> > And Windis is 16bit ugh....
>
> I think you'll find WinDIS is 32 bit, and has been since it's last
> release in around 1995.
>
First, thanks to John McCabe, for emailing me about these posts
which deserve a response from me.
And agreed windis32 was indeed issued as a 32 bit app.
Now to the (email) bug.
The last issue of windis32 was 3.1, which I used up until the 1997 when
the bug killed my then companies mail system. :( The bug is in
the "full stop" detector. I documented this back in Jan 97
and file describing the issue is still there.
ftp://ftp.demon.co.uk/pub/ibmpc/win95/winsock/apps/windis/windis32_v3_1_BUG.txt
It's short enough to quote in full (see at end).
If that seem's unimportant get someone to send you an email with a
line in the middle WITH only a single fullstop on it. {NOT HTML of course}
And ouch ....
If email is all you need then there a number of low cost mail gateways
around. Including one that is specifically designed to work with Demon
(either by SMTP delivery or by POP3 collection) and is currently FREE.
(And yes, I am involved with its development.)
I hope this answers your question and thanks again for the prompt.
--
Roger
------------------------------- full quote of bug report ------------------------
This file should be placed as below (please)
ftp://ftp.demon.co.uk/public/pub/ibmpc/win95/winsock/apps/windis/windis32_v3_1_BUG.txt
In it with great sadness that I have to report a data
corrupting bug in
windis32 v3.1 (and windis32 v3.0)
Description
===========
Any data lines in your email that start with a "." ie a full stop
or dot will have the first dot removed as they pass through
winds32 v3.1
If your data contains any line that is ONLY a "." then this will
be taken as the end of you email!!!! the rest of that email will
be discarded !!!!!!!!!!
Cause - it is part of the SMTP protocol RFC 821 section 4.5.2
that transmission is terminated by having a line containing ONLY a
full stop. To allow real lines that have only a full-stop
additional full-stops need to be added and later remove from the
front of certain lines. Needless to say the bug is in this area.
As it does NOT manifest with all emailers at all times!!! Here is
proof. Forward this email to xxx@yy it will then sit in the
windis32 output queue dir. normally called mqueue. If you then
tell windis that the SMTP gateway is its self then every time you
push the send button this email will go out of and into windis32.
And on every cycle it looses a dot from the front of each line.
Inspect it in the dir.
...four dots
..three dots
.two dots
one dot
If the number of dots given above is not correct - that’s because
I sent this email via WinDis32 !!! :(
NOTE the problem WILL NOT BE SEEN IN THE FTP COPY (of course)
If anyone can successfully contact the original authors and get
them to get permission to do, and then do, a windis32 v3.2 it is
much needed.
In the mean time use at your peril.
--
end Jan 1997
Raj Rijhwani
john....@emrad.ns.com "John McCabe" wrote:
> On Tue, 02 Apr 2002 10:53:56 +0100, BMG support
> <sup...@BusinessMailGateway.com> wrote:
>
> >Actually it doesn't work WINDIS has a killer bug in it when it handles
> >email.
> Explain please.
The flag for e-mail rejection is a global, so when any one e-mail being
received triggers a rejection, odds are that every e-mail being received
concurrently will be summarily rejected.
> > And Windis is 16bit ugh....
> I think you'll find WinDIS is 32 bit, and has been since it's last
> release in around 1995.
Better not be. I'm running it on Win 3.11 here...
John McCabe
Rijhwani) wrote:
>> >Actually it doesn't work WINDIS has a killer bug in it when it handles
>> >email.
>
>> Explain please.
>
>The flag for e-mail rejection is a global, so when any one e-mail being
>received triggers a rejection, odds are that every e-mail being received
>concurrently will be summarily rejected.
First I've heard of that one.
>> > And Windis is 16bit ugh....
>> I think you'll find WinDIS is 32 bit, and has been since it's last
>> release in around 1995.
>Better not be. I'm running it on Win 3.11 here...
Well there are different strains :-) What I should have said is that
there is a native Win32 version available as well as a Win16 one. Same
source though as far as I'm aware (conditionally compiled).
John McCabe
McCabe) wrote:
>On Tue, 02 Apr 2002 10:53:56 +0100, BMG support
><sup...@BusinessMailGateway.com> wrote:
>
>>Actually it doesn't work WINDIS has a killer bug in it when it handles
>>email.
>
>Explain please.
I haven't seen Roger's reply arrive on the news server yet, but I have
been speaking to him by email and he's explained that there is a bug
in WinDIS related to the transmission and receipt of data that has
period character '.' at the start of a line.
According to RFC2821 (the SMTP) one, messages transmission from a
client is completed using a <CRLF>.<CRLF> sequence, i.e. a "." on a
line by itself. This should be sent by the SMTP client. A message
however may want to have a "." on a line by itself, so RFC2821 defines
a scheme whereby the SMTP client adds an extra "." at the start of any
line that alread has a "." as the first character. It would appear
that WinDIS32 V3.1 implements this incorrectly and only inserts an
extra "." if the line contains *only* a "."!
For example:
If the message contains a line:
.
WinDIS should send
..
To the remote server. That's implemented correctly.
If the message contains:
. Hello
WinDIS should send
.. Hello
To the remote server. It doesn't - it sends:
. Hello
and the remote server, based on the rules in RFC2821, will strip off
the leading "." and deliver:
Hello
to the recipient.
For anyone using a different version of WinDIS, it would be
interesting to know if this has been introduced at a later stage. For
example does WinDIS 2.x handle this correctly (i.e. anything before
June 1995)
John McCabe
<sup...@BusinessMailGateway.com> wrote:
Roger
I've just had a go with this again with a fresh version of WinDis32
V3.1 downloaded from the Demon website. Note that the version ident in
the Help => About box is:
Version 3.1 (ml. 08) Built on Dec 11 1995 16:19:49
>------------------------------- full quote of bug report ------------------------
>Description
>===========
>Any data lines in your email that start with a "." ie a full stop
>or dot will have the first dot removed as they pass through
>winds32 v3.1
That is, incoming email will have the initial dot removed as required
by RFC2821 section 4.5.2.
>If your data contains any line that is ONLY a "." then this will
>be taken as the end of you email!!!! the rest of that email will
>be discarded !!!!!!!!!!
I can't see that happening with the version above.
>As it does NOT manifest with all emailers at all times!!! Here is
>proof. Forward this email to xxx@yy it will then sit in the
>windis32 output queue dir. normally called mqueue. If you then
>tell windis that the SMTP gateway is its self then every time you
>push the send button this email will go out of and into windis32.
>And on every cycle it looses a dot from the front of each line.
>Inspect it in the dir.
>
>
>...four dots
>..three dots
>.two dots
>one dot
>
>If the number of dots given above is not correct - that’s because
>I sent this email via WinDis32 !!! :(
That, of course, does happen because this version of WinDis only
inserts an extra "." on lines that contain *only* a ".", rather than
on any line starting with a "." as required by RFC2821.
BMG support
> From: john....@emrad.ns.com (John McCabe)
> Newsgroups: demon.ip.support.winsock
> Subject: Re: Windis & KA9Q
> Date: Fri, 05 Apr 2002 12:56:51 GMT
>
> On Thu, 04 Apr 2002 11:49:41 +0100, BMG support
> <sup...@BusinessMailGateway.com> wrote:
>
> Roger
>
> I've just had a go with this again with a fresh version of WinDis32
> V3.1 downloaded from the Demon website. Note that the version ident in
> the Help => About box is:
>
> Version 3.1 (ml. 08) Built on Dec 11 1995 16:19:49
>
[snips]
> ... this version of WinDis only
> inserts an extra "." on lines that contain *only* a ".", rather than
> on any line starting with a "." as required by RFC2821.
>
Yep - and that means that it can and does damage data as it passes
through. Any email system that does that is .......
I'll leave you to fill in the blank. :) and :(
--
Roger
John McCabe
>Yep - and that means that it can and does damage data as it passes
>through.
Oh yes, that is true, but it doesn't seem to be detecting an end of
message incorrectly which is what your earlier message suggested.
Of course if the damaged information is used in some other program....
Best Regards
John McCabe <jo...@assen.demon.co.uk>