MCIsendCommand() vs. MCIsendString()
Walter F. Dexter
In a previous article, iv...@syl.sj.nec.com (Ivan Fuller) says:
>
>What are the disadvantages (if any) of using MCISendString() instead of
>MCISendCommand()?
>
>I still need all of the callback capability and I need to be able to send
>window handles, etc to the devices, but I think MCISendString() can handle
>that. (I have sent string commands containing window handles via IconAuthor
>before...)
>
>It seems to be much easier to support many device types by using the string
>interface instead of the API interface... This way, we can provide support for
>any media type that has a string interface without having to recompile our code
>(e.g. just let our user pass the device name as part of an MCI string command.)
>
>Does anyone know of any features that are available ONLY in MCISendCommand()
>and not MCISendString()?
>
>-Thank You
>
>Ivan Fuller <iv...@syl.sj.nec.com>
>
>
Ivan,
I'm not aware of any features that are MCISendCommand() only. The big issue,
depending on the host system, is probably speed. I've written MCI drivers; what
the driver gets for a command, whether sent as a string or a C structure,
is the MCISendCommand() format. MMSYSTEM has to translate the strings into
the C structure format. I don't know how quick/effecient it is at doing that.
FWIW, I've always used strings.
Walt