K9 forum?

  • Thread starter Thread starter ms
  • Start date Start date
M

ms

I'm just starting with K9.
In Configuration, I can't enter the username as it states, as K9 won't
accept that number of characters in the entry slot.

On the website is no reference to a forum.

Is there a URL for the K9 forum?

Mike Sa
 
ms said:
I'm just starting with K9.
In Configuration, I can't enter the username as it states, as K9 won't
accept that number of characters in the entry slot.

On the website is no reference to a forum.

Is there a URL for the K9 forum?

Mike Sa

K9 doesn't have a forum.

If you mail me your e-mail address (don't post it on the forum), I'll be
happy to send you the authors e-mail address.

--
Regards

Steven Burn
Ur I.T. Mate Group CEO
www.it-mate.co.uk

Disclaimer:
I know I'm probably wrong, I just like taking part ;o)
 
I'm just starting with K9.
In Configuration, I can't enter the username as it states, as K9 won't
accept that number of characters in the entry slot.

Eh? You don't enter a user name into K9, it's just a proxy. Are you talking
about the "Server/Port/User Delimiter" field? That's only a single
character, which defaults to "/". The changes are made to your mail
program, not K9.

For example, say I am user JOE on mail server POP.ROGERS.COM.

Before K9
user: JOE
server: POP.ROGERS.COM

After K9
user: POP.ROGERS.COM/110/JOE
server: localhost

Just make sure that your mail program goes to port 9999 on your PC for mail,
rather than port 110.

On the website is no reference to a forum.

Is there a URL for the K9 forum?

There's no web forum, but Robin posts on the grc.spam newsgroup of the
news.grc.com news server.
 
H does?..... cool (I wasn't aware of that)

--
Regards

Steven Burn
Ur I.T. Mate Group CEO
www.it-mate.co.uk

Disclaimer:
I know I'm probably wrong, I just like taking part ;o)
 
I'll try again ;) Make your messages easier to read
and understand (here: snip - and - don't top post, provide context
first - then your reply) See for example:
<http://home.online.no/~shughes/a57998/quote.html#004>

(no - I am not out to "get you" <g>)

All the best,
Bjorn Simonsen

lol, I was on the phone at the time so was more concerned with not screwing
up my spelling etc (which I seem to have done anyway). I'll remember not to
scew up next time ;o)

--
Regards

Steven Burn
Ur I.T. Mate Group CEO
www.it-mate.co.uk

Disclaimer:
I know I'm probably wrong, I just like taking part ;o)
 
William said:
Eh? You don't enter a user name into K9, it's just a proxy. Are you talking
about the "Server/Port/User Delimiter" field? That's only a single
character, which defaults to "/". The changes are made to your mail
program, not K9.

For example, say I am user JOE on mail server POP.ROGERS.COM.

Before K9
user: JOE
server: POP.ROGERS.COM

After K9
user: POP.ROGERS.COM/110/JOE
server: localhost

Just make sure that your mail program goes to port 9999 on your PC for mail,
rather than port 110.


There's no web forum, but Robin posts on the grc.spam newsgroup of the
news.grc.com news server.

Thanks for the data.
Ver. 1.15
In /Advanced:
Account Name: I assume this can be my normal Earthlink name.

Server Name/IP: 127.0.0.1:9999 (this from another post when I use
Netscape, is the :9999 correct?)

Port: 9999

User name: here is the problem, according to the help file, it should
look like- pop.west.cox.net/110/robin, in my case:
pop.onemain.com/110/XXXXXXXXX where X is my account number.

I get as far as pop.onemain.com/110/, then additional characters move to
the beginning of the line, cutting off the end of the previous entry.
It's like an absolute character limit, when the help file example is
longer than the window space anyway?

above in your post:
Eh? You don't enter a user name into K9, it's just a proxy. Are you talking
about the "Server/Port/User Delimiter" field? That's only a single
character, which defaults to "/".
Yes, that's the default on the /Configuration page.
The changes are made to your mail program, not K9.
I don't understand this comment, if K9 is the first interface with my
ISP server, the user name string tells K9 how to receive my mail?

Advice?

Mike Sa
 
Thanks for the data.
Ver. 1.15
In /Advanced:
Account Name: I assume this can be my normal Earthlink name.

Server Name/IP: 127.0.0.1:9999 (this from another post when I use
Netscape, is the :9999 correct?)

No, just 127.0.0.1. You can specify the mail server by name (say
POP.BROADBAND.ROGERS.COM) or by IP (66.185.95.101), but not both. In the
case of K9, you can use the server name of LOCALHOST or the IP address of
127.0.0.1. You don't (and can't) use both.
Port: 9999
Correct.

User name: here is the problem, according to the help file, it should
look like- pop.west.cox.net/110/robin, in my case:
pop.onemain.com/110/XXXXXXXXX where X is my account number.

I get as far as pop.onemain.com/110/, then additional characters move to
the beginning of the line, cutting off the end of the previous entry.
It's like an absolute character limit, when the help file example is
longer than the window space anyway?

I'm afraid I don't know the Netscape mailer well enough to answer this one.
Sorry.
I don't understand this comment, if K9 is the first interface with my
ISP server, the user name string tells K9 how to receive my mail?

Correct. What happens is that Netscape (or any mailer) fetches mail by
connecting to MAILSERVER on PORT and using the POP3 protocol to get your
mail. Using (e-mail address removed) as an example, Netscape would connect to
POP.BROADBAND.ROGERS.COM, and log in as "JOE" to get the mail. By default,
this all happens on TCP/IP port 110.

With K9 in place, Netscape never talks to POP.BROADBAND.ROGERS.COM directly.
Instead, it talks to K9, and K9 talks to the Rogers server. Other than
changing the server name, port, and username, Netscape sees no difference.

What you normally would have is

Netscape ---(110)---> POP.BROADBAND.ROGERS.COM ("Hi, I'm JOE, password XXX")
ROGERS --(Netscape)--> "Hi Joe, here's your mail"

With K9, you have

Netscape ---(9999)---> LOCALHOST ("Hi, I'm POP.BROADBAND.ROGERS.COM/110/JOE
password XXX")
K9 ---(110)----> POP.BROADBAND.ROGERS.COM ("Hi, I'm JOE, password XXX")
ROGERS --(K9)--> "Hi Joe, here's your mail"
K9 ---(Netscape)---> "Hi Joe, here's your mail; spam is marked with [spam]")


Since the POP3 protocol only allows for username and password, there has to
be a way to tell K9 which mail server and port to use. So Robin encodes it
into the username that you pass to K9, and K9 parses it.

I'm afraid that with respect to Netscape mail. But changing the server name
to just LOCALHOST and removing the port will help, but I don't know about
why you are limited in your field entry. Unfortunately, you are using
Netscape 4, which is fairly buggy, so it could well be that this is a bug
in the Netscape mailer.
 
Steven said:
K9 doesn't have a forum.

If you mail me your e-mail address (don't post it on the forum), I'll be
happy to send you the authors e-mail address.

--
Regards

Steven Burn
Ur I.T. Mate Group CEO
www.it-mate.co.uk

Disclaimer:
I know I'm probably wrong, I just like taking part ;o)

Thanks, but I have author's address from the web page, he states he
likely will not answer emails, certainly not newbie questions like mine.

Mike Sa
 
I've notice from past posts, you seem very familiar with K9.
No, just 127.0.0.1. You can specify the mail server by name (say
POP.BROADBAND.ROGERS.COM) or by IP (66.185.95.101), but not both. In the
case of K9, you can use the server name of LOCALHOST or the IP address of
127.0.0.1. You don't (and can't) use both.


I'm afraid I don't know the Netscape mailer well enough to answer this one.
Sorry.
I was talking about K9, as Netscape is properly configured as my mail
client.
See below:
I don't understand this comment, if K9 is the first interface with my
ISP server, the user name string tells K9 how to receive my mail?

Correct. What happens is that Netscape (or any mailer) fetches mail by
connecting to MAILSERVER on PORT and using the POP3 protocol to get your
mail. Using (e-mail address removed) as an example, Netscape would connect to
POP.BROADBAND.ROGERS.COM, and log in as "JOE" to get the mail. By default,
this all happens on TCP/IP port 110.

With K9 in place, Netscape never talks to POP.BROADBAND.ROGERS.COM directly.
Instead, it talks to K9, and K9 talks to the Rogers server. Other than
changing the server name, port, and username, Netscape sees no difference.

What you normally would have is

Netscape ---(110)---> POP.BROADBAND.ROGERS.COM ("Hi, I'm JOE, password XXX")
ROGERS --(Netscape)--> "Hi Joe, here's your mail"

With K9, you have

Netscape ---(9999)---> LOCALHOST ("Hi, I'm POP.BROADBAND.ROGERS.COM/110/JOE
password XXX")
K9 ---(110)----> POP.BROADBAND.ROGERS.COM ("Hi, I'm JOE, password XXX")
ROGERS --(K9)--> "Hi Joe, here's your mail"
K9 ---(Netscape)---> "Hi Joe, here's your mail; spam is marked with [spam]")

Since the POP3 protocol only allows for username and password, there has to
be a way to tell K9 which mail server and port to use. So Robin encodes it
into the username that you pass to K9, and K9 parses it.

I'm afraid that with respect to Netscape mail. But changing the server name
to just LOCALHOST and removing the port will help, but I don't know about
why you are limited in your field entry. Unfortunately, you are using
Netscape 4, which is fairly buggy, so it could well be that this is a bug
in the Netscape mailer.
Update:
I checked Enable this Account, so I have
Account Name: Earthlink
Server Name: LOCALHOST
Port: 9999
User Name: I kept having the above problem again, entered
pop.onemain.com/110/abcd.. and the next 5 numbers started at the
beginning of the line, again.
BUT, I opened it a minute later, and the username entry is correct!
It is my version of the example- pop.west.cox.net/110/robin, where in
place of "robin", I have my 9-didget account number.

At this point, I've made no changes to Netscape for K9...? I was not
sure from the above if my mail client has to change to suit K9.

Your reply above-
I'm JOE, password XXX")- IMO you meant account number, not password? The password goes in another entry.

Now that I want to try K9, there is no spam, so I sent myself an email.
It is on the server.
In K9, I clicked Read, nothing happens, it does not see the email, so I
still have a configuration issue. (?)

For me, Netscape 4.79 has always been reliable.

Thanks,

Mike Sa
 
In 1.13 settings would be something like ;

Earthlink

pop.west.cox.net

110

msa

password

This is to do a server check. Should be the same in 1.15.

At this point, I've made no changes to Netscape for K9...? I was not
sure from the above if my mail client has to change to suit K9.

If you don't change your emailer settings then the emailer will not
use k9.

Your POP server in your emailer should be 127.0.0.1

Username something like ;

pop.west.cox.net/110/msa

Regards, John.

--
****************************************************
,-._|\ (A.C.F FAQ) http://clients.net2000.com.au/~johnf/faq.html
/ Oz \ John Fitzsimons - Melbourne, Australia.
\_,--.x/ http://www.aspects.org.au/index.htm
v http://clients.net2000.com.au/~johnf/
 
John said:
In 1.13 settings would be something like ;

Earthlink

pop.west.cox.net

110

msa

password

This is to do a server check. Should be the same in 1.15.



If you don't change your emailer settings then the emailer will not
use k9.

Your POP server in your emailer should be 127.0.0.1

Username something like ;

pop.west.cox.net/110/msa

Regards, John.

--
****************************************************
,-._|\ (A.C.F FAQ) http://clients.net2000.com.au/~johnf/faq.html
/ Oz \ John Fitzsimons - Melbourne, Australia.
\_,--.x/ http://www.aspects.org.au/index.htm
v http://clients.net2000.com.au/~johnf/

Thanks, John.

No luck when I changed Netscape to interface with K9.

As I understand it, K9 should not care if the mail client is properly
configured, as K9 is the first interface with the server.

As in my earlier post, IMO K9 is now setup OK in the Advanced page, I
enabled the account parameters so it should work...... but when I click
Read, it still does nothing, so it does not see the server.

Pending comments here, I will post to the grc.spam newsgroup, and see if
I can get it to work.

Mike Sa
 
I've notice from past posts, you seem very familiar with K9.

Actually, I've only been using it for a month.
I was talking about K9, as Netscape is properly configured as my mail
client.

Ah, *now* I understand what's been going on. You've been talking about
configuring the "POP3 Server Check" function in the Advanced tab of K9,
right? Okay, that's why things aren't working, we're talking about two
different things.

K9 has two basic features:
1) spam filtering, and
2) server status checking for APOP accounts

I've been talking about (1), you've been talking about (2).

Okay, here's what to do:

First, in K9, under the POP3 server check, disable (uncheck) the POP3 Server
Check. You don't need it; it's optional, and we can always enable that
later, after we get the more important issue of your email going.

Secondly, in the Netscape mailer, change your user name to
"pop.onemain.com/110/abcd", the server name to LOCALHOST, and the port to
9999. If you can't set the port to 9999 (I don't know if Netscape 4
supported that or not), then in the K9 Configuration tab, change the "Listen
on Port" from 9999 to 110.
At this point, I've made no changes to Netscape for K9...? I was not
sure from the above if my mail client has to change to suit K9.

Yup, as above. You've been configuring the server check feature, not the
spam filter proxy.
Now that I want to try K9, there is no spam, so I sent myself an email.
It is on the server.
In K9, I clicked Read, nothing happens, it does not see the email, so I
still have a configuration issue. (?)

The "Read" button on both the Recent Emails and Storage Area sections is
used to expand an individual message from the database of mail already in
K9. It doesn't pull mail from the server.

Okay, now that we're talking about the same thing, things should go a lot
smoother. Make the changes above and let us know how it's working out.
 
As I understand it, K9 should not care if the mail client is properly
configured, as K9 is the first interface with the server.

K9 cares very much that the mailer is properly configured, because it's the
mail client that pulls the mail from the server. If the mail client doesn't
talk to K9, K9 won't do a thing, because K9 won't pull email off a server by
itself.
 
John Fitzsimons wrote:

< snip >

Hi Mike,
Thanks, John.
No luck when I changed Netscape to interface with K9.
As I understand it, K9 should not care if the mail client is properly
configured, as K9 is the first interface with the server.

It cares because there are two ways of K9 working. Directly with the
server via the "Advanced" configuration tab and via the emailer on
the "Configuration" tab.
As in my earlier post, IMO K9 is now setup OK in the Advanced page, I
enabled the account parameters so it should work...... but when I click
Read, it still does nothing, so it does not see the server.

Well, it has to have something to read. Do you get what is on the
server listed BEFORE you select the item you want to read ?

Very hard to "read" something that isn't there !

You press the first button "Check" before you select the email you
want to read.
Pending comments here, I will post to the grc.spam newsgroup, and see if
I can get it to work.

Regards, John.

--
****************************************************
,-._|\ (A.C.F FAQ) http://clients.net2000.com.au/~johnf/faq.html
/ Oz \ John Fitzsimons - Melbourne, Australia.
\_,--.x/ http://www.aspects.org.au/index.htm
v http://clients.net2000.com.au/~johnf/
 
William said:
Actually, I've only been using it for a month.


Ah, *now* I understand what's been going on. You've been talking about
configuring the "POP3 Server Check" function in the Advanced tab of K9,
right? Okay, that's why things aren't working, we're talking about two
different things.

K9 has two basic features:
1) spam filtering, and
2) server status checking for APOP accounts

I've been talking about (1), you've been talking about (2).

Okay, here's what to do:

First, in K9, under the POP3 server check, disable (uncheck) the POP3 Server
Check. You don't need it; it's optional, and we can always enable that
later, after we get the more important issue of your email going.

Secondly, in the Netscape mailer, change your user name to
"pop.onemain.com/110/abcd", the server name to LOCALHOST, and the port to
9999. If you can't set the port to 9999 (I don't know if Netscape 4
supported that or not), then in the K9 Configuration tab, change the "Listen
on Port" from 9999 to 110.


Yup, as above. You've been configuring the server check feature, not the
spam filter proxy.


The "Read" button on both the Recent Emails and Storage Area sections is
used to expand an individual message from the database of mail already in
K9. It doesn't pull mail from the server.

Okay, now that we're talking about the same thing, things should go a lot
smoother. Make the changes above and let us know how it's working out.
Sorry it took so long to reply.
Thanks for the help. I will save the data for later (see below):

As you described:
K9 requires that I change settings in my Netscape mail client. Also, I
need to create a Netscape filter to automate the process. When it is
working, a message comes in, K9 changes the header if it thinks it's
spam, the message goes to Netscape, the filter puts it in a separate
folder. When it's all done, I would want to check that folder and decide
for sure if it's spam by the header, not wanting to open it, because now
it's on my hard drive.
If I change my mail client, I need to change it's settings the same way.

Presently, I use Popcorn. I look on the server at the mail header. If
there's a question, I open the message on the server, and if it's spam,
I can delete it there. The only mail I download from the server is
"desired" messages. So far, this has worked well. Presently, no changes
needed to my mail client, if I change to another one, no problem as
Popcorn is independent of the mail client.

Either way, I would have to check the header manually to be sure it's
spam. In my situation, it looks like Popcorn is an easier way to go for
now.

Please comment on the above.

Thanks,

Mike Sa
 
As you described:
K9 requires that I change settings in my Netscape mail client. Also, I
need to create a Netscape filter to automate the process. When it is
working, a message comes in, K9 changes the header if it thinks it's
spam, the message goes to Netscape, the filter puts it in a separate
folder.
Correct.

When it's all done, I would want to check that folder and decide
for sure if it's spam by the header, not wanting to open it, because now
it's on my hard drive.
If I change my mail client, I need to change it's settings the same way.
Correct.

Presently, I use Popcorn. I look on the server at the mail header. If
there's a question, I open the message on the server, and if it's spam,
I can delete it there. The only mail I download from the server is
"desired" messages. So far, this has worked well. Presently, no changes
needed to my mail client, if I change to another one, no problem as
Popcorn is independent of the mail client.

I used to use Magic Mail Monitor (http://mmm3.sourceforge.net) the same way.
Either way, I would have to check the header manually to be sure it's
spam. In my situation, it looks like Popcorn is an easier way to go for
now.

Please comment on the above.

If you're looking to filter only on subject, then none of the Bayesian
filters (K9, POPFile, Spamihilator, SpamPal, etc.) will do what you want. By
definition, a Bayesian filter bases its' decision as to whether or not a
message is spam from reading the content of the message. That means messages
*must* be downloaded to your PC. If you are trying to save bandwidth by not
downloading messages, then Bayesian is not the way to go.

For my part, I used to use Magic Mail Monitor to kill offending messages off
the server via a whitelist (known good), delete list (known bad), blacklist
(suspect) rule set. Whitelisted messages, such as known senders, or a
subject of "NEWSGROUP POSTING" for this account, always pass. Delete lists
were known bad posters, such as @hotmail.com, or subject lines. Blacklist
was presumed spam, and downloaded to the mail client for processing. This
worked, except that spam that arrived in between the MMM3 checks every 10
minutes would go to the client. It also meant occassional race conditions,
if both MMM3 and the client tried to pop mail at the same time.

When my mail (The Bat; a commercial product) upgraded to v2 recently, it
included a number of features that allowed it to do much of what MMM3 did
externally, so I stopped using MMM3 and only use K9 and the Bat. It sounds
like you're doing the same thing I did.

What I found was that the white/black/delete lists were accurate to about
88%. That mean 10%-12% of spam got through; about 50 or so a day. Of course,
even with v1 of the Bat, 45 of those 50 were identified. With K9, I'm
sitting at 99.2%, and there's no 10 minute window where stuff can slip past
MMM3 any more.

In the end, it comes down to just doing what you are comfortable with. Some
people value bandwidth more (especially if on dialup), others are more
interested in convenience. There's no "right" answer, it's individual
preference.
 
William said:
On Thu, 13 Nov 2003 17:14:44 GMT, ms wrote: snip

I used to use Magic Mail Monitor (http://mmm3.sourceforge.net) the same way.
snip
For my part, I used to use Magic Mail Monitor to kill offending messages off
the server via a whitelist (known good), delete list (known bad), blacklist
(suspect) rule set. Whitelisted messages, such as known senders, or a
subject of "NEWSGROUP POSTING" for this account, always pass. Delete lists
were known bad posters, such as @hotmail.com, or subject lines. Blacklist
was presumed spam, and downloaded to the mail client for processing. This
worked, except that spam that arrived in between the MMM3 checks every 10
minutes would go to the client. It also meant occassional race conditions,
if both MMM3 and the client tried to pop mail at the same time.

When my mail (The Bat; a commercial product) upgraded to v2 recently, it
included a number of features that allowed it to do much of what MMM3 did
externally, so I stopped using MMM3 and only use K9 and the Bat. It sounds
like you're doing the same thing I did.

What I found was that the white/black/delete lists were accurate to about
88%. That mean 10%-12% of spam got through; about 50 or so a day. Of course,
even with v1 of the Bat, 45 of those 50 were identified. With K9, I'm
sitting at 99.2%, and there's no 10 minute window where stuff can slip past
MMM3 any more.

In the end, it comes down to just doing what you are comfortable with. Some
people value bandwidth more (especially if on dialup), others are more
interested in convenience. There's no "right" answer, it's individual
preference.
That explains the value of K9, looking in the message text.

I have DUN, but it's not a bandwidth issue. Due to age/other issues, I
like to make computing easier if I can.

Now, I get about 5 spams a day, so it's manageable, you're getting a lot
more.
Good data on Magic Mail, I'll look at it again, as it will certainly
help. Hotmail.com is not a clear cut item for me as some ebay contacts
and some of my wife's friends use hotmail, sad to say. I tell them it's
a risk, but they don't know to be concerned.

Mike Sa
 
I have DUN, but it's not a bandwidth issue. Due to age/other issues, I
like to make computing easier if I can.

Aren't we all.
Now, I get about 5 spams a day, so it's manageable, you're getting a lot
more.

5 a day? I have recieved 2 a *minute* in the past. Judicious mods to the
mail server helped, but I still get 50 a day. Of course, now I have to look
at the K9 statistics page to quote that, rather than knowing it.
Good data on Magic Mail, I'll look at it again, as it will certainly
help. Hotmail.com is not a clear cut item for me as some ebay contacts
and some of my wife's friends use hotmail, sad to say. I tell them it's
a risk, but they don't know to be concerned.

I had the same. That's why I used the whitelist/blacklist/deletelist scheme
I mentioned. If (e-mail address removed) and (e-mail address removed) are the only people on
Hotmail I want to let in, then I put both of their accounts in the
whitelist.from file, and "@hotmail.com" in the deletelist.from file.
Anything in the whitelist is passed through, so Joe and Jane could still
email me, but everyone else at hotmail.com can't get through, except when
they send in the period between the time MMM3 checks the server and my
mailer checks the server.

I don't really care what anti-spam measures people take, as long as they
take something.
 
Back
Top