home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!mcsun!Germany.EU.net!gmd.de!newsserver.jvnc.net!yale.edu!yale!gumby!destroyer!gatech!udel!darwin.sura.net!zaphod.mps.ohio-state.edu!rpi!ghost.dsi.unimi.it!univ-lyon1.fr!frmop11!psuvm!cunyvm!ccvqc
- Organization: City University of New York/ University Computer Center
- Date: Tuesday, 26 Jan 1993 10:50:57 EST
- From: Christopher Vickery <CCVQC@CUNYVM.BITNET>
- Message-ID: <93026.105057CCVQC@CUNYVM.BITNET>
- Newsgroups: comp.realtime
- Subject: Re: What's wrong with DOS in real-time?
- References: <93005.121935CCVQC@CUNYVM.BITNET> <sT3XwB1w165w@ade.no>
- <nhyrx5#@quantum.uucp> <1993Jan22.160747.9625@iccgcc.decnet.ab.com>
- Lines: 47
-
- In article <1993Jan22.160747.9625@iccgcc.decnet.ab.com>,
- sinclair@iccgcc.decnet.ab.com (Andrew Sinclair) says:
-
- >I'd like to add a column to capture the approx. number of sites doing
- >active software development in each o/s. (Especially development
- >involving windows).
-
- I would be leery of answers to this one that are supplied by vendors
- or users. The vendors may (or may not) be willing to say how many
- copies of their OS they have sold, but that doesn't tell how many of
- those sales represent true development sites. It's like Microsoft's
- claims about how many copies of Windows have been installed compared
- to what many people suspect is a significantly smaller number of
- actual Windows users. The problem is particularly pernicious for an
- OS that has been around a long time, is available in different
- versions, etc. The only meaningful answer would come from an
- independently-conducted survey, and I'm not sure the market is large
- enough to generate interest in conducting such a survey.
-
- >
- >What about the impact of the rtos on the integrity of Windows and its apps?
-
- Murky question there. What do you mean by "integrity"? If you mean,
- "does the memory management policy, or any other property of the rtos,
- cause DOS and/or Windows apps to fail", then you get an answer like:
- "Such bevior represents a bug in the rtos, which will be fixed 'real
- soon.'" On the other hand, if you are asking if the rtos will have a
- negative effect on the *performance* of DOS/Windows applications, then
- you are asking the wrong question. By definition, it is the real-time
- side of the application that must be given highest processing priority
- so that it does not miss deadlines. If you don't have deadlines, you
- aren't doing real-time. The real question, then, is whether DOS and/or
- Windows apps can interfere with the ability of the real-time side to
- meet its deadlines. I can answer for iRMX for Windows, where DOS (and
- Windows, if run) is/are encapsulated as a Virtual Machine-86 (VM86)
- hardware task. By running this task at the lowest priority, iRMX for
- Windows guarantees that real-time tasks will always get the CPU when
- they need it. This may mean poor DOS-side responsiveness when there is
- real-time processing to be done, but that's the way it's supposed to
- be. If you are doing real real-time work, you can't let DOS or Windows
- or their applications interfere with the responsiveness of the real-time
- side. That's why I don't understand what Dan says when he says that QNX
- runs DOS in real mode rather than VM86 mode. A true real-mode app would
- be able to starve the real-time side.
- -------
- Christopher Vickery, Computer Science Department, Queens College (CUNY)
- Flushing, NY 11367-0904 vickery@ipc1.cs.qc.edu (718)997-3500
-