1. Bug in eu2c

Ditulis 10-9-2000, 14:31 pake kalong

this is from dos prompt window.....

read the comment starting with #


# Using interpreter
F:\AKU>ex acakaja c:\windows\system.dat c:\cthbak 212356715376
Ukuran file:         7344164 bytes
# this means: size of the file = 7344164
Waktu diperlukan:         15.16 detik
# this means: time to finish = 15.16 seconds
Kecepatan:            484443 bytes/detik
# speed = 484443 bytes/second


# using compiler
F:\AKU>acakaja c:\windows\system.dat c:\cthbak 212356715376
DOS/4GW Protected Mode Run-time  Version 1.97
Copyright (c) Rational Systems, Inc. 1990-1994

        Visit http://www.RapidEuphoria.com/
        to register the Euphoria to C translator.
        You'll eliminate this message and delay,
        and you'll get improved debug facilities.

Ukuran file:         7344164 bytes
# this means: size of the file = 7344164, same as above
Waktu diperlukan:        140.70 detik
# the time to finish: 140.70 seconds, but it is not correct,
# I think it is 14.07 seconds
Kecepatan:             52197 bytes/detik
# speed = 52197 bytes/second
# it is too slow, because the time is not 140.70 seconds
# but 14.07 seconds


It is created by using time()
so i think time() is not correct.

(sorry if my english is bad)


--
Aku

new topic     » topic index » view message » categorize

2. Re: Bug in eu2c

Aku writes:
> F:\AKU>acakaja c:\windows\system.dat c:\cthbak 212356715376
> DOS/4GW Protected Mode Run-time  Version 1.97
> Copyright (c) Rational Systems, Inc. 1990-1994
>...
> # the time to finish: 140.70 seconds, but it is not correct,
> # I think it is 14.07 seconds

Mic also reports a similar factor of ten error in the reported time.
I assume he used DOS4GW too.

I'm not interested in hearing about any more problems with
DOS4GW. Please download the latest translator package for
DOS32. It contains the CauseWay extender, and it links in
CauseWay automatically. As a bonus, it also compresses
your .exe using the CauseWay compressor. CauseWay is
what the Euphoria interpreter uses.

I used DOS4GW for Euphoria, versions 1.3 (5 years ago)
and earlier. I found it to be rather buggy, and I don't think
they updated it much after Windows 95 came out.
CauseWay is far superior. I tried a few small programs
using DOS4GW a while ago and they ran ok, so I thought
it might still be a viable option. But then I ran some larger ones
that failed, and people have reported lots of problems with DOS4GW
on this mailing list. I will not recommend that anyone try
DOS4GW or other extenders anymore.

And, yes, I'm aware that most people don't have Watcom.
I'm trying to find out if they can release some of the Watcom
executables, before they release their open source version.
If they can't, I will consider other compilers.

Since I've been building the Euphoria interpreter with Watcom,
it's easy for me to provide a translator that works with Watcom.
Providing support for other DOS/WIN compilers requires a
fair bit of work, so I'm not jumping into it right away.

Regards,
   Rob Craig
   Rapid Deployment Software
   http://www.RapidEuphoria.com

new topic     » goto parent     » topic index » view message » categorize

3. Re: Bug in eu2c

>Mic also reports a similar factor of ten error in the reported time.
>I assume he used DOS4GW too.

Nope. I'm using CauseWay.
Wich leads us to the question: "does time() work the way it should?"
_________________________________________________________________________
Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com.

Share information about yourself, create your own public profile at
http://profiles.msn.com.

new topic     » goto parent     » topic index » view message » categorize

4. Re: Bug in eu2c

Ditulis 11-9-2000, 22:12 > 10-9-2000 23:55 Robert Craig

R> Aku writes:
>> F:\AKU>acakaja c:\windows\system.dat c:\cthbak 212356715376
>> DOS/4GW Protected Mode Run-time  Version 1.97
>> Copyright (c) Rational Systems, Inc. 1990-1994
>>...
>> # the time to finish: 140.70 seconds, but it is not correct,
>> # I think it is 14.07 seconds

R> Mic also reports a similar factor of ten error in the reported time.
R> I assume he used DOS4GW too.

Now I have downloaded the update, and i have compiled using causeway.
But I still get the time error.....


F:\AKU>acakaja c:\windows\system.dat NUL 3198230982
#### No DOS4GW AGAIN... (no dos4gw message, and it is CAUSEWAY.)

        Visit http://www.RapidEuphoria.com/
        to register the Euphoria to C translator.
        You'll eliminate this message and delay,
        and you'll get improved debug facilities.

Ukuran file:         7344164 bytes
Waktu diperlukan:        108.70 detik
# Still 108.70... is should be 10.87 seconds...
# in the interpreter it is 12.85 seconds.....
Kecepatan:             67563 bytes/detik


---------------------

Maybe the error is in watcom, the readme.txt of watcom says:

Changes in 11.0 that may Require Recompilation:
-----------------------------------------------
clock() The clock function accuracy has changed from 100 ticks per
second to 1000 ticks per second (i.e., CLOCKS_PER_SEC has changed).
Source code that uses the clock function and CLOCKS_PER_SEC in its
calculations must be recompiled before linking the application with
new libraries.




--
Aku udah ngebales imel seukuran 1792, nih...

new topic     » goto parent     » topic index » view message » categorize

5. Re: Bug in eu2c

Mic writes:
>>Mic also reports a similar factor of ten error in the reported time.
>>I assume he used DOS4GW too.
>
>Nope. I'm using CauseWay.
>Wich leads us to the question: "does time() work the way it should?"

I'm using WATCOM 10.6 where the value of
CLOCKS_PER_SEC is 100.
I wonder if WATCOM 11.0 defines CLOCKS_PER_SEC
as 1000?
Can someone check for
#define CLOCKS_PER_SEC ...
in \watcom\h\time.h?
I get the time by dividing (double)clock() / CLOCKS_PER_SEC

If clock() on 11.0 is returning a number 10x bigger than I expect,
then that would explain the problem.

I guess I should #include <time.h> to allow for any differences.

Regards,
   Rob Craig
   Rapid Deployment Software
   http://www.RapidEuphoria.com

new topic     » goto parent     » topic index » view message » categorize

6. Re: Bug in eu2c

Aku writes:
> Changes in 11.0 that may Require Recompilation:
> -----------------------------------------------
> clock() The clock function accuracy has changed from 100 ticks per
> second to 1000 ticks per second (i.e., CLOCKS_PER_SEC
> has changed).

OK. Great. I just read your message.
I'll fix that and upload a new version for DOS32,
and also WIN32, within a day or so.

Regards,
   Rob Craig
   Rapid Deployment Software
   http://www.RapidEuphoria.com

new topic     » goto parent     » topic index » view message » categorize

Search



Quick Links

User menu

Not signed in.

Misc Menu