Getting Started

[StevenUnderwood 0]

Julian, several questions about the new system before I dive in.

I have a free account for reporting spam from my work servers and a paid email account for reporting from my home accounts as I want notifications going to different servers.

I have not converted my account to the cookie style because I regularly submit from many different systems (5 at work, 2 at home) that I do not control completely and prefer the HTML auth. Will accepting the cookie here stop me from submitting spam using HTML login? I read it as being YES.

I submit both work and home spam on the same machines. Will I be able to have 2 cookies on the machine at one time?

My home email account has 4 poboxes.com addresses forwarded to it and a yahoo accont POPped. I know that the spamcop email account will be configured when I follow the link (and accept the cookie) and think that I need to configure each of the four poboxes.com addresses even though they are aliases for the main poboxes.com account. Do I need to do anything for the yahoo account?

My work email gets forwarded through postini.com which sorts out most of the spam, then forwards it on to my server. They are our primary MX and we have no contact with the messages there. Will those intermediate servers be "authorized" automatically?

What happens when something happens and either postini or my work kicks over to the secondary servers? This happens every Sunday for the work servers as the main servers are offline for backups and the secondary server accepts the mail until the primary is back online. Postini has brought their secondary site online a couple of times this past year as well. They are the 3rd and 4th MX for our domain and the web page says these failed when I tried from work, which is correct because they are not active.

My work account is used to report spam that my users complain about. There are 5 accounts that I regularly report as they are role accounts that can not send, only receive email. There is also my main account, which I regularly report all messages from. However, regularly one of my 250 users complains about a specific spam message, usually porn related, that I offer to report for them. Will I be able to report these or is this one of the instances where "Some "unique" users may not be able to report all the spam they have in the past."?

Will we still be able to parse messages from these other accounts and simply not report through the system? I need to manually paste these into the parser because I have never been able to get Lotus Notes to forward the message in a way acceptable to the parser and I need to export the message to a text file anyway.

Thanks for the improvements

Edited March 18, 2004 by StevenUnderwood

[julian 0]

> Julian, several questions about the new system before I dive in.

I'd prefer you dive in first, then ask me any *remaining* questions. There *is* a method to back out if you don't like it.

> I have not converted my account to the cookie style because I regularly submit from many different systems (5 at work, 2 at home) that I do not control completely and prefer the HTML auth.

There is no conversion required. You can mix & match login styles. For this trial, you *must* take the cookie.

> Will accepting the cookie here stop me from submitting spam using HTML login?

HTTP login? No. Mix and match. Do whatever fits your mood.

> I submit both work and home spam on the same machines. Will I be able to have 2 cookies on the machine at one time?

No, but you can report spam for both accounts with one login. Or you can log in & out as you please. You can't be logged in as two different users using either method.

[snip boring description of complex forwarding rules]

Hopefully the new system will handle all this stuff if you follow through with the configuration. But that's what you're supposed to tell me. Try it out and see what problems you encounter.

> Will those intermediate servers be "authorized" automatically?

You tell me.

> What happens when something happens and either postini or my work kicks over to the secondary servers

Ditto - this is why I need you to do some testing.

> Will I be able to report these or is this one of the instances where "Some "unique" users may not be able to report all the spam they have in the past."?

If you all share the same configuration, you should be able to continue reporting them. But again, try it and let me know. If they get their spam from a mailhost you don't have an account on, then you probably won't be able to report it, but spamcop should fail with a sensible error in that case.

-=Julian "not sure I like this new forum" Haight=-

[StevenUnderwood 0]

Thanks for the quick turn around. I will try and let you know how it works (actually, I started the work one this afternoon but had not received the test messages yet). I think I will hold off a couple days on my paid account.

I do like to know what to expect before trying something normally.

Edited March 18, 2004 by StevenUnderwood

[StevenUnderwood 0]

Julian:

I sent off the 2 confirmations. Are these handled automatically the way that the spam submittals are? The reason I ask is that in Lotus Notes, to get the full message with headers, I need to export the message to a text file. I then attached the text file to a reply of the confirmation message. The parser has never been able to get the headers from these attachments.

[StevenUnderwood 0]

Another update... My attaching the text message with heareds did not work. I have tried both messages simply pasting the extracted message into the email body with nothing else.

And that did not work because it looks like either Lotus Notes outbound MTA or your inbound MTA is wrapping the lines. I did not want to mess with them too much, but there were breaks in the 2 received lines with white space that I forced to the next line and tried again.

I will post any parts you want here, but it seems my special code is mentioned several places.

Edited March 18, 2004 by StevenUnderwood

[StevenUnderwood 0]

Last test for the night here...

That test failed. I looked and found that Lotus Notes v5 defaults to wrapping to 75 characters when converting Notes Rich Text to plain text for SMTP transactions.

What would be a reasonable amount to wrap at? What is the relevant RFC for this?

Also, keep in mind that while I admin this server and can change these defaults, that will not be the case for all Lotus Notes users. I notice your X_SpamCop_Test1: line is 140 characters.

[Wazoo 0]

What is the relevant RFC for this

admit that it's been a while since I've read up, but off the top of my head, I don't think wrapping was really a covered issue ..?? Line length in the header contents, use of special characters, lengths of certain fields, etc. .... and again, off the top of my head, I'd say turn it up to 255, which would allow anything "legal" to pass through unmodified.

[StevenUnderwood 0]

Julian:

My attempt to confirm also failed because of my MTA (I believe). Even setting the outgoing wrap to 255 characters, it came back failed.

Please post when you have a fix for this and I will try again. At least this was my first server I can not confirm so submitting via web page is still working.

[julian 0]

Julian:

I sent off the 2 confirmations.  Are these handled automatically the way that the spam submittals are?  The reason I ask is that in Lotus Notes, to get the full message with headers, I need to export the message to a text file.  I then attached the text file to a reply of the confirmation message.  The parser has never been able to get the headers from these attachments.

> Are these handled automatically the way that the spam submittals are?

Generally speaking, yes.

The trick with line wrapping is not that you have to do it at any specific column, but that when you wrap, you do it correctly - called "folding" in the RFC. The wrapped part should always be preceeded by some white space. For example:

Header:  This is a correctly
   "folded" line
Header2: This line is broken - wrapped
normally, without folding

Since headers never start with whitespace, this distinguishes the "continuation" part from the other headers.

Lotus notes is hard to deal with.

If you want to send samples or get into more detail on this just email me:

ngexpires040404[at]mail.julianhaight.com

-=Julian=-

[StevenUnderwood 0]

Lotus notes is hard to deal with.

If you want to send samples or get into more detail on this just email me:

I tried one more time manually wrapping one extremely long (287 character) line $MIMETrack: into 3 lines.

Otherwise, I will forward it to your attention.

This test failed as well with the message sent back showing all lines correctly wrapped with the exception on your X_SpamCop_Test1: header line. I have forwarded these error pages to you.

Edited March 18, 2004 by StevenUnderwood

[StevenUnderwood 0]

Julian:

One new question. In the past, specifically if there is a problem with spamcop or my ISP's email system, have changed the configuration to forward differently.

In the future if I do this, do I need to update my mailhosts configuration to reflect this?

For example, I used to have my poboxes.com addresses forward to spamcop and spamcop forward to my ISP where I pooped it from.

Now, I have poboxes.com forward directly to spamcop and my ISP forward directly to spamcop and POP it from there (while still popping my empty ISP account so I can send through their SMTP).

These obviously have two different routes, but the same hosts are involved.

[julian 0]

One new question.  In the past, specifically if there is a problem with spamcop or my ISP's email system, have changed the configuration to forward differently.

In the future if I do this, do I need to update my mailhosts configuration to reflect this?

I'll let you experiment and draw your own conclusions. If nothing else, you should be very careful whenever you change your mail routeing. Theoretically (and the way it's set up now), once all the mailhosts are configured, the order *should* be able to change and spamcop can keep up. But I can't garantee that. You just have to try it yourself and see if it works for whatever new configuration you are using. If in doubt, reconfigure the mailhosts involved. You should be able to always "add mailhost" on an existing address and it'll just get update the existing record while it tests to make sure it is parseable.

-=Julian=-

[StevenUnderwood 0]

Thanks

You should be able to always "add mailhost" on an existing address

If it were only that easy.

[StevenUnderwood 0]

Success!!!

Since others had apparently had success in getting this mailhost configuraton working, I tried it on my home setup and found things went very well.

Because I use Eudora at home to POP from spamcop, I simply stopped retrieving messages and forwarded the replies like I full report my spam from inside of the webmail application. All accounts currently forward into my SpamCop account so it was very easy.

And reporting seems to be working properly as well. I am back to full reporting until I get confidence that all is working.

I do note that I have yet to see the parser work back beyond the server that sent it into my configuration, which I think is the ultimate goal of this experiment. I don't know of anyway to test that legitimate outgoing relaying is being passed over.

Now if I can get my Lotus Notes configuration at work setup, I will be very happy.

One thing that the new configuration does is make it impossible to parse the headers posted in these forums to be able to help people solve their problems. Would it be possible to create a single parser page that anyone could use to determine the source of a message with no reports ever generated?

Edited March 19, 2004 by StevenUnderwood