Thread: Re: [Logicmail-users] LogicMail 1.9.1.213

SourceForge Headquarters 225 Broadway Suite 1600 San Diego, CA 92101 +1 (858) 422-6466

Derek, did you get my PM? Reply here please yes/no there may be delivery problems in both directions.

-- Peter

  Yes, and testing now.  It looks like you have a very weird POP 
server.  I'll mention details here, in case anyone else is curious.

With a normal POP server, the login sequence is like such:
----
+OK <12...@fo...> mail.foo.com [CR][LF]
USER jdoe[CR][LF]
+OK[CR][LF]
PASS password[CR][LF]
+OK Mailbox locked and ready[CR][LF]
----

But with your POP server, it looks like this:
----
+OK <12...@ba...>[CR][LF]
USER jdoe[CR][LF]
+OK[CR][LF]
PASS password[CR][LF]
+OK qmail-pop3d-gemini 1766: auth_init: login=jdoe, 
authdata=password[LF][LF]
+OK qmail-pop3d-gemini 1766: Pod 7 Gemini_POP, country: '' found in the 
auth database[LF][LF]
+OK[CR][LF]
----

That awkwardly terminated/separated multi-line response to the PASS 
command is what's screwing up LogicMail.  I now need to fix my 
line-termination-handler code to detect and work around this situation.

-Derek

On 10/15/2010 1:14 PM, Peter P. wrote:
> Derek, did you get my PM? Reply here please yes/no there may be 
> delivery problems in both directions.
>
> -- Peter
>
>
>
> ------------------------------------------------------------------------------
> Download new Adobe(R) Flash(R) Builder(TM) 4
> The new Adobe(R) Flex(R) 4 and Flash(R) Builder(TM) 4 (formerly
> Flex(R) Builder(TM)) enable the development of rich applications that run
> across multiple browsers and platforms. Download your free trials today!
> http://p.sf.net/sfu/adobe-dev2dev
>
>
> _______________________________________________
> Logicmail-users mailing list
> Log...@li...
> https://lists.sourceforge.net/lists/listinfo/logicmail-users

-- 
----------------------------
  Derek Konigsberg
  dko...@lo...
----------------------------

Ok, I got the problem. Weird, but it screwed up other clients too. Alpine is one of the few which handle it right. I think that one has to implement the \r\n line termination detection required by RFCs, and not rely on \n alone in headers. Although that IS weird nowadays. Strange godaddy are not exactly script kiddies. Maybe it's a temporary bug introduced by an update, but I do not believe to be the only one to have noticed it.

Thanks for answering, looking forward to a solution.

-- Peter

--- On Fri, 10/15/10, Derek Konigsberg <dko...@lo...> wrote:
Yes, and testing now.  It looks like you have a very weird POP
    server.  I'll mention details here, in case anyone else is curious.

    With a normal POP server, the login sequence is like such:

    ----

    +OK <12...@fo...> mail.foo.com [CR][LF]

    USER jdoe[CR][LF]

    +OK[CR][LF]

    PASS password[CR][LF]

    +OK Mailbox locked and ready[CR][LF]

    ----

    But with your POP server, it looks like this:

    ----

    +OK <12...@ba...>[CR][LF]

    USER jdoe[CR][LF]

    +OK[CR][LF]

    PASS password[CR][LF]

    +OK qmail-pop3d-gemini 1766: auth_init: login=jdoe,
    authdata=password[LF][LF]

    +OK qmail-pop3d-gemini 1766: Pod 7 Gemini_POP, country: '' found in
    the auth database[LF][LF]

    +OK[CR][LF]

    ----

    That awkwardly terminated/separated multi-line response to the PASS
    command is what's screwing up LogicMail.  I now need to fix my
    line-termination-handler code to detect and work around this
    situation.

    -Derek

  I was actually originally keying on LF or CRLF.  Obviously it started 
by finding the first instance of an LF in the response, hence the 
problem.  In any case, I just checked in a fix.  (I also checked in a 
fix to the config screen layout issue an hour or two ago.)

However, I've been neck-deep in some far more involved coding on 
LogicMail this week (check the site's changelog/timeline if curious), 
and am thus not quite ready to turn the automated builds back on.  I'll 
probably post to the list sometime in the next few days with more 
details on what I've been working on.

-Derek

On 10/15/2010 2:16 PM, Peter P. wrote:
> Ok, I got the problem. Weird, but it screwed up other clients too. 
> Alpine is one of the few which handle it right. I think that one has 
> to implement the \r\n line termination detection required by RFCs, and 
> not rely on \n alone in headers. Although that IS weird nowadays. 
> Strange godaddy are not exactly script kiddies. Maybe it's a temporary 
> bug introduced by an update, but I do not believe to be the only one 
> to have noticed it.
>
> Thanks for answering, looking forward to a solution.
>
> -- Peter
>
> --- On *Fri, 10/15/10, Derek Konigsberg 
> /<dko...@lo...>/* wrote:
>
>     Yes, and testing now.  It looks like you have a very weird POP
>     server.  I'll mention details here, in case anyone else is curious.
>
>     With a normal POP server, the login sequence is like such:
>     ----
>     +OK <12...@fo...> </mc/compose?to=12...@fo...> mail.foo.com
>     [CR][LF]
>     USER jdoe[CR][LF]
>     +OK[CR][LF]
>     PASS password[CR][LF]
>     +OK Mailbox locked and ready[CR][LF]
>     ----
>
>     But with your POP server, it looks like this:
>     ----
>     +OK <12...@ba...> </mc/compose?to=12...@ba...>[CR][LF]
>     USER jdoe[CR][LF]
>     +OK[CR][LF]
>     PASS password[CR][LF]
>     +OK qmail-pop3d-gemini 1766: auth_init: login=jdoe,
>     authdata=password[LF][LF]
>     +OK qmail-pop3d-gemini 1766: Pod 7 Gemini_POP, country: '' found
>     in the auth database[LF][LF]
>     +OK[CR][LF]
>     ----
>
>     That awkwardly terminated/separated multi-line response to the
>     PASS command is what's screwing up LogicMail.  I now need to fix
>     my line-termination-handler code to detect and work around this
>     situation.
>
>     -Derek
>
>
>
> ------------------------------------------------------------------------------
> Download new Adobe(R) Flash(R) Builder(TM) 4
> The new Adobe(R) Flex(R) 4 and Flash(R) Builder(TM) 4 (formerly
> Flex(R) Builder(TM)) enable the development of rich applications that run
> across multiple browsers and platforms. Download your free trials today!
> http://p.sf.net/sfu/adobe-dev2dev
>
>
> _______________________________________________
> Logicmail-users mailing list
> Log...@li...
> https://lists.sourceforge.net/lists/listinfo/logicmail-users

-- 
----------------------------
  Derek Konigsberg
  dko...@lo...
----------------------------

Okay, thanks for the quick fix. I'll check it again next week, I hope it will work by then.

So automated builds are off now? I see. I understand that 1.9.1 is a maintenance release anyway. True?

I do not currently have access to the build system (need to boot Windows) so rebuilding it from scratch is not an available option for me so far. We will see how things will work out later, next week. If you find the time to rebuild 1.9.1. please notify, I'll be watching the mailing list.

Thanks for the great effort, the client works great apart from the strange bug I found.

-- Peter

--- On Fri, 10/15/10, Derek Konigsberg <dko...@lo...> wrote:

  I was actually originally keying on LF or CRLF.  Obviously it
    started by finding the first instance of an LF in the response,
    hence the problem.  In any case, I just checked in a fix.  (I also
    checked in a fix to the config screen layout issue an hour or two
    ago.)

    However, I've been neck-deep in some far more involved coding on
    LogicMail this week (check the site's changelog/timeline if
    curious), and am thus not quite ready to turn the automated builds
    back on.  I'll probably post to the list sometime in the next few
    days with more details on what I've been working on.

    -Derek

    On 10/15/2010 2:16 PM, Peter P. wrote:

            Ok, I got the
              problem. Weird, but it screwed up other clients too.
              Alpine is one of the few which handle it right. I think
              that one has to implement the \r\n line termination
              detection required by RFCs, and not rely on \n alone in
              headers. Although that IS weird nowadays. Strange godaddy
              are not exactly script kiddies. Maybe it's a temporary bug
              introduced by an update, but I do not believe to be the
              only one to have noticed it.

              Thanks for answering, looking forward to a solution.

              -- Peter

              --- On Fri, 10/15/10, Derek Konigsberg <dko...@lo...>
              wrote:

                Yes, and testing now.  It looks
                  like you have a very weird POP server.  I'll mention
                  details here, in case anyone else is curious.

                  With a normal POP server, the login sequence is like
                  such:

                  ----

                  +OK <12...@fo...>
                  mail.foo.com [CR][LF]

                  USER jdoe[CR][LF]

                  +OK[CR][LF]

                  PASS password[CR][LF]

                  +OK Mailbox locked and ready[CR][LF]

                  ----

                  But with your POP server, it looks like this:

                  ----

                  +OK <12...@ba...>[CR][LF]

                  USER jdoe[CR][LF]

                  +OK[CR][LF]

                  PASS password[CR][LF]

                  +OK qmail-pop3d-gemini 1766: auth_init: login=jdoe,
                  authdata=password[LF][LF]

                  +OK qmail-pop3d-gemini 1766: Pod 7 Gemini_POP,
                  country: '' found in the auth database[LF][LF]

                  +OK[CR][LF]

                  ----

                  That awkwardly terminated/separated multi-line
                  response to the PASS command is what's screwing up
                  LogicMail.  I now need to fix my
                  line-termination-handler code to detect and work
                  around this situation.

                  -Derek

  1.9.1.x is not a maintenance release (that was 1.1.x).  Its the 
development build.  Or in other words, what will eventually become 2.0.  
Normally its built automatically whenever I check in new code.

However, since so many people (besides me) are actually running 1.9.1.x 
now (since 1.1 has been stale forever), I occasionally turn off 
automatic builds during times of major code slush.

-Derek

On 10/15/2010 2:40 PM, Peter P. wrote:
> Okay, thanks for the quick fix. I'll check it again next week, I hope 
> it will work by then.
>
> So automated builds are off now? I see. I understand that 1.9.1 is a 
> maintenance release anyway. True?
>
> I do not currently have access to the build system (need to boot 
> Windows) so rebuilding it from scratch is not an available option for 
> me so far. We will see how things will work out later, next week. If 
> you find the time to rebuild 1.9.1. please notify, I'll be watching 
> the mailing list.
>
> Thanks for the great effort, the client works great apart from the 
> strange bug I found.
>
> -- Peter
>
> --- On *Fri, 10/15/10, Derek Konigsberg 
> /<dko...@lo...>/* wrote:
>
>     I was actually originally keying on LF or CRLF.  Obviously it
>     started by finding the first instance of an LF in the response,
>     hence the problem.  In any case, I just checked in a fix.  (I also
>     checked in a fix to the config screen layout issue an hour or two
>     ago.)
>
>     However, I've been neck-deep in some far more involved coding on
>     LogicMail this week (check the site's changelog/timeline if
>     curious), and am thus not quite ready to turn the automated builds
>     back on.  I'll probably post to the list sometime in the next few
>     days with more details on what I've been working on.
>
>     -Derek
>
>     On 10/15/2010 2:16 PM, Peter P. wrote:
>>     Ok, I got the problem. Weird, but it screwed up other clients
>>     too. Alpine is one of the few which handle it right. I think that
>>     one has to implement the \r\n line termination detection required
>>     by RFCs, and not rely on \n alone in headers. Although that IS
>>     weird nowadays. Strange godaddy are not exactly script kiddies.
>>     Maybe it's a temporary bug introduced by an update, but I do not
>>     believe to be the only one to have noticed it.
>>
>>     Thanks for answering, looking forward to a solution.
>>
>>     -- Peter
>>
>>     --- On *Fri, 10/15/10, Derek Konigsberg
>>     /<dko...@lo...>
>>     </mc/compose?to=dko...@lo...>/* wrote:
>>
>>         Yes, and testing now.  It looks like you have a very weird
>>         POP server.  I'll mention details here, in case anyone else
>>         is curious.
>>
>>         With a normal POP server, the login sequence is like such:
>>         ----
>>         +OK <12...@fo...> mail.foo.com [CR][LF]
>>         USER jdoe[CR][LF]
>>         +OK[CR][LF]
>>         PASS password[CR][LF]
>>         +OK Mailbox locked and ready[CR][LF]
>>         ----
>>
>>         But with your POP server, it looks like this:
>>         ----
>>         +OK <12...@ba...>[CR][LF]
>>         USER jdoe[CR][LF]
>>         +OK[CR][LF]
>>         PASS password[CR][LF]
>>         +OK qmail-pop3d-gemini 1766: auth_init: login=jdoe,
>>         authdata=password[LF][LF]
>>         +OK qmail-pop3d-gemini 1766: Pod 7 Gemini_POP, country: ''
>>         found in the auth database[LF][LF]
>>         +OK[CR][LF]
>>         ----
>>
>>         That awkwardly terminated/separated multi-line response to
>>         the PASS command is what's screwing up LogicMail.  I now need
>>         to fix my line-termination-handler code to detect and work
>>         around this situation.
>>
>>         -Derek
>>
>
>
> ------------------------------------------------------------------------------
> Download new Adobe(R) Flash(R) Builder(TM) 4
> The new Adobe(R) Flex(R) 4 and Flash(R) Builder(TM) 4 (formerly
> Flex(R) Builder(TM)) enable the development of rich applications that run
> across multiple browsers and platforms. Download your free trials today!
> http://p.sf.net/sfu/adobe-dev2dev
>
>
> _______________________________________________
> Logicmail-users mailing list
> Log...@li...
> https://lists.sourceforge.net/lists/listinfo/logicmail-users

-- 
----------------------------
  Derek Konigsberg
  dko...@lo...
----------------------------

Maybe relevant (I don't know if exim handles such horrors right - why would it), for the argument's sake:

http://www.exim.org/exim-html-current/doc/html/spec_html/ch44.html

# 2. Line Endings

-- Peter