Well, all of what you wrote i read about but i didn't expect it to be that bad.. I'd just like to mention that i probably won't need DMus but rather simply the MIDI part of the Portcls. Also I'm always happy to learn, I started programming with a tutor ( a good one^^ ) and learned the rest by reading into stuff. So if you really think that i need a Tutor, do you know any one in this community who is happy to teach? Because I really want this application done and (as usual) its just windows beeing "special" and not offering virtual midi ports by default... > Date: Sun, 15 Jul 2012 01:18:02 +0200 > Subject: [wdmaudiodev] Re: Starting to dev. a virtual Midi Port > From: noise@xxxxxxxx > To: wdmaudiodev@xxxxxxxxxxxxx > > How much effort did you plan to put into this? > I would say its probably impossible to start MIDI portclass driver > development by reading the WDK documentation. You will need to learn > how to develop and debug a Windows driver. (There is some Literature), > but I would say its pretty trial and error only - if reading thru WDK > and MSDN is the only source of information, which in turn is pretty > cluttered information. Then the messy portclass/adapter, with its > unnecessary COM MACRO obsession, for which some here probably still > think there isn't even a well-tested Microsoft compiler, because its > C++ and C++ is forbidden in the kernel (see the rather religious > discussion on the Windows driver list). But who would expect any > Literature? On top of it the IMiniportDMus bureaucracy, which you will > need for your adventure. How would somebody explain how to set the > MIDI pin names in the .inf to expose to applications? And then all the > kernel bugs and tons of undocumented or wrongly documented behavior > you need to ship around, which we learnt to bypass and circumvent as > ignored by Microsoft. Each BSOD which blocks your test machine for 2 > minutes to boot up again and login and setup everything for BSOD again > until something destroyed some "important" info in the registry and > you need to re-install your OS, which will disallow registering your > Windows key after 10? times, so no chance without a 3000$ MSDN > account. Oh, and you will probably need to test on XP, Phista, 7 in 32 > and 64bit flavors. Because each of the OS behave differently and have > a different level of bugs, no, those are features. Well, did I mention > driver signing and installation? You will learn that a difx merge > module can't have the WDF coinstaller in the same component, because > it will not get uninstalled, and uninstallation is one of the most > important features when testing your software. Oh, right your > portclass might not need WDF, but even that might not be written > somewhere. > > I would go down to the beach an order some beer and make my life fun, > as long as you don't have a good tutor and access to a working > IMiniportDMus example and somebody showing you how to install and > debug this. > > On Sat, Jul 14, 2012 at 8:43 PM, Revan Foly <xaseriii@xxxxxxxxxx> wrote: > > Hi, > > > > I decided to read though the WDK documenation again and I'm now back at the > > FMSynth example. I compiled it to a lib, however the documentation of the > > sample states that I'm supposed > > to link somewhere. ("which can be linked into an audio adapter driver.") Any > > ideas whats meant by this? > > > > I also tried to use the INF file of the AC97 sample, however when trying to > > install the AC97 driver with that INF, I get an error (translated) "This > > installation method is not supported by this INF file". > > > > - Xaser > ****************** > > WDMAUDIODEV addresses: > Post message: mailto:wdmaudiodev@xxxxxxxxxxxxx > Subscribe: mailto:wdmaudiodev-request@xxxxxxxxxxxxx?subject=subscribe > Unsubscribe: mailto:wdmaudiodev-request@xxxxxxxxxxxxx?subject=unsubscribe > Moderator: mailto:wdmaudiodev-moderators@xxxxxxxxxxxxx > > URL to WDMAUDIODEV page: > http://www.wdmaudiodev.com/ >