[Tickets #11639] Re: ActiveSync :: Extend the Autodiscover feature to return also valid answers to MS Outlook queries
bugs at horde.org
bugs at horde.org
Thu Nov 15 15:54:34 UTC 2012
DO NOT REPLY TO THIS MESSAGE. THIS EMAIL ADDRESS IS NOT MONITORED.
Ticket URL: http://bugs.horde.org/ticket/11639
------------------------------------------------------------------------------
Ticket | 11639
Updated By | Michael Rubinsky <mrubinsk at horde.org>
Summary | ActiveSync :: Extend the Autodiscover feature to return
| also valid answers to MS Outlook queries
Queue | Horde Framework Packages
Version | Git develop
Type | Enhancement
State | Feedback
Priority | 1. Low
Milestone |
Patch | 1
Owners | Michael Rubinsky
------------------------------------------------------------------------------
Michael Rubinsky <mrubinsk at horde.org> (2012-11-15 15:54) wrote:
>> How about now?
>>
>>
>> http://lists.horde.org/archives/commits/2012-November/017210.html
>>
>> http://lists.horde.org/archives/commits/2012-November/017212.html
>
> Hi,
> The respond schema is now detected correctly for both, Outlook and RCA.
>
>
> Regarding teh logic fix: Maybe I'm missing the point, but for me it
> does make still no sense.
>
> My observation was, that outlook 2013 performs two independet
> autodiscover steps: It tries to discover a mobilesync setting, and
> it tries to discover a conventional Outlook setting as Outlook 2010
> and Outlook 2007 do.
>
> Some suggestions for Autodiscover.php:
> - Line 113 and 136 check both for empty repsonse schema. this could
> be done directly after entering the _buildResponseString(..)
> function to make the rest of the function easier to read.
Line 113 checks for an empty *request* schema. It's needed because
some broken android clients don't send the XML payload at all. So, we
default to the assuming a mobilesync *request* schema.
> - Line 135 opens a conditionalblock for the "autodiscover/outlook"
> schema, so within this block, it does not make sense to check for a
> mobilesync schema as done in lines 140:144. -> what was the
> intention of the line? In my opinion we must drop that lines,
> because the comparison in line 140 is always true and thus the
> function returns a failure response.
I misunderstood what you were telling me about the multiple requests
that outlook makes. The idea was to catch the mobilesync request made
by Outlook and reject it, but I now see that the request from Outlook
for mobilesync looks like any other mobilesync request. I'll have to
check the useragent, which I was hoping to avoid since future versions
of Outlook can have different values for this.
> Bye the way:
> the failure response has always the response schema for a
> mobilesync, so RCA seems not to unterstand that an error occured.
> the _buildFailureResponse(..) function should use the correct
> response schema in line 216 to allow Outlook & RCA to unterstand
> that the error code is the response to teir query.
Ok. Thanks for the info. Will fix.
> Torben
More information about the bugs
mailing list