🚀 go-pugleaf

chucks@skypoint.com (Erik Funkenbusch) writes:
>Illya Vaes (ivaes@hr.ns.nl) wrote:
>: Why don't you ask Microsoft, why they left in all the Windows 3.x nonreentrant
>: stuff that stops the multitasking when accessed (which is also done for win32
>: apps).
>You know.. i keep seeing people say this.  Now, I'm an amiga person through
>and through, have been since 89.. but lying is not going to get you anywhere.
>The Win16Mutex in Windows95 does not stop multi-tasking, it causes certain
>threads to wait if they need to make use of certain 16 bit functions.
>This does not prevent applications making calls to non 16 bit code from 
>being scheduled (and pre-empting the 16 bit subsystem) nor does it stop
>device drivers and/or interrupt processing from being handled.  Nor does
>it do anything else that might lead you to believe it's not pre-emptive.

Every "Windows" app in Windows 95 (I repeat, *every*) depends on the non-
reentrant win3.x stuff they left in. This includes win32 obviously. Now you
might say it will not have "noticable" effect in "reality", but it *is* there.
It is not for nothing that almost everyone thinks OS/2 multi-tasks better than
Windows 95.
"Causing certain threads to wait" is also called blocking, which means that
the threads are no longer runnable. Therefore, the threads have stopped being
multitasked. Only runnable threads are multitasked. According to you, when a
win16 app hogs the system in win3.x (or has hung) , the system is still multi-
tasking it and other (Windows) apps.

The inventivity of people trying to "explain away" negative aspects of MS 
software...
-- 
Illya Vaes  (ivaes@hr.ns.nl)
Holland Railconsult BV, Signalling CAD
Postbus 2855, 3500 GW Utrecht
Tel +31.30.358586, Fax 357208           "Do...or do not, there is no try" - Yoda