home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.sys.amiga.misc,comp.sys.amiga.programmer
- Path: netcom.com!mbanet
- From: mbanet@netcom.com (David Waters)
- Subject: Re: toy operating systems, like AmigaOS
- Message-ID: <mbanetDptG63.HJG@netcom.com>
- Followup-To: comp.sys.amiga.misc,comp.sys.amiga.programmer
- Organization: NETCOM On-line Communication Services (408 261-4700 guest)
- X-Newsreader: TIN [version 1.2 PL1]
- References: <skllsf.984983.4.5@groomlake.mil>
- Date: Sat, 13 Apr 1996 19:44:27 GMT
- Sender: mbanet@netcom8.netcom.com
-
- Duanne Barry (dbarry@groomlake.mil) wrote:
- : I've just spent the last 2 months coding a large and complex algorithm
- : on a Unix box (Sun Sparc 10).
-
- <Readers: prepare yourself for an *AMAZING* comparison...>
-
- Duanne Barry (dbarry@groomlake.mil) wrote:
- : I've have made hundreds of critical bugs that generally
- : just cause the entire code to blow up (eg. memory overwriting, runaway
- : loops, etc). But never ever did any of my bugs actually crash the Unix
- : box (ie. it never needed a reboot) due to the lovely memory protection.
-
- First of all, take a moment to consider the manhours that went into the
- development of UNIX versus AmigaDOS/Workbench. Consider all of the
- overhead that UNIX requires. Now make the same comparison to MicroSoft
- and Apple's operating systems. Besides, UNIX (SVR4) is one of the boot
- options on my A3000 but I only use UNIX to keep my UNIX System
- Administration skills sharpened.
-
- The tradeoff of memory protection is efficiency and speed at the risk of
- someone being careless with the use of memory pointers. Without memory
- protection, an independent process (p2) may have *direct* access to the
- internal data of another one (p1). This scenario prevents the need to
- copy contents to a "public" location and the use of a "watch dog" to
- guard your "private territory".
-
- One of the greatest strength's of the Amiga's operating system is its
- inter-process message handling system *without* memory protection. Doing
- it this way maximizes efficiency by avoiding AT LEAST one level of
- overhead.
-
- In a perfect world developers would "play by the rules" and take the
- appropriate precautions. Even if they don't, the marketplace should be
- a great regulator of those who play by the rules and those who don't.
- I cite the existence of ARexx ports on major Amiga applications as an
- example of this.
-
- Duanne Barry (dbarry@groomlake.mil) wrote:
- : This is completely different from coding on an Amiga 4000
- : (which i did for a few years), where a serious bug just takes out
- : the entire machine (thus significantly lengthening coding time)... from
- : this experience, AmigaOS looks like a complete toy to me, even though the
- : machine has a MMU available...
-
- Okay, let's combine your statement(s) with what I said above to develop
- a pardigm using bicycles. Would you consider a bicycle with "training
- wheels" more of a toy than one without them? Assuming that you would, can
- you think of extra layers of overhead as "training wheels"? If so, can
- you see that the Amiga is an operating system that let's you ride without
- training wheels?
-
- Sure, it's potentially more dangerous to ride a bicycle without training
- wheels but if you're good, and careful enough, you'll have more fun and
- freedom because you'll carry less weight, can squeeze into more narrow
- openings, and go a lot faster!
-
- Duanne Barry (dbarry@groomlake.mil) wrote:
- : So, to conclude this monologue, in my view the Amiga will not be taken
- : seriously until it has fast hardware and a robust operating system...
- : I don't see this happening for at least 2 years unless someone ports
- : Linux to the upcoming PowerAmiga (if it ever comes out).
-
- Ignoring the fact that you're comparing systems that aren't even in
- direct competition (AmigaDOS/Workbench is SINGLE USER multitasking while
- UNIX is MULTI-USER multitasking), somehow you seem to have reached the
- conclusion that the overwhelming majority of computer buyers/users based
- their choice on the strength and weaknesses of the operation system alone.
-
- I'd say that marketing and the desire to be "compatible" are the two
- major factors behind more than 95% of all computer purchases. Most
- computer users only care about being able to use a few specific
- applications that they saw somewhere else. They don't give anymore
- thought to the OS than to say "oh, that's the thinggy that helps me run
- <insert random application program here>".
-
- David "Always chooses CAPABILITY over compatibility!" Waters
- --
- _/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/
- _/_/ David C. Waters, Jr. _/_/ Integrated Business Solutions _/_/
- _/_/ Multimedia Business Answers _/_/ Video Audio CD-ROM Publishing _/_/
- _/_/ E-Mail: mbanet@netcom.com _/_/ World Wide Web and TV Access! _/_/
- _/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/
-