[Muscle] bug in SCardReleaseContext?

[Eric Walter]

Hi Ludovic,

> We can generalize this and call SCardRemoveContext() every time the
> daemon is not present anymore (before returning SCARD_E_NO_SERVICE) to
> free the allocated context.
> So future PC/SC call would get SCARD_E_INVALID_HANDLE.

That make sense in a way, but I thought SCARD_E_INVALID_HANDLE would  
apply to errors concering an SCARDHANDLE, not an SCARDCONTEXT? It  
doesn't really matter so much, as there really isn't anything much to  
do with an error returned from SCardReleaseContext except for maybe  
logging it.

> Does any body knows the Windows PC/SC behavior in such a situation?
>
> Do you have a particular setup so you (often) get the  
> SCARD_E_NO_SERVICE error?

Not particularly, just trying to get things as stable as possible. We  
have users from Mac OS X 10.2 to 10.5 with all sorts of different  
configurations. And quite a few users that seem to like keeping our  
application running 24/7 with phases of sleep mode in between.

Regards, Eric