From: agate!spool.mu.edu!wupost!zaphod.mps.ohio-state.edu!ub!dsinc!gvlf3.gvl.un isys.com!email!parkhill@ucbvax.Berkeley.EDU (parkhill)
Subject: Re: Ada Tasking problem
Date: 25 Nov 91 17:02:16 GMT [thread overview]
Message-ID: <5860@email.sp.unisys.com> (raw)
Humble suggestion:
Why not treat your output console as a shared device and write a
standard interface to the device? Write a package specification
that provides all the operations you require (locking the device,
time sharing, type and format of output, whatever). Hide the exact
method of implementation of output and locking of the device in
the body of a package. The body of the package will probably require
another task be added.
I have found that it can be unwise to let tasks use an output device
in an uncontrolled/unsynchronized manner. Putting the control of
the synchronization in a single package keeps the code in the package
instead of every task that might want to access a device (text_IO
console). The body may have to be implemented in a target dependant
manner but the other code remains more or less portable.
Robert Parkhill
next reply other threads:[~1991-11-25 17:02 UTC|newest]
Thread overview: 6+ messages / expand[flat|nested] mbox.gz Atom feed top
1991-11-25 17:02 agate!spool.mu.edu!wupost!zaphod.mps.ohio-state.edu!ub!dsinc!gvlf3.gvl.un [this message]
-- strict thread matches above, loose matches on Subject: below --
1991-11-26 9:10 Ada Tasking problem paul goffin
1991-11-25 4:04 cis.ohio-state.edu!pacific.mps.ohio-state.edu!linac!uwm.edu!ogicse!milton
1991-11-25 0:04 Bob Kitzberger
1991-11-23 0:03 cis.ohio-state.edu!zaphod.mps.ohio-state.edu!uakari.primate.wisc.edu!usen
1991-11-21 13:04 Eric E. Mays 52202
replies disabled
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox