Quoting Oliver Schwartz <[email protected]>: > Hi, > > > snapscan isn't compiled at all if ipc/semphore headers aren't found. > > At least that's the intention. > > > > > My fix for this is to do something like > > > > > > #ifdef OS2 > > > #include "snapscan-usb-stubs.c" > > > #else > > > #include "snapscan-usb.c" > > > #endif > > > > > > where the 'stubs' version doesn't include the offending header file, > but > > > it does have empty function bodies to keep the linker happy. > > Not a very nice solution in my opinion.
Expedient. > > To the snapscan developers: can this be implemented? Or something > > similar, e.g. #ifdefing everything that uses ipc.h and sem.h? Is > > sem.h/ipc.h needed at all? > > The usb part uses semaphores to provide atomic commands. It may be > possible > to replace the semaphore code by something similar, it's only used at > one > place. But then, I'm quite sure that the OS/2 API also supports > semaphores > (DosSemGet() etc., if my memory serves me right). So I'd suggest to map > the > semaphore operations on OS/2 semaphores. > > For a start we can use separate functions for the semaphore operations > (and > maybe bundle them in a separate file), e.g. snapscani_semaphore_open(), > > snapscani_semaphore_close(), snapscani_semaphore_wait(), > snapscani_semaphore_signal. Until an OS/2 implementation is provided the > OS/2 > version of the functions may be left empty (maybe issuing a warning > message). > However, the OS/2 implementation should be simple. Are there any developers that have OS/2 and a USB scanner? For that matter, are there any backends other than snapscan that use semaphores with USB? A bientot Paul -- Paul Floyd http://paulf.free.fr for those with time to waste What's the point? The sharp thing on the end.
