SetVendorcastDeviceLabel - previous test-label not cleared

[Niels J]

I am having a little problem with SetVendorcastDeviceLabel.

When the RDM Testing tool tests 'vendorcast label', it transmits 'vendorcast label ascii\\xc0'.

Some of a previous label 'string with\x0d non ascii\xc0' is still in.

The tool is running on Raspberry Pi.

Here is the full report:

SetVendorcastDeviceLabel
SET the device label using the vendorcast address.

 SET: uid: 4d50:ffffffff, pid: DEVICE_LABEL (0x0082), sub device: 0, args: ['vendorcast label']
 Request status: Request was broadcast
 GET: uid: 4d50:010a2aca, pid: DEVICE_LABEL (0x0082), sub device: 0, args: []
 Response: RDMResponse: Get ACK, PID: 0x0082, PDL: 23, data: {'label': 'vendorcast label ascii\\xc0'}
 Failed: Labels didn't match, expected "vendorcast label", got "vendorcast label ascii\xc0"
 SET: uid: 4d50:010a2aca, pid: DEVICE_LABEL (0x0082), sub device: 0, args: ['vendorcast labe']
 Response: RDMResponse: Set ACK, PID: 0x0082, PDL: 0, data: {}

 

To me it looks like a bug in the test executer.

Will someone look into this?

[Simon Newton]

On Wed, Jan 15, 2014 at 5:13 AM, Niels J <nielskrus...@gmail.com> wrote:
> I am having a little problem with SetVendorcastDeviceLabel.
> When the RDM Testing tool tests 'vendorcast label', it transmits 'vendorcast
> label ascii\\xc0'.
> Some of a previous label 'string with\x0d non ascii\xc0' is still in.
> The tool is running on Raspberry Pi.

This is a bug in the responder, it's not handling non-NULL terminated
strings. See my comments inline below.

> Here is the full report:
> SetVendorcastDeviceLabel
> SET the device label using the vendorcast address.
> SET: uid: 4d50:ffffffff, pid: DEVICE_LABEL (0x0082), sub device: 0, args:
> ['vendorcast label']

This sends a SET DEVICE_LABEL with param data "'vendorcast label" to
4d50:ffffffff.

> Request status: Request was broadcast
> GET: uid: 4d50:010a2aca, pid: DEVICE_LABEL (0x0082), sub device: 0, args:
> []
> Response: RDMResponse: Get ACK, PID: 0x0082, PDL: 23, data: {'label':
> 'vendorcast label ascii\\xc0'}

This then sends a GET DEVICE_LABEL to "4d50:010a2aca". The device
returns 'vendorcast label ascii\\xc0'. Note that the
SetNonAsciiDeviceLabel test would have previously set the string
'string with\x0d non ascii\xc0' , so your responder isn't handling
non-NULL terminated strings, see section 10.1 of E1.20

> Failed: Labels didn't match, expected "vendorcast label", got "vendorcast
> label ascii\xc0"

The test then fails because the returned data doesn't match the value
it tried to set.

Simon

> SET: uid: 4d50:010a2aca, pid: DEVICE_LABEL (0x0082), sub device: 0, args:
> ['vendorcast labe']
> Response: RDMResponse: Set ACK, PID: 0x0082, PDL: 0, data: {}
>
> To me it looks like a bug in the test executer.
> Will someone look into this?
>

> --
>
> ---
> You received this message because you are subscribed to the Google Groups
> "RDM Testing" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to rdm-testing...@googlegroups.com.
> For more options, visit https://groups.google.com/groups/opt_out.

[Niels J]

>This sends a SET DEVICE_LABEL with param data "'vendorcast label" to
>4d50:ffffffff.
 

No, it actually transmits ''vendorcast  label ascii\\xc0'.

I have check the transmission on RDM.

 

[Simon Newton]

That's odd. Can you attach the capture?

[Niels J]

I will capture it Monday.

[Niels J]

I misinterpreted the data.
The error was in the responder.
Thanks for your support.