Nokia WSI MM7 support

Nokia WSI MM7 support SearchSearch
Author Message
avism
Unregistered guest
Posted on Sunday, November 30, 2003 - 03:54 pm:   

Hi,

Does the Now.SMS support the Nokia WSI MM7 interface.
I could not succeed to send MMS messages to now.sms as a vasp application.

Thanks Avi
avism
Unregistered guest
Posted on Monday, December 01, 2003 - 01:28 pm:   

Hi,

More info on the above problem.
I get the following error from the Nowsms upon submiting the MMS over the MM7:

HTTP/1.0 200 OK
Content-length: 731
Content-type: text/xml
Connection: close

<?xml version='1.0' ?>
<env:Envelope xmlns:env="http://schemas.xmlsoap.org/soap/envelope/">
<env:Header>
<mm7:TransactionID xmlns:mm7="http://www.3gpp.org/ftp/Specs/archive/23_series/23.140/schema/REL-5-MM7-1-0" env:mustUnderstand="1">
unknown
</mm7:TransactionID>
</env:Header>
<env:Body>
<mm7:RSErrorRsp xmlns:mm7="http://www.3gpp.org/ftp/Specs/archive/23_series/23.140/schema/REL-5-MM7-1-0">
<MM7Version>5.3.0</MM7Version>
<Status>
<StatusCode>4003</StatusCode>
<StatusText>Unsupported Operation</StatusText>
<Details>Closing MIME boundary for second content part not found</Details>
</Status>
</mm7:RSErrorRsp>
</env:Body>
</env:Envelope>

What could be the problem?
Thanks Avi.
Bryce Norwood - NowSMS Support
Board Administrator
Username: Bryce

Post Number: 1285
Registered: 10-2002
Posted on Monday, December 01, 2003 - 06:10 pm:   

Avi,

I think I'd want to see the MM7 data that is being posted to us.

But before we troubleshoot this further, I can think of a bug that we've fixed which could be causing this problem.

Try downloading http://www.nowsms.com/download/nowsms5-patch3.zip. It includes an updated MMSC.EXE which fixes a problem in the MM7 receiver where the MM7 receiver did not like MM7 messages which included binary content and had embedded NULLs (content with embedded NULLs could only be submitted using base64 encoding).

An embedded NULL is the only logical reason that I could see this particular error being triggered (unless there is something very odd or corrupt about the message formatting). So try this patch.

If that doesn't fix the problem, then enable the debug mode in the MMSC so that we can see the format of the MM7 post. To enable debug mode, manually edit MMSC.INI and under the [MMSC] section header, add Debug=Yes. Then repeat your MM7 attempt, and e-mail the mmscdebug.log to nowsms@now.co.uk (or post in reply to this message).

-bn
avism
Unregistered guest
Posted on Tuesday, December 02, 2003 - 01:36 pm:   

Thanks Bryce,

The new version worked fine, messages are being successfully received (as MM7 VASP) and processed by the now.sms.

Although the Nokia WSI MM7 API throws an exception of "com.nokia.mobile.services.api.DriverException: remote call failed".
Did you tested the now.sms with the new Nokia WSI MM7 API?

Thanks Avi.
Bryce Norwood - NowSMS Support
Board Administrator
Username: Bryce

Post Number: 1299
Registered: 10-2002
Posted on Wednesday, December 03, 2003 - 02:54 pm:   

Avi,

No, we have not yet tested with the Nokia MM7 API.

Of course, the type of problem that you describe is a difficult one to troubleshoot. It is easy to troubleshoot if we are unable to understand a request that the other software generates ... but it is often difficult to guess why another application does not like a response that we generate.

We'll have to schedule some time to test that library.
avism
Unregistered guest
Posted on Tuesday, December 09, 2003 - 12:42 pm:   

Hi Bryce,

Could it be that the now.sms response is not a valid MM7 SOAP response?

Avi.
avism
Unregistered guest
Posted on Tuesday, December 09, 2003 - 01:21 pm:   

Hi Bryce,

I thinks the proble is with wrong schema URI.
Now.sms response with URI to: http://www.3gpp.org/ftp/Specs/archive/23_series/23.140/schema/REL-5-MM7-1-0

but it should point to http://www.3gpp.org/ftp/Specs/archive/23_series/23.140/schema/REL-5-MM7-1-2
because the http://www.3gpp.org/ftp/Specs/archive/23_series/23.140/schema/REL-5-MM7-1-0 does not include the <MessageID> element in the <mm7:SubmitRsp>

Could you please relaese a new patch that sends the correct schema URI.

Thanks Avi.

Bryce Norwood - NowSMS Support
Board Administrator
Username: Bryce

Post Number: 1337
Registered: 10-2002
Posted on Tuesday, December 09, 2003 - 08:55 pm:   

Avi,

You're a good detective.

Send an e-mail to nowsms@now.co.uk referencing this thread, and I'll e-mail you an update.

The update will also be included in the next patch release, but I'd be interested to find out if this resolves your problem prior to that patch release.

-bn
avism
Unregistered guest
Posted on Monday, December 15, 2003 - 08:59 am:   

Hi Bryce,

I have continued in the investigation of the problem.

The response of NowSMS is incorrect because it does not exactly follow
the MM7 specification:

<mm7:SubmitRsp xmlns:mm7="http://www.3gpp.org/.../REL-5-MM7-1-2">
<MM7Version>5.3.0</MM7Version>
<Status>
...

The XML namespace is missing from <MM7Version>, <Status> and all the
other child elements of SubmitRsp (They should have been written
<mm7:MM7Version>, <mm7:Status>, etc)


Could you please confirm and if possible send me a patch version of Now.sms for further check.

Thanks Avi.
Bryce Norwood - NowSMS Support
Board Administrator
Username: Bryce

Post Number: 1385
Registered: 10-2002
Posted on Monday, December 15, 2003 - 09:11 pm:   

Hi Avi,

I replied to you via e-mail.

We based our MM7 implementation on the definition in 3GPP TS 23.140 version 5.4.0. It looks like there was the same bug in the examples cited in that version, and earlier versions of the 3GPP MMS specification. The MM7 examples were corrected in 3GPP TS 23.140 version 5.5.0 and later releases.

From an XML perspective, it's definitely not correct. We followed the examples in the earlier version of the 3GPP spec too closely.

Instead of:

<mm7:SubmitRsp xmlns:mm7="http://www.3gpp.org/.../REL-5-MM7-1-2">
<MM7Version>5.3.0</MM7Version>
<Status>

We're going to encode this as:

<SubmitRsp xmlns="http://www.3gpp.org/.../REL-5-MM7-1-2">
<MM7Version>5.3.0</MM7Version>
<Status>

I've already sent you an update, which seems to be working now.

-bn