No problem. Please let us know of any other MIDI issues you discover.
Frank Yerrace
From: wdmaudiodev-bounce@xxxxxxxxxxxxx
[mailto:wdmaudiodev-bounce@xxxxxxxxxxxxx] On Behalf Of Basileios Bu
Sent: Friday, May 1, 2015 11:26 AM
To: wdmaudiodev@xxxxxxxxxxxxx
Subject: [wdmaudiodev] Re: Question for MIDI on older W10 Technical Preview
(10061)
Thanks for taking the time to answer!
On Fri, May 1, 2015 at 1:53 PM, Frank Yerrace
<Frank.Yerrace@xxxxxxxxxxxxx<mailto:Frank.Yerrace@xxxxxxxxxxxxx>> wrote:
I believe the OS was rejecting devices that did not support
KSPROPERTY_GENERAL_COMPONENTID.
Frank Yerrace
From: wdmaudiodev-bounce@xxxxxxxxxxxxx<mailto:wdmaudiodev-bounce@xxxxxxxxxxxxx>
[mailto:wdmaudiodev-bounce@xxxxxxxxxxxxx<mailto:wdmaudiodev-bounce@xxxxxxxxxxxxx>]
On Behalf Of Basileios Bu
Sent: Friday, May 1, 2015 10:34 AM
To: wdmaudiodev@xxxxxxxxxxxxx<mailto:wdmaudiodev@xxxxxxxxxxxxx>
Subject: [wdmaudiodev] Re: Question for MIDI on older W10 Technical Preview
(10061)
Hello,
Out of curiosity, is it possible to describe the nature of the bug? Was it
related to the pin factory FriendlyName? Or something else in the topology? I
only ask out of curiosity, as all is so far well now on 10074. But if I could
put in the one (or more) extra changes that would have made the difference for
10061, things might be better.
Thank you,
Basil
On Fri, May 1, 2015 at 12:22 PM, Frank Yerrace
<Frank.Yerrace@xxxxxxxxxxxxx<mailto:Frank.Yerrace@xxxxxxxxxxxxx>> wrote:
There was indeed an OS bug that broke most MIDI devices for a series of builds.
From: wdmaudiodev-bounce@xxxxxxxxxxxxx<mailto:wdmaudiodev-bounce@xxxxxxxxxxxxx>
[mailto:wdmaudiodev-bounce@xxxxxxxxxxxxx<mailto:wdmaudiodev-bounce@xxxxxxxxxxxxx>]
On Behalf Of Basileios Bu
Sent: Friday, May 1, 2015 9:10 AM
To: wdmaudiodev@xxxxxxxxxxxxx<mailto:wdmaudiodev@xxxxxxxxxxxxx>
Subject: [wdmaudiodev] Question for MIDI on older W10 Technical Preview (10061)
Hello.
Under Windows 10 technical preview build 10061, our MIDI implementations based
both on portcls and AVStream suddenly stopped working, in that they did not
appear in the list of MIDI devices on the system. Indeed, WPP traces indicated
that the filters in question were simply not being opened. That convinced me
that it was a matter of stricter topology requirements. So based on this, I
tried to make sure that everything in our topologies was as close as that of
MS, whose devices of course still functioned. Unfortunately, I was not close
enough, and our implementations started to work again under build 10074. I
think that all that I had missed in my earlier attempts were the friendly names
of the pin factories, and I cannot easily revert builds to test. Can someone
at MS (or anyone who knows this information) please confirm if that is indeed
the case, or if not, what were the changes from 10061 to 10074 for MIDI?
Thank you.
Basil
