[Alpine-info] alpine 2.00 and Exchange 2007

[Mark Crispin]

On Thu, 4 Jun 2009, John Hodrien wrote:

>> This was supposed to have been made the default (and only) setting years

>> ago.

> In the sense that this was something Microsoft had agreed to do?


In the sense that Microsoft said it was done a decade ago!


> Heh, all I meant was, against Exchange, I have minimal confidence that it

> doesn't just return me gibberish to convince me to use MAPI instead ;)


I wish that that was just a joke.


> <snip for tidiness>

> Sure.  Exchange 2007 is hideously bad (for whatever reasons) as an IMAP

> server, and these are kludges to make alpine usable against it.

> Basically I'm making patches out of a pragmatic need to interface with

> Exchange, while getting all of these issues raised as faults with Exchange.


Yes, and I sympathize.

Just understand that I can't say "it is alright to do this kludge".  If 
someone gets bitten in the ass (arse for you UK folks) by the kludge in 
the future, I don't want anyone to say "but Mark Crispin said that it was 
alright!"

Hence what I can say is "at this point in time, this kludge is probably 
relatively harmless."  Weasel words galore, but the current climate favors
the mustelidae and they are in abundance...


>> Exchange has a more serious bug in that some versions present EXISTS and

>> EXPUNGE values out of order.  That actually causes real harm because the

>> client will interpret it in the wrong way.  There's no way to patch around

>> that problem; if you can't get a patch from MS to fix the problem in

>> Exchange you are screwed.

> Eek, we've not obviously seen this bug, although I'd read about the

> possibility.


I don't know the status of MS' patch for this bug.  But the fact that this 
bug got into distribution indicates that MS did zero regression testing 
(much less any quality control) on Exchange 2007.


> Is there a good list of Exchange 2007 bugs?


I came up with a partial list while I was still at UW, but I no longer 
have it.

As I recall, I came up with well over a dozen issues before the futility 
of further testing in that version stopped me.  That is, some of the 
problems were serious enough to block further testing in that area.  This 
was a "fix these bugs first, then we redo the test to find more bugs" 
situation.

For my efforts, I came under considerable pressure to shut up.  Some 
months later, I was directly told by Microsoft that they had no intention 
of fixing any IMAP bugs in Exchange.


> I've found microsoft documents

> that explain how they fail to adhere to the RFCs but it seems woefully

> incomplete.


Incomplete and self-contradictory.  I am one of the (probably very few) 
individuals outside of MS who has ever figured out how to use the SSPI API 
for TLS/SSL and Kerberos.  I quickly discovered that there are lies, 
damned lies, and Microsoft documentation.

To make things worse, many of the example programs don't work!  Often the 
clue is to found in an example program for some completely different task; 
you see them do something weird (with no explanation!), and puzzle over 
"why did they do that?"  This is how you learn how SSPI buffers work.

I guess that it's my personal pergatory for complaining about the quality 
of Unix man pages.

-- Mark --

http://panda.com/mrc
Democracy is two wolves and a sheep deciding what to eat for lunch.
Liberty is a well-armed sheep contesting the vote.