the Exchange 2007 fix

the Exchange 2007 fix

[Michael Harnois]

It doesn't seem to be the only fix needed, but I think it's the Big Kahuna.

mharnois:lisp mdharnois$ diff -urN nnimap.el nnimap.el.orig
--- nnimap.el    2008-03-24 14:16:40.000000000 -0500
+++ nnimap.el.orig    2008-02-07 11:36:36.000000000 -0600
@@ -555,7 +555,7 @@
           (imap-mailbox-select group examine))
       (let (minuid maxuid)
     (when (> (imap-mailbox-get 'exists) 0)
-      (imap-fetch "1:*" "UID" nil 'nouidfetch)
+      (imap-fetch "1,*" "UID" nil 'nouidfetch)
       (imap-message-map (lambda (uid Uid)
                   (setq minuid (if minuid (min minuid uid) uid)
                     maxuid (if maxuid (max maxuid uid) uid)))

Re: the Exchange 2007 fix

[Ted Zlatanov]

On Mon, 24 Mar 2008 14:31:16 -0500 "Michael Harnois" <mdharnois@gmail.com> wrote: 

MH> It doesn't seem to be the only fix needed, but I think it's the Big Kahuna.
MH> mharnois:lisp mdharnois$ diff -urN nnimap.el nnimap.el.orig
MH> --- nnimap.el    2008-03-24 14:16:40.000000000 -0500
MH> +++ nnimap.el.orig    2008-02-07 11:36:36.000000000 -0600
MH> @@ -555,7 +555,7 @@
MH>            (imap-mailbox-select group examine))
MH>        (let (minuid maxuid)
MH>      (when (> (imap-mailbox-get 'exists) 0)
MH> -      (imap-fetch "1:*" "UID" nil 'nouidfetch)
MH> +      (imap-fetch "1,*" "UID" nil 'nouidfetch)
MH>        (imap-message-map (lambda (uid Uid)
MH>                    (setq minuid (if minuid (min minuid uid) uid)
MH>                      maxuid (if maxuid (max maxuid uid) uid)))

Since it was a tiny fix I comitted it.  It seems to work fine with
Exchange 2007 and Courier IMAP.

If you find any other places to fix the IMAP handling with Exchange,
many uses will greatly appreciate it.

Ted

Re: the Exchange 2007 fix

[Nathan J. Williams]

Ted Zlatanov <tzz@lifelogs.com> writes:

> On Mon, 24 Mar 2008 14:31:16 -0500 "Michael Harnois" <mdharnois@gmail.com> wrote: 
>
> MH> It doesn't seem to be the only fix needed, but I think it's the Big Kahuna.
> MH> mharnois:lisp mdharnois$ diff -urN nnimap.el nnimap.el.orig
> MH> --- nnimap.el    2008-03-24 14:16:40.000000000 -0500
> MH> +++ nnimap.el.orig    2008-02-07 11:36:36.000000000 -0600
> MH> @@ -555,7 +555,7 @@
> MH>            (imap-mailbox-select group examine))
> MH>        (let (minuid maxuid)
> MH>      (when (> (imap-mailbox-get 'exists) 0)
> MH> -      (imap-fetch "1:*" "UID" nil 'nouidfetch)
> MH> +      (imap-fetch "1,*" "UID" nil 'nouidfetch)
> MH>        (imap-message-map (lambda (uid Uid)
> MH>                    (setq minuid (if minuid (min minuid uid) uid)
> MH>                      maxuid (if maxuid (max maxuid uid) uid)))
>
> Since it was a tiny fix I comitted it.  It seems to work fine with
> Exchange 2007 and Courier IMAP.

Please see: http://comments.gmane.org/gmane.emacs.gnus.general/65855

I think we can use "1,*:*" instead of "1:*" to get the min-max effect
without the effect of pulling down all the UIDs in the group; it's
what I've been using in my local copy of Gnus since December against
an Exchange 2007 server.

	- Nathan

Re: the Exchange 2007 fix

[Simon Josefsson]

nathanw@MIT.EDU (Nathan J. Williams) writes:

> Ted Zlatanov <tzz@lifelogs.com> writes:
>
>> On Mon, 24 Mar 2008 14:31:16 -0500 "Michael Harnois" <mdharnois@gmail.com> wrote: 
>>
>> MH> It doesn't seem to be the only fix needed, but I think it's the Big Kahuna.
>> MH> mharnois:lisp mdharnois$ diff -urN nnimap.el nnimap.el.orig
>> MH> --- nnimap.el    2008-03-24 14:16:40.000000000 -0500
>> MH> +++ nnimap.el.orig    2008-02-07 11:36:36.000000000 -0600
>> MH> @@ -555,7 +555,7 @@
>> MH>            (imap-mailbox-select group examine))
>> MH>        (let (minuid maxuid)
>> MH>      (when (> (imap-mailbox-get 'exists) 0)
>> MH> -      (imap-fetch "1:*" "UID" nil 'nouidfetch)
>> MH> +      (imap-fetch "1,*" "UID" nil 'nouidfetch)
>> MH>        (imap-message-map (lambda (uid Uid)
>> MH>                    (setq minuid (if minuid (min minuid uid) uid)
>> MH>                      maxuid (if maxuid (max maxuid uid) uid)))
>>
>> Since it was a tiny fix I comitted it.  It seems to work fine with
>> Exchange 2007 and Courier IMAP.
>
> Please see: http://comments.gmane.org/gmane.emacs.gnus.general/65855
>
> I think we can use "1,*:*" instead of "1:*" to get the min-max effect
> without the effect of pulling down all the UIDs in the group; it's
> what I've been using in my local copy of Gnus since December against
> an Exchange 2007 server.

So maybe this patch would work better?  However, it doesn't take care of
the imap.el problems you mentioned in that thread.  Perhaps the
bug-workaround variable should be a imap.el variable rather than a
nnimap.el variable?  Hm.

--- nnimap.el	01 Apr 2008 16:27:27 +0200	7.45
+++ nnimap.el	02 Apr 2008 00:38:11 +0200	
@@ -568,7 +568,7 @@
 	      (imap-mailbox-select group examine))
       (let (minuid maxuid)
 	(when (> (imap-mailbox-get 'exists) 0)
-	  (imap-fetch (if nnimap-enable-minmax-bug-workaround "1:*" "1,*")
+	  (imap-fetch (if nnimap-enable-minmax-bug-workaround "1,*:*" "1,*")
 		      "UID" nil 'nouidfetch)
 	  (imap-message-map (lambda (uid Uid)
 			      (setq minuid (if minuid (min minuid uid) uid)

/Simon

Re: the Exchange 2007 fix

[Michael D. Harnois]

nathanw@MIT.EDU (Nathan J. Williams) writes:

> Ted Zlatanov <tzz@lifelogs.com> writes:
>
>> On Mon, 24 Mar 2008 14:31:16 -0500 "Michael Harnois" <mdharnois@gmail.com> wrote: 
>>
>> MH> It doesn't seem to be the only fix needed, but I think it's the Big Kahuna.
>> MH> mharnois:lisp mdharnois$ diff -urN nnimap.el nnimap.el.orig
>> MH> --- nnimap.el    2008-03-24 14:16:40.000000000 -0500
>> MH> +++ nnimap.el.orig    2008-02-07 11:36:36.000000000 -0600
>> MH> @@ -555,7 +555,7 @@
>> MH>            (imap-mailbox-select group examine))
>> MH>        (let (minuid maxuid)
>> MH>      (when (> (imap-mailbox-get 'exists) 0)
>> MH> -      (imap-fetch "1:*" "UID" nil 'nouidfetch)
>> MH> +      (imap-fetch "1,*" "UID" nil 'nouidfetch)
>> MH>        (imap-message-map (lambda (uid Uid)
>> MH>                    (setq minuid (if minuid (min minuid uid) uid)
>> MH>                      maxuid (if maxuid (max maxuid uid) uid)))
>>
>> Since it was a tiny fix I comitted it.  It seems to work fine with
>> Exchange 2007 and Courier IMAP.
>
> Please see: http://comments.gmane.org/gmane.emacs.gnus.general/65855
>
> I think we can use "1,*:*" instead of "1:*" to get the min-max effect
> without the effect of pulling down all the UIDs in the group; it's
> what I've been using in my local copy of Gnus since December against
> an Exchange 2007 server.
>
> 	- Nathan

A thing of beauty. Actually, my great source of perplexment is that I
pored over RFC 3501 for hours and could not find an example of the "1,*"
syntax anywhere.

Re: the Exchange 2007 fix

[Jake Colman]

>>>>> "SJ" == Simon Josefsson <simon@josefsson.org> writes:

   SJ> nathanw@MIT.EDU (Nathan J. Williams) writes:
   >> Ted Zlatanov <tzz@lifelogs.com> writes:
   >> 
   >>> On Mon, 24 Mar 2008 14:31:16 -0500 "Michael Harnois"
   SJ> <mdharnois@gmail.com> wrote:
   >>> 
   MH> It doesn't seem to be the only fix needed, but I think it's the Big
   SJ> Kahuna.
   MH> mharnois:lisp mdharnois$ diff -urN nnimap.el nnimap.el.orig
   MH> --- nnimap.el    2008-03-24 14:16:40.000000000 -0500
   MH> +++ nnimap.el.orig    2008-02-07 11:36:36.000000000 -0600
   MH> @@ -555,7 +555,7 @@
   MH> (imap-mailbox-select group examine))
   MH> (let (minuid maxuid)
   MH> (when (> (imap-mailbox-get 'exists) 0)
   MH> -      (imap-fetch "1:*" "UID" nil 'nouidfetch)
   MH> +      (imap-fetch "1,*" "UID" nil 'nouidfetch)
   MH> (imap-message-map (lambda (uid Uid)
   MH> (setq minuid (if minuid (min minuid uid) uid)
   MH> maxuid (if maxuid (max maxuid uid) uid)))
   >>> 
   >>> Since it was a tiny fix I comitted it.  It seems to work fine with
   >>> Exchange 2007 and Courier IMAP.
   >> 
   >> Please see: http://comments.gmane.org/gmane.emacs.gnus.general/65855
   >> 
   >> I think we can use "1,*:*" instead of "1:*" to get the min-max effect
   >> without the effect of pulling down all the UIDs in the group; it's
   >> what I've been using in my local copy of Gnus since December against
   >> an Exchange 2007 server.

   SJ> So maybe this patch would work better?  However, it doesn't take care of
   SJ> the imap.el problems you mentioned in that thread.  Perhaps the
   SJ> bug-workaround variable should be a imap.el variable rather than a
   SJ> nnimap.el variable?  Hm.

   SJ> --- nnimap.el   01 Apr 2008 16:27:27 +0200      7.45
   SJ> +++ nnimap.el   02 Apr 2008 00:38:11 +0200
   SJ> @@ -568,7 +568,7 @@
   SJ>               (imap-mailbox-select group examine))
   SJ>        (let (minuid maxuid)
   SJ>         (when (> (imap-mailbox-get 'exists) 0)
   SJ> -         (imap-fetch (if nnimap-enable-minmax-bug-workaround "1:*" "1,*")
   SJ> +         (imap-fetch (if nnimap-enable-minmax-bug-workaround "1,*:*" "1,*")
   SJ>                       "UID" nil 'nouidfetch)
   SJ>           (imap-message-map (lambda (uid Uid)
   SJ>                               (setq minuid (if minuid (min minuid uid) uid)

I've applied this patch and it works - but I don't think I see any
performance improvement.

For example, when I enter a group with a very large number of read
messages and only 2 or 3 unread messages, I see the following message
for an extended period of time:

nnimap: Updating info for nnimap+exchange-d01:INBOX/Principia...

For the record, I am not using nnimap.el from cvs.  Instead I patched my
own copy of nnimap.el from two weeks ago to use "1,*:*".  I don't think
there is any other substantive change that I missed.

-- 
Jake Colman
Director of Software Development
Principia Partners LLC
101 West Elm Street
Suite 620
Conshohocken, PA  19428
+1 (610) 755-9786
www.principiapartners.com

Re: the Exchange 2007 fix

[Jake Colman]

>>>>> "m" == mdharnois  <mdharnois@gmail.com> writes:

   m> nathanw@MIT.EDU (Nathan J. Williams) writes:
   >> Ted Zlatanov <tzz@lifelogs.com> writes:
   >> 
   >>> On Mon, 24 Mar 2008 14:31:16 -0500 "Michael Harnois"
   m> <mdharnois@gmail.com> wrote:
   >>> 
   MH> It doesn't seem to be the only fix needed, but I think it's the Big
   m> Kahuna.
   MH> mharnois:lisp mdharnois$ diff -urN nnimap.el nnimap.el.orig
   MH> --- nnimap.el    2008-03-24 14:16:40.000000000 -0500
   MH> +++ nnimap.el.orig    2008-02-07 11:36:36.000000000 -0600
   MH> @@ -555,7 +555,7 @@
   MH> (imap-mailbox-select group examine))
   MH> (let (minuid maxuid)
   MH> (when (> (imap-mailbox-get 'exists) 0)
   MH> -      (imap-fetch "1:*" "UID" nil 'nouidfetch)
   MH> +      (imap-fetch "1,*" "UID" nil 'nouidfetch)
   MH> (imap-message-map (lambda (uid Uid)
   MH> (setq minuid (if minuid (min minuid uid) uid)
   MH> maxuid (if maxuid (max maxuid uid) uid)))
   >>> 
   >>> Since it was a tiny fix I comitted it.  It seems to work fine with
   >>> Exchange 2007 and Courier IMAP.
   >> 
   >> Please see: http://comments.gmane.org/gmane.emacs.gnus.general/65855
   >> 
   >> I think we can use "1,*:*" instead of "1:*" to get the min-max effect
   >> without the effect of pulling down all the UIDs in the group; it's
   >> what I've been using in my local copy of Gnus since December against
   >> an Exchange 2007 server.
   >> 
   >> - Nathan

   m> A thing of beauty. Actually, my great source of perplexment is that I
   m> pored over RFC 3501 for hours and could not find an example of the "1,*"
   m> syntax anywhere.

Nathan,

Does a 'B m' command work for you to move messages between folders?

...Jake

-- 
Jake Colman
Director of Software Development
Principia Partners LLC
101 West Elm Street
Suite 620
Conshohocken, PA  19428
+1 (610) 755-9786
www.principiapartners.com

Re: the Exchange 2007 fix

[Ted Zlatanov]

On Wed, 02 Apr 2008 00:41:17 +0200 Simon Josefsson <simon@josefsson.org> wrote: 

SJ> So maybe this patch would work better?  However, it doesn't take care of
SJ> the imap.el problems you mentioned in that thread.  Perhaps the
SJ> bug-workaround variable should be a imap.el variable rather than a
SJ> nnimap.el variable?  Hm.

FWIW, Exchange announces itself:

* OK The Microsoft Exchange IMAP4 service is ready."

so we could have specific fixes for it predicated on that, rather than
user-controlled variables (and the user could override the
auto-detection, naturally).  I don't relish the idea, but it's certainly
better than leaving a large percentage of Gnus users out in the cold.

Ted

Re: the Exchange 2007 fix

[Simon Josefsson]

Jake Colman <colman@ppllc.com> writes:

>    SJ> So maybe this patch would work better?  However, it doesn't take care of
>    SJ> the imap.el problems you mentioned in that thread.  Perhaps the
>    SJ> bug-workaround variable should be a imap.el variable rather than a
>    SJ> nnimap.el variable?  Hm.
>
>    SJ> --- nnimap.el   01 Apr 2008 16:27:27 +0200      7.45
>    SJ> +++ nnimap.el   02 Apr 2008 00:38:11 +0200
>    SJ> @@ -568,7 +568,7 @@
>    SJ>               (imap-mailbox-select group examine))
>    SJ>        (let (minuid maxuid)
>    SJ>         (when (> (imap-mailbox-get 'exists) 0)
>    SJ> -         (imap-fetch (if nnimap-enable-minmax-bug-workaround "1:*" "1,*")
>    SJ> +         (imap-fetch (if nnimap-enable-minmax-bug-workaround "1,*:*" "1,*")
>    SJ>                       "UID" nil 'nouidfetch)
>    SJ>           (imap-message-map (lambda (uid Uid)
>    SJ>                               (setq minuid (if minuid (min minuid uid) uid)
>
> I've applied this patch and it works - but I don't think I see any
> performance improvement.
>
> For example, when I enter a group with a very large number of read
> messages and only 2 or 3 unread messages, I see the following message
> for an extended period of time:
>
> nnimap: Updating info for nnimap+exchange-d01:INBOX/Principia...
>
> For the record, I am not using nnimap.el from cvs.  Instead I patched my
> own copy of nnimap.el from two weeks ago to use "1,*:*".  I don't think
> there is any other substantive change that I missed.

What's in *imap-log* if you enable imap-log after this has happened?

For the 1,*:* command do you get a two-line response or a many-line
response?

/Simon

Re: the Exchange 2007 fix

[Simon Josefsson]

Jake Colman <colman@ppllc.com> writes:

>    m> A thing of beauty. Actually, my great source of perplexment is that I
>    m> pored over RFC 3501 for hours and could not find an example of the "1,*"
>    m> syntax anywhere.
>
> Nathan,
>
> Does a 'B m' command work for you to move messages between folders?

Likely not -- one of Nathan's other patches was for imap.el related to
appending/copying messages between mailboxes.  So you'll need that patch
too in order to make moving work.

/Simon

Re: the Exchange 2007 fix

[Simon Josefsson]

Simon Josefsson <simon@josefsson.org> writes:

> Jake Colman <colman@ppllc.com> writes:
>
>>    m> A thing of beauty. Actually, my great source of perplexment is that I
>>    m> pored over RFC 3501 for hours and could not find an example of the "1,*"
>>    m> syntax anywhere.
>>
>> Nathan,
>>
>> Does a 'B m' command work for you to move messages between folders?
>
> Likely not -- one of Nathan's other patches was for imap.el related to
> appending/copying messages between mailboxes.  So you'll need that patch
> too in order to make moving work.

I have applied all patches to CVS, you'll need to update both imap.el
and nnimap.el, set the variable imap-enable-exchange-bug-workaround, and
recompile (or remove the stale *.elc files) to have things work.  This
should make B m work, does it?

/Simon

Re: the Exchange 2007 fix

[Simon Josefsson]

Ted Zlatanov <tzz@lifelogs.com> writes:

> On Wed, 02 Apr 2008 00:41:17 +0200 Simon Josefsson <simon@josefsson.org> wrote: 
>
> SJ> So maybe this patch would work better?  However, it doesn't take care of
> SJ> the imap.el problems you mentioned in that thread.  Perhaps the
> SJ> bug-workaround variable should be a imap.el variable rather than a
> SJ> nnimap.el variable?  Hm.
>
> FWIW, Exchange announces itself:
>
> * OK The Microsoft Exchange IMAP4 service is ready."
>
> so we could have specific fixes for it predicated on that, rather than
> user-controlled variables (and the user could override the
> auto-detection, naturally).  I don't relish the idea, but it's certainly
> better than leaving a large percentage of Gnus users out in the cold.

Maybe you could implement some logic that would set the newly introduced
imap-enable-exchange-bug-workaround based on the server greeting?

/Simon

Re: the Exchange 2007 fix

[Jake Colman]

>>>>> "SJ" == Simon Josefsson <simon@josefsson.org> writes:

   >> Likely not -- one of Nathan's other patches was for imap.el
   >> related to appending/copying messages between mailboxes.  So
   >> you'll need that patch too in order to make moving work.

   SJ> I have applied all patches to CVS, you'll need to update both
   SJ> imap.el and nnimap.el, set the variable
   SJ> imap-enable-exchange-bug-workaround, and recompile (or remove the
   SJ> stale *.elc files) to have things work.  This should make B m
   SJ> work, does it?

Any idea how I can get nightly tarball?  I tried to ftp to ftp.gnus.org
and to ftp.quimby.org but I time out.

Can someone send me the snapshot or point me to location from which I
can download it?

-- 
Jake Colman
Director of Software Development
Principia Partners LLC
101 West Elm Street
Suite 620
Conshohocken, PA  19428
+1 (610) 755-9786
www.principiapartners.com

Re: the Exchange 2007 fix

[Ted Zlatanov]

On Wed, 02 Apr 2008 18:13:56 -0400 Jake Colman <colman@ppllc.com> wrote: 

Jake> Any idea how I can get nightly tarball?  I tried to ftp to ftp.gnus.org
Jake> and to ftp.quimby.org but I time out.

Jake> Can someone send me the snapshot or point me to location from which I
Jake> can download it?

FWIW, I always check out/update from CVS.  The snapshots use up too much
bandwidth.

Ted

Re: the Exchange 2007 fix

[Ted Zlatanov]

On Wed, 02 Apr 2008 16:50:24 +0200 Simon Josefsson <simon@josefsson.org> wrote: 

SJ> Ted Zlatanov <tzz@lifelogs.com> writes:
>> On Wed, 02 Apr 2008 00:41:17 +0200 Simon Josefsson <simon@josefsson.org> wrote: 
>> 
SJ> So maybe this patch would work better?  However, it doesn't take care of
SJ> the imap.el problems you mentioned in that thread.  Perhaps the
SJ> bug-workaround variable should be a imap.el variable rather than a
SJ> nnimap.el variable?  Hm.
>> 
>> FWIW, Exchange announces itself:
>> 
>> * OK The Microsoft Exchange IMAP4 service is ready."
>> 
>> so we could have specific fixes for it predicated on that, rather than
>> user-controlled variables (and the user could override the
>> auto-detection, naturally).  I don't relish the idea, but it's certainly
>> better than leaving a large percentage of Gnus users out in the cold.

SJ> Maybe you could implement some logic that would set the newly introduced
SJ> imap-enable-exchange-bug-workaround based on the server greeting?

I am not sure if Exchange-specific handling should be on or off by
default.  The user will be able to override it the other way, but the
base default is a bit of a philosophical question and therefore must
take up more discussion time than a technical issue :)

Ted

Re: the Exchange 2007 fix

[Jake Colman]

>>>>> "SJ" == Simon Josefsson <simon@josefsson.org> writes:

   SJ> I have applied all patches to CVS, you'll need to update both
   SJ> imap.el and nnimap.el, set the variable
   SJ> imap-enable-exchange-bug-workaround, and recompile (or remove the
   SJ> stale *.elc files) to have things work.  This should make B m
   SJ> work, does it?

Yes it does.  I am using last night's snapshot and working with Exchange
is much, much better.  I still have an issue with 'updating info' taking
a very long time on a group with a huge number of read messages.  I will
enable imap-log and post later today or tomorrow.  What do you think:
add the message to this thread or create a new subject?

-- 
Jake Colman
Director of Software Development
Principia Partners LLC
101 West Elm Street
Suite 620
Conshohocken, PA  19428
+1 (610) 755-9786
www.principiapartners.com

Re: the Exchange 2007 fix

[Simon Josefsson]

Jake Colman <colman@ppllc.com> writes:

>>>>>> "SJ" == Simon Josefsson <simon@josefsson.org> writes:
>
>    SJ> I have applied all patches to CVS, you'll need to update both
>    SJ> imap.el and nnimap.el, set the variable
>    SJ> imap-enable-exchange-bug-workaround, and recompile (or remove the
>    SJ> stale *.elc files) to have things work.  This should make B m
>    SJ> work, does it?
>
> Yes it does.  I am using last night's snapshot and working with Exchange
> is much, much better.

Great.  Thanks for confirming.

> I still have an issue with 'updating info' taking a very long time on
> a group with a huge number of read messages.  I will enable imap-log
> and post later today or tomorrow.  What do you think: add the message
> to this thread or create a new subject?

Please add a new thread for it.  If the amount of data isn't excessive,
I'm not sure we'll be able to do much about it.  But seeing exactly what
is transferred would be useful (to make sure there isn't something wrong
with the recent patches).

/Simon

Re: the Exchange 2007 fix

[Reiner Steib]

On Thu, Apr 03 2008, Ted Zlatanov wrote:

>>> FWIW, Exchange announces itself:
>>> 
>>> * OK The Microsoft Exchange IMAP4 service is ready."
>>> 
>>> so we could have specific fixes for it predicated on that, rather
>>> than user-controlled variables (and the user could override the
>>> auto-detection, naturally).  I don't relish the idea, but it's
>>> certainly better than leaving a large percentage of Gnus users out
>>> in the cold.
>
> SJ> Maybe you could implement some logic that would set the newly
> SJ> introduced imap-enable-exchange-bug-workaround based on the
> SJ> server greeting?
>
> I am not sure if Exchange-specific handling should be on or off by
> default.  The user will be able to override it the other way, but
> the base default is a bit of a philosophical question and therefore
> must take up more discussion time than a technical issue :)

The subject say that the problem is only with Exchange 2007.  If older
versions don't require it, we should probably not enable it just
because of the Exchange server greeting (which doesn't tell the
version).

Bye, Reiner.
-- 
       ,,,
      (o o)
---ooO-(_)-Ooo---  |  PGP key available  |  http://rsteib.home.pages.de/

Re: the Exchange 2007 fix

[Jake Colman]

>>>>> "RS" == Reiner Steib <reinersteib+gmane@imap.cc> writes:

   RS> On Thu, Apr 03 2008, Ted Zlatanov wrote:
   >>>> FWIW, Exchange announces itself:
   >>>> 
   >>>> * OK The Microsoft Exchange IMAP4 service is ready."
   >>>> 
   >>>> so we could have specific fixes for it predicated on that,
   >>>> rather than user-controlled variables (and the user could
   >>>> override the auto-detection, naturally).  I don't relish the
   >>>> idea, but it's certainly better than leaving a large percentage
   >>>> of Gnus users out in the cold.
   >> 
   SJ> Maybe you could implement some logic that would set the newly
   SJ> introduced imap-enable-exchange-bug-workaround based on the
   SJ> server greeting?
   >> 
   >> I am not sure if Exchange-specific handling should be on or off by
   >> default.  The user will be able to override it the other way, but
   >> the base default is a bit of a philosophical question and
   >> therefore must take up more discussion time than a technical issue
   >> :)

   RS> The subject say that the problem is only with Exchange 2007.  If
   RS> older versions don't require it, we should probably not enable it
   RS> just because of the Exchange server greeting (which doesn't tell
   RS> the version).

I can vouch for the fact that Exchange 5.5 worked like a charm.  It's
amazing how badly MickySoft broke things when it released Exchange 2007.

I can also vouch for the fact that the patches contained in the latest
release of imap.el and nnimap.el go a very long way towards addressing
the issues.  We're not quite there but we're edging ever close.

-- 
Jake Colman
Director of Software Development
Principia Partners LLC
101 West Elm Street
Suite 620
Conshohocken, PA  19428
+1 (610) 755-9786
www.principiapartners.com

Re: the Exchange 2007 fix

[Ted Zlatanov]

On Thu, 03 Apr 2008 21:37:07 +0200 Reiner Steib <reinersteib+gmane@imap.cc> wrote: 

RS> On Thu, Apr 03 2008, Ted Zlatanov wrote:
>>>> FWIW, Exchange announces itself:
>>>> 
>>>> * OK The Microsoft Exchange IMAP4 service is ready."
>>>> 
>>>> so we could have specific fixes for it predicated on that, rather
>>>> than user-controlled variables (and the user could override the
>>>> auto-detection, naturally).  I don't relish the idea, but it's
>>>> certainly better than leaving a large percentage of Gnus users out
>>>> in the cold.
>> 
SJ> Maybe you could implement some logic that would set the newly
SJ> introduced imap-enable-exchange-bug-workaround based on the
SJ> server greeting?
>> 
>> I am not sure if Exchange-specific handling should be on or off by
>> default.  The user will be able to override it the other way, but
>> the base default is a bit of a philosophical question and therefore
>> must take up more discussion time than a technical issue :)

RS> The subject say that the problem is only with Exchange 2007.  If older
RS> versions don't require it, we should probably not enable it just
RS> because of the Exchange server greeting (which doesn't tell the
RS> version).

OK.  In that case I'll leave things as they are right now.

Thanks
Ted

Re: the Exchange 2007 fix

[Steinar Bang]

>>>>> Simon Josefsson <simon@josefsson.org>:

> I have applied all patches to CVS, you'll need to update both imap.el
> and nnimap.el, set the variable imap-enable-exchange-bug-workaround,
> and recompile (or remove the stale *.elc files) to have things work.
> This should make B m work, does it?

Is that variable setting server local?

I have folders on an Exchange server at work, and folders on a cyrus
server at home, side by side in Gnus.