sancho1980 wrote:
> hi

> first of all: why have there been so few posts these last few days???
> no on actually answered to my last to posts...not that anyone would
> *have* to answer, i'm just curious...
> i finally did get problem 44 at end of chapter 2 done with the help of
> tom?s (thanx)

I don't have the book (only the second edition) so I can't really help you
with questions related to "the book".

--
Jens de Smit
Student Computer Science | Vrije Universiteit Amsterdam
jfdsmit@few.vu.nl | http://www.few.vu.nl/~jfdsmit

"sancho1980" writes:

> hi
>
> first of all: why have there been so few posts these last few days???

Must be upcoming chrismas. :-). I noticed the same lack of messages in
other usenet groups.

> i finally did get problem 44 at end of chapter 2 done with the help of

> now i was starting to have a look at the remaining hacks and it seems
> to me like problem no 50 is more or less the same, or what is meant by
> "statistics"??


BTW: To which version of OSDI do you refer?

Regards -- Markus

Markus E Leypold schrieb:

> "sancho1980" writes:
>
> > hi
> >
> > first of all: why have there been so few posts these last few days???
>
> Must be upcoming chrismas. :-). I noticed the same lack of messages in
> other usenet groups.
>
> > i finally did get problem 44 at end of chapter 2 done with the help of
>
> > now i was starting to have a look at the remaining hacks and it seems
> > to me like problem no 50 is more or less the same, or what is meant by
> > "statistics"??
>
>
> BTW: To which version of OSDI do you refer?

3rd edition. Problem 44 says: extend the kernel in a way that it keeps
track of which processes have sent how many messages to which other
processes, then print out this matrix when the user hits f4 (i used f12
so i didnt override any pre-built function
and the last problem says something about providing statistical
information about messages in the system...what does this mean? shall i
printf out the mean message number all processes receive and send? of
course, you could make a ranking showing which processes are the most
copious letter-writer but i'm not sure this is what is meant here

merry christmas (ach, i hate it...) and thanx,

martin

>
> Regards -- Markus

"sancho1980" writes:

> Markus E Leypold schrieb:
>
>> "sancho1980" writes:
>>
>> > hi
>> >
>> > first of all: why have there been so few posts these last few days???
>>
>> Must be upcoming chrismas. :-). I noticed the same lack of messages in
>> other usenet groups.
>>
<...>

>> BTW: To which version of OSDI do you refer?
>
> 3rd edition. Problem 44 says: extend the kernel in a way that it keeps
> track of which processes have sent how many messages to which other
> processes, then print out this matrix when the user hits f4 (i used f12
> so i didnt override any pre-built function
> and the last problem says something about providing statistical

Perhaps the exact wording could provide clue what they mean? Difficult
to say, since I don't own OSDI/3 (yet) -- I'll wait until I really
need it or the first used copies hit the market.

> information about messages in the system...what does this mean? shall i
> printf out the mean message number all processes receive and send? of

Not having hte exact wording of problem 44 I can only guess, but
something like the average message frequency, perhaps lenght, somehow
averaged over the last minutes, the total number of messages send,
when the last message where send, would perhaps be useful for
debugging. I'm only guessing, of course. Concentrate on something like
the message frequency received/sent from/to a particular process
averaged over the last N minutes: That should be the typical problem,
all other statistical information would be recorded / printed in a
very similar way.

> course, you could make a ranking showing which processes are the most
> copious letter-writer but i'm not sure this is what is meant here

>
> merry christmas (ach, i hate it...) and thanx,

I feel with you :-).

Regards -- Markus

> Not having hte exact wording of problem 44 I can only guess, but
> something like the average message frequency, perhaps lenght, somehow

s/hte/the/; s/lenght/length/

> averaged over the last minutes, the total number of messages send,
> when the last message where send, would perhaps be useful for

s/send/sent/; s/where/were/ ...

Terrible. This was written in the last stages of coffein + sugar
withdrawal. Please excuse the errors ... :-).

I also should run my spellchecker now and then.

Regards -- Markus