Re: "Real" Multitasking (Re: GNO questions...)

[slp@vax1.mankato.msus.edu]

In article <1993Apr22.154817.5639@slab.slip.uiuc.edu>, derek@slab.slip.uiuc.edu (Derek Taubert) writes:
> slp@vax1.mankato.msus.edu said:
>> Sorry for the second post in a row, but I just reread my post and I think I
>> failed to emphasize what was in fact the point of my original post, and the
> 
> It would be more appropriate that you apologize for making the FIRST post.  

Alright, my apologies for stating my opinions in the first place.  As Derek has
pointed out, no one has the right to state their own opinions unless they agree
with his.  I should have cleared these posts with him first.  I'm truly sorry.

>> Likewise, if someone asks, "Which is better, cooperative or preemptive
>> multitasking?" an ignorant individual would reply "[one or the other] is
>> better."  A more knowledgable programmer knows that each has its advantages
>> and disadvantages.  Only the ignorant believe one is better than the other,
>> period.
> 
> So why are you raising the issue without any solid grounds whatsoever?  You
> seem to admit that your own actions are counter productive.

Derek, you just haven't got a clue, have you?  The whole point of my posts has
been to try and explain that both approaches have their own merits.  I have
plenty of solid ground to stand on for this point.  You seem to think my point
is something other than this.  What, I don't know...

>> Someone actually sent me mail saying they were glad I posted what I originally
>> did, because everytime they mention [their favorite multitasker], people jump
>> on them and berate them because it isn't preemptive.  This is ludicrous
>> behavior.
> 
> Well, indeed this behavior is unfounded, but who made you the defender of the
> earth?

Again, my sincere apologies for stating my own opinions and defending my owbn
beliefs.  I forgot I had no right to do so when they conflict with yours.

>> I rose to the defense of cooperative multitasking because everyone
>> seems to be putting it down.
> 
> Cooperative multitasking is neither a physical entity nor a religion, it does
> not need to be defended, especially not by you.

If this is the case, why is everyone so convinced of the need to attack it?
It neither NEEDS to be attacked not defended.  But as long as so many people
are doing the one, why can't I do the other?  Ooops, I forgot, it's because
your opinion is different.  I keep forgetting I only have the right to defend
my opinions if you agree with them.  Sorry.

>> So, what I'm really saying is this: stop ragging on one or the other.  It's
>> pointless.
> 
> Pardon me for being the one to point out to you that you were in fact the
> aggressor of this entire discussion, and really should practice more of what
> you preach (right or wrong), or stop preaching.

I think you're confused.  I was not the aggressor.  I saw people posting about
cooperative multitasking not being "real", apparently in completely ignorance
that the only difference between the two is when they switch processes.  I
attempted to DEFEND cooperative multitasking.  I was then jumped on be people
like yourself who seem upset that I'm willing to put forward an opinion counter
to their own.

>> If you believe [type1] multitasking is better than [type2]
>> multitasking in all situations, you're an idiot, and you only display your
>> ignorance whenever you rag on one type or the other because you believe
>> [type1] is always better in all situations.
> 
> This may or may not be true, but so far you haven't presented any valid
> arguments that would place cooperative above preemptive.  Yes, the comparison
> fields are extremely situational, but I am positive that you don't have enough
> experience with operating system implementations to be making any sort of
> conclusion.  Your entire argument centers on the falicy that since cooperative
> multitasking has less overhead involved in switching, it is therefore better
> for single user systems.  The other aspects involved in making an operating
> system work far outweigh this single issue, and you haven't even given them any
> consideration.

There are a number of other reasons, some of which I've touched on briefly.
I've always found one of the nicest things about cooperative multitasking is
that you don't have to muddle with semaphores everytime different threads in
your program want to access global data, OS routines, and so on.  This is the
main reason why cooperative is usually best for running multiple threads in a
single program.  If you're implementing multitasking on the OS level, this is
less of a concern and the preemptive approach becomes more attractive.  Still,
there may be other considerations.  You probably see this admision as a
contradiction to my previous posts, but then you have this tendency to see
everything in black and white, and have thus missed the whole point of my
previous posts (which was that the whole issue is essentially grey).

> And, oh, thank you for judging the entire population's mental capacity based
> upon a single issue.  Your impetuous behaviour is outdone only by your lack of
> sagacity.  Take your "expert" opinion and present it to a crowd of fools, then
> you may have something...
> 
> from previous article:
>> People who refer to preemptive multitasking as "real" multitasking are
>> people who know sh*t about what's what.
> 
> current:
>> Neither approach is the best, nor can you
>> know which is the best for a situation until you've fully explored the pros
>> and cons of that situation.  I gave my rules of thumb that I use in my
>> original post, but I never meant them to be taken as anything more than that.
>> They were only rules of thumb, and weak ones at that.
> 
> Well, now.  Seems your tone has changed substantially in just a few short
> sentences.  Don't back out now, you're just about to be clobbered.  Tell me
> more about how efficient cooperative multitasking is.

I did refer to ignorance in my previous post, and thankt you for just making my
point.  Your quotes "from previous article" and "current" are not in any way
contradictory.  I have neither backed down one iota nor has my tone changed
substantially.  My original point was that neither cooperative nor preemptive
multitasking is more "real" than the other, that each has its merits, and I
have yet to say anything contradictory to this.

>> There is no one answer to the question, so let's stop stomping on [such and
>> such a system] just because it uses [such and such an approach].
> 
> There's the smartest thing I've heard you say.  I challenge you to listen to it.
> Especially since you are obviously gasping for a way to get out of an argument
> you shouldn't have started in the first place.

Heh!  I should know better than to get into a battle of wits with someone who
is obviously unarmed.

-GTR