Author |
Message |
|
Post subject:
Posted: Jan 09, 2006 - 11:05
|
|
Order of the Butterfly
Joined: Jan 14, 2004
Posts: 1940
Location: Karlsborg,Sverige
|
|
yes but does it say something about good applications like browsers and office? |
_________________ RR!
|
|
|
|
|
|
Post subject:
Posted: Jan 09, 2006 - 11:17
|
|
Newbie
Joined: Jul 12, 2005
Posts: 14
|
|
Regardless of what operating system one prefers I think there's only one word worth uttering about that article: CRAP |
|
|
|
|
|
|
Post subject:
Posted: Jan 09, 2006 - 11:24
|
|
Order of the Butterfly
Joined: Jan 14, 2004
Posts: 1940
Location: Karlsborg,Sverige
|
|
|
|
|
|
Post subject:
Posted: Jan 09, 2006 - 11:32
|
|
Order of the Butterfly
Joined: Aug 26, 2003
Posts: 1784
Location: Malmö
|
|
I like how he has allocated one column for "Linux" as if all distributions are exactly the same. Very scientific study.
Quote: | There is not many likely articles to find which includes MorphOS yet, because its still a quite young operating system, but it also clearly have 'roots' from the Amiga, and they are very close to each other, why this also 'can' be considered appliceable also for a MorphOS system. |
MorphOS is mentioned as being something without Zen!
Quote: |
With the Zen established, we can consider what other things reflect similar Zen:
* PowerPC
* AmigaOS 4
* Chris Hodges' Poseidon USB stack
* ReAction
* DataTypes that recognise files by their content
* INet225 and socket.library
and those which don't:
* Linux
* x86
* MorphOS
* AmigaDE
* GCC
* PalmOS (the Zen of which is more akin to the original MacOS than anything else)
* Using file extension to identify a file's types
* AmiTCP and bsdsocket.library
|
|
_________________ I need this baby in a month send me nine women!
|
|
|
|
|
|
Post subject:
Posted: Jan 09, 2006 - 12:05
|
|
Order of the Butterfly
Joined: Jan 14, 2004
Posts: 1940
Location: Karlsborg,Sverige
|
|
Zen for me Zero Effort network or networking |
_________________ RR!
|
|
|
|
|
|
Post subject:
Posted: Jan 09, 2006 - 15:04
|
|
Order of the Pegasos
Joined: Jan 22, 2004
Posts: 2517
Location: Göteborg
|
|
Pardon my French..
How the fsck can someone say that MacOS X is only 'Good' when it comes
to ease of use. There is no easier OS in the whole world than MacOS X.
Windows is years behind OS X when it comes to ease of use. |
|
|
|
|
|
|
Post subject:
Posted: Jan 09, 2006 - 21:05
|
|
Order of the Pegasos
Joined: Aug 16, 2003
Posts: 2538
Location: Göteborg
|
|
Hello,
Ok, Im back again
Thread moved and splitted yes,.. ok as the germans tends to say,.. Keine Problem !
The article I linked too is of course the words of the writer of it. And my aim was to show that people do have opinions, and of course its not anything wrong with that, instead everybody is free to have any opinion, at least in in my opinion.
Anyway, results might not always be what You expect them to be. I got some results concerning memory allocation that I would like to share then.
The below test is said to have been made on a Pegasos 1 in MorphOS and in Linux operating system. Please note that I have not run this myself yet, and my guess is also that if You do run the tests on a Pegasos II, you might get completely different results. I just compiled it here now.
occount is the outer loop
icount is the inner loop
size is the size in bytes thats allocated
then column which tells how long time it took in MorphOS respectively in Linux
Here is the table and source code:
Code: |
/*
ocount icount size Mos Linux
10000000 2 10 56 5
10000000 2 64 53 5
10000000 2 256 55 10
10000000 2 1024 50 9
10000000 2 5120 58 9
10000000 2 10240 59 9
1000000 2 1024000 5 60
1000000 20 10 54 5
1000000 20 64 53 5
1000000 20 256 53 8
1000000 20 1024 58 8
1000000 20 5120 60 8
1000000 20 10240 60 144
100000 20 1024000 6 61
100000 200 10 54 5
100000 200 64 53 5
100000 200 256 54 8
100000 200 1024 63 40
100000 200 5120 65 309
100000 200 10240 64 316
10000 200 1024000 18 75
10000 2000 10 54 5
10000 2000 64 61 10
10000 2000 256 70 29
10000 2000 1024 105 100
10000 2000 5120 129 408
10000 2000 10240 411
*/
//#define HAVE_EXEC
#include <stdio.h>
#include <stdlib.h>
#include <time.h>
#ifdef HAVE_EXEC
# include <exec/memory.h>
#endif
int main(int argc, char *argv[])
{
void **mem;
int ocount, icount, size, j;
time_t t1, t2;
if (argc != 4) {
printf("Usage: mtest <outercount> <innercount> <size>\n");
return 0;
}
ocount = atoi(argv[1]);
icount = atoi(argv[2]);
size = atoi(argv[3]);
mem = malloc(sizeof(void *) * icount);
time(&t1);
for (j = 0; j < ocount; j++) {
int i;
for (i = 0; i < icount; i++) {
#ifndef HAVE_EXEC
mem[i] = malloc(size);
#else
mem[i] = AllocVec(size, MEMF_PUBLIC);
#endif
}
for (i = 0; i < icount; i++) {
#ifdef HAVE_EXEC
FreeVec(mem[i]);
#else
free(mem[i]);
#endif
}
}
time(&t2);
free(mem);
printf("Done... Time elapsed: %fs\n", difftime(t2, t1));
return 0;
}
|
More to discuss |
_________________ Mvh Gunne
|
|
|
|
|
|
Post subject:
Posted: Jan 10, 2006 - 04:09
|
|
Order of the Butterfly
Joined: Jan 14, 2004
Posts: 1940
Location: Karlsborg,Sverige
|
|
ja koden ser b0rkad ut.. |
_________________ RR!
|
|
|
|
|
|
Post subject:
Posted: Jan 10, 2006 - 07:13
|
|
Order of the Butterfly
Joined: Aug 26, 2003
Posts: 1784
Location: Malmö
|
|
@lisardman:
This forum is in english
@GGS:
Why not run a whetstone or dhrystone benchmark instead. They aren't very high regarded as benchmarks these days but they are a lot better than what you pasted above. The source code to the original whetstone and dhrystone benchmarks are freely available here.
The above benchmark means nothing if it doesn't state what kind of system it was supposed to benchmark (server/desktop/embedded) and how the Linux kernel was compiled to accomodate that particular configuration. Was it compiled to do caching, does it run with preemption (low latency higher context overhead) and so on.
BTW, try enclosing your code in [ code] .. [/code] for readability in the forum since it will be treated as verbatim rather than prose. |
_________________ I need this baby in a month send me nine women!
Last edited by dholm on Jan 10, 2006 - 10:43; edited 3 times in total
|
|
|
|
|
|
Post subject:
Posted: Jan 10, 2006 - 07:17
|
|
Order of the Butterfly
Joined: Jan 14, 2004
Posts: 1940
Location: Karlsborg,Sverige
|
|
dholm wrote: | @lisardman:
This forum is in english
|
okej then.. The code is b0rked |
|
|
|
|
|
|
Post subject:
Posted: Jan 10, 2006 - 08:39
|
|
Order of the Pegasos
Joined: Aug 16, 2003
Posts: 2538
Location: Göteborg
|
|
Hello,
Well, at least this guy who made this program and tests have tried to do something himself.
Neither have he tried to explain in the forum why Linux feels behave slower then MorphOS
All the best,
Gunne |
_________________ Mvh Gunne
|
|
|
|
|
|
Post subject:
Posted: Jan 10, 2006 - 09:35
|
|
Order of the Butterfly
Joined: Aug 26, 2003
Posts: 1784
Location: Malmö
|
|
|
|
|
|
Post subject:
Posted: Jan 10, 2006 - 09:56
|
|
Journeyman
Joined: Jun 20, 2004
Posts: 42
Location: Munkedal
|
|
So linux is several times faster than morphos in that test? Numbers in time? |
_________________ Amigaföredetting
|
|
|
|
|
|
Post subject:
Posted: Jan 10, 2006 - 10:04
|
|
Order of the Pegasos
Joined: Aug 16, 2003
Posts: 2538
Location: Göteborg
|
|
d'olen,
Well, as dholm wrote above, its difficult to say how good this code really is. I feel myself the results looks, hmmm.. little bit strange. I compiled and did run it on my Pegasos II with MorphOS and got very different results.
How elapsed time can be so like for different sizes as in the table above looks for me also little strange. |
_________________ Mvh Gunne
|
|
|
|
|
|
Post subject:
Posted: Jan 10, 2006 - 10:49
|
|
Order of the Butterfly
Joined: Jan 14, 2004
Posts: 1940
Location: Karlsborg,Sverige
|
|
in Linux you have working Altivec such thing doesn't exist in morphos |
_________________ RR!
|
|
|
|
|
|