[fpc-pascal]Amiga snapshot (was heap and stack size )

[Carl Eric Codere]

> Could you make lha archive of your current fpc amiga 
> version and place it > somewhe in net as I didn't 
> succeed to crosscompile my own.

Greetings,
 I've finally done the script for creating
snapshots for the amiga target. I'll upload a 
an amiga snapshot this monday. Since this is not
a full release, you'll have to configure the compiler
yourself, as well as install the GNU as/ld/ixemul lib yourself
to make it work. Please let me know if you have problems
with it 

Thanks!



___
fpc-pascal maillist  -  [EMAIL PROTECTED]
http://lists.freepascal.org/mailman/listinfo/fpc-pascal

[fpc-pascal]Test (please ignore)

[Aitor Santamaría Merino]

Apologies for this little test (I got no mails from the 14th)

___
fpc-pascal maillist  -  [EMAIL PROTECTED]
http://lists.freepascal.org/mailman/listinfo/fpc-pascal

[fpc-pascal][Fwd: Re: [wdosx] Future of WDOSX]

[Aitor Santamaría]

Hi,

In the list of discussion about the WDOSX DOS extender, this interesting 
question has been raised. It is about legal issues with the LGPL license 
and the linkage to propietary code. As FreePascal has been directly 
mentioned, and I am subscribed to both lists, and quite interested in 
the topic, I thought it would worth to post this item here, to see what 
people has to say about it. The next forward contains quite a good 
summary of what has been happening after the announcement that WDOSX 
developers may be thinking of distributing WDOSX under the LGPL license.

Aitor
=

I noticed that if you go to the FreePascal project at http://www.freepascal.org/ and go down 
to the Licence section at the bottom of the page it says "The packages and runtime library 
come under a modified Library GNU Public License (LGPL) to allow  the use of static 
libraries when creating applications."

This may be worth investigating because it seems that the static linking of the WDOSX 
code described below is the primary sticking point that has been raised in this topic.

Cheers

Garry Wood.

"Alain M." wrote:

> 2) distribute my proprietary sw in .OBJ form and have all
> LGPL sw with source included so that they can be rebuilt and
> liked by the user (at his own risc). This cannot be as some libs are
> forbidden to be distributed if not linked with an executable.
> 
> If WDOSX could be distributed embeded in my binaries,
> I could make a .EXE that has all my proprietary code plus
> the restricted libraries and place all LGLP sw in DLLs
> and then I could be ok with every licence.
> 
> I hope this explains my problem (and other's I believe).
> I am bringing this discussion because I have tested WDOSX
> with good results and I like it.

Alain,

I think you have a serious point here. My guess is that you are
referring to LPGL Section 6.

The idea behind this section is to enable the user to completely
rebuild the executable using a modified version of the "library".

For anything that does not make use of the Win32 emulation this
can be achieved by simply running stubit.exe on the executable again.
As far as WDLs, however, there currently is no mechanism to replace
them with updated modules once they are part of the executable.
I can see a possible solution for this. However, it requires some
work: Stubit.exe needs to be modified such that it replaces any
WFSE module in the executable where it finds a matching new .WDL
for it and leave all the others alone. Guess that would (be) work...

Guess LGPL wasn't especially written with DOS extenders in mind :-/

I'm going to throw this subject into discussion as I might have
missed something very obvious and easy which might make this issue
a non non- issue...

Best regards,


Michael Tippach







___
fpc-pascal maillist  -  [EMAIL PROTECTED]
http://lists.freepascal.org/mailman/listinfo/fpc-pascal

[fpc-pascal]1. Mandrake & 2. FPC with WINE and crosscompiling

[Holger Peterrs]

Hi,

1. At home I'm working with FreePascal for Windows. But in extra curricular
lessons, I work with Linux (Mandrake). I tried to install FreePascal for
Linux on this comuter, but it didn't work. For example there is no IDE and if I
start fpc in the console, there is no output. Is it because I am not logged in
as a root user? How can I solve this?

2. If there is no solution, can I run the windows version of FreePascal with
WINE and compile the programs for Linux (crosscompiling)? If yes, how can I
do this? 

Thanks for your awnsers. 

Holger 

-- 
+++ GMX - Mail, Messaging & more  http://www.gmx.net +++
NEU: Mit GMX ins Internet. Rund um die Uhr für 1 ct/ Min. surfen!

___
fpc-pascal maillist  -  [EMAIL PROTECTED]
http://lists.freepascal.org/mailman/listinfo/fpc-pascal

[fpc-pascal]graphiX & NVIDIA

[Peter Beisel]

compiler: 1.0.6
target: go32v2
OS: a) DOS 6.22; b) Win98SE

I've got serious problems running graphix with a NVIDIA GeForce4 440 Go
64MB graphics adapter on my notebook.
Has anybody some hints??

Regards Peter

___
fpc-pascal maillist  -  [EMAIL PROTECTED]
http://lists.freepascal.org/mailman/listinfo/fpc-pascal

[fpc-pascal]TurboPower

[Andreas K. Foerster]

Announcement from :



TurboPower discontinues retail business

Company announces sweeping open source initiative

Colorado Springs - TurboPower Software Company today announces their
immediate withdrawal from the retail component and developer tools
market. As part of the move, TurboPower announces its intention to
release their award winning component libraries as open source to the
maximum extent possible.

"We've been a big part of the developer community for nearly 18 years,"
said TurboPower President Gary Frerking. "Open sourcing our products
seems like an appropriate gesture of appreciation to the community that
supported us so well over the years. We sincerely hope these products
take on a life of their own and continue to prosper for many years to
come."

TurboPower's component libraries cover a wide spectrum of development
needs including compression, serial communication, faxing, Internet
communication, scheduling, data entry, encryption, and XML
manipulation. There are nearly twenty commercial-quality libraries
being considered for release, containing hundreds of components and
thousands of classes and routines - over one million lines of source
code in all!

The process of preparing the libraries for release to open source is
already underway, and the current goal is to have the libraries
released to the maximum extent possible by the end of January. The
resulting open source projects will be hosted on SourceForge.




-- 
Tschuess
Andreas
___
fpc-pascal maillist  -  [EMAIL PROTECTED]
http://lists.freepascal.org/mailman/listinfo/fpc-pascal

[fpc-pascal]graphiX & NVIDIA

[Peter Beisel]

compiler: 1.0.6
target: go32v2
OS: a) DOS 6.22; b) Win98SE

I've got serious problems running graphix with a NVIDIA GeForce4 440 Go
64MB graphics adapter on my notebook.
Has anybody some hints??

Regards Peter

___
fpc-pascal maillist  -  [EMAIL PROTECTED]
http://lists.freepascal.org/mailman/listinfo/fpc-pascal

[fpc-pascal]Units for Windows

[Hans Mårtensson]

According to the FreePascal Documentation the Windows unit includes the API
calls for the Windows OS as doucumented in Borland's Win32 help file.

But not all of them.

I discovered another unit, named mmsystem, that took care of the functions
to be used in connection with the sound card. With this unit I was able to
make a sound recording program.

Now I would like to make a program to manipulate printers. For this purpose
I need to call functions like OpenPrinter, EnumJobs and SetJob, specified
in the Win32 help file. But they are not recognized by the Windows unit.

Does there exist a unit for this purpose?
Or another way to use these functions?

Hans Maartensson

___
fpc-pascal maillist  -  [EMAIL PROTECTED]
http://lists.freepascal.org/mailman/listinfo/fpc-pascal

[fpc-pascal]Dynamic Arrays (Pointers)

[James Mills]

Hi,

I'm using FPC 1.0.4

Would someone be so kind as to show me an example of using dynamic
arrays of an object, creating new objects on the fly in code and
disposing of them when they're no longer needed. Use of this would be in
a irc services program I'm porting to linux from delphi (windows), ie:
new nick joins server, create new object respresenting that nick...

Thank you
James
___
fpc-pascal maillist  -  [EMAIL PROTECTED]
http://lists.freepascal.org/mailman/listinfo/fpc-pascal

[fpc-pascal]1. FPC and Mandrake 2. Crosscompiling

[Holger Peters]

Hi,
1. At home I'm working with FreePascal for Windows. But in extra 
curricular lessons, I work with Linux (Mandrake). I tried to install 
FreePascal for Linux on this comuter, but it didn't work. For example 
there is no IDE and if I start fpc in the console, there is no output.
Is it because I am not logged in as a root user? How can I solve this?

2. If there is no solution, can I run the windows version of FreePascal 
with WINE and compile the programs for Linux (crosscompiling)? If yes, 
how can I do this?

Thanks for your awnsers.

Holger


___
fpc-pascal maillist  -  [EMAIL PROTECTED]
http://lists.freepascal.org/mailman/listinfo/fpc-pascal

[fpc-pascal]64bit constant

[Stefan Ziegenbalg]

Hello,

how can I define a 64 bit constant, e.g.

const csr_register_base = $f000;


This would produce:

csr.pas(5,35) Error: Invalid integer expression



Regards Stefan


--
mailto:[EMAIL PROTECTED]
http://www.sziegenbalg.de
http://www.simage.org

___
fpc-pascal maillist  -  [EMAIL PROTECTED]
http://lists.freepascal.org/mailman/listinfo/fpc-pascal

Re: [fpc-pascal]64bit constant

[Jonas Maebe]

On vrijdag, jan 24, 2003, at 12:23 Europe/Brussels, Stefan Ziegenbalg 
wrote:

how can I define a 64 bit constant, e.g.

const csr_register_base = $f000;


You can only do that in version 1.1 and later (see development section 
of the website for beta snapshots).


Jonas

___
fpc-pascal maillist  -  [EMAIL PROTECTED]
http://lists.freepascal.org/mailman/listinfo/fpc-pascal

Re: [fpc-pascal]TurboPower

[memsom]

> Announcement from :
> 
> 
> 
> TurboPower discontinues retail business
> ..
> The process of preparing the libraries for release to open source is
> already underway, and the current goal is to have the libraries
> released to the maximum extent possible by the end of January. The
> resulting open source projects will be hosted on SourceForge.

Orpheus is already available from SF. I got it at the end of last 
week/beginning of this week. Looks nice, but I haven't really had a chance to 
do much with it.


Matt

-
This message was sent using Mistral WebMail.
http://www.mistral.co.uk/


___
fpc-pascal maillist  -  [EMAIL PROTECTED]
http://lists.freepascal.org/mailman/listinfo/fpc-pascal

[fpc-pascal]SetUid

[Balázs Csaba]

Hello masters!

I am new on maillist. I have got a question. A cannot find function called
setuid() like in c++. In the source of compiler/units a have found a section
where the linux like function declared. There is a contant called
"syscall_nr_setuid", but it is NOT used in any function. A can declare a
setuid function but i do not know which registers are user int this
function.
How can I switch to a low-level user (not root) after a fork()
Can somebody help me?

Regards!
Balázs Csaba

___
fpc-pascal maillist  -  [EMAIL PROTECTED]
http://lists.freepascal.org/mailman/listinfo/fpc-pascal

[fpc-pascal]FPDoc reference manual?

[Preben Mikael Bohn]

Hi all, I have just tried to compile FPDoc, and it went fine except that I had
to change the makefile to add a few more include-dirs than were in it as
default. Anyway, I am now looking at the reference manual located at
http://www.freepascal.org/cgi-bin/viewcvs.cgi/*checkout*/projects/fpdoc/doc/fpdoc.tex?rev=1.2&content-type=text/plain
however I am not able to compile it with latex due to missing preambles, etc.
Does anyone have a ps, pdf, etc -version available, or could tell me how to
compile it?

Best regards Preben

-- 
Don't take life too seriously. You'll never get out of it alive.


___
fpc-pascal maillist  -  [EMAIL PROTECTED]
http://lists.freepascal.org/mailman/listinfo/fpc-pascal

Re: [fpc-pascal]Dynamic Arrays (Pointers)

[memsom]

> Would someone be so kind as to show me an example of using dynamic
> arrays of an object, creating new objects on the fly in code and
> disposing of them when they're no longer needed. Use of this would be in
> a irc services program I'm porting to linux from delphi (windows), ie:
> new nick joins server, create new object respresenting that nick...

To be honest, I would advise using a TList. This will do all the donley work 
for you, and even has a nice method 'pack' that will remove dead pointers for 
you. It doesn't sound like you need uber speed or anything, so even though the 
TList does add a slight performance hit over a dynamic array, the pro's 
outweigh the cons. All you would do is something like this, (TMYCLASSTYPE would 
be replaced with yout TObject descendent):

TWrapList = class
private
  FList: TList;
protected
  procedure SetListValue(Value: TMYCLASSTYPE); virtual;
  function  GetListValue(const Index: integer): TMYCLASSTYPE; virtual;
public
  property ListValues[index: integer]: TMYCLASSTYPE 
   read GetListValue write SetListValue;

  constructor Create; virtual;
  destructor Destroy; override;

  function Count: integer;
  procedure Add(Item: TMYCLASSTYPE);
  procedure Delete(Index: Integer);
end;


constructor TWrapList.Create;
begin
  FList := TList.Create;
end;
 
destructor TWrapList.Destroy;
begin
  // NB: remember remove any items left in list
  
  FList.free;
  inherited;
end;
 
procedure TWrapList.SetListValue(Value: TMYCLASSTYPE);
begin
  // NB. range checking will be usefull
  FList[index] := Value;
end;
 
function  TWrapList.GetListValue(const Index: integer): TMYCLASSTYPE; 
begin
  Result := TMYCLASSTYPE( FList[index] );
end;

function TWrapList.Count: integer;
begin
  Result := FList.Count;
end;

procedure TWrapList.Add(Item: TMYCLASSTYPE);
begin
  FList.Add(Item); //if you make this a finction you can
   //return the index here
end;

procedure TWrapList.Delete(Index: Integer);
begin
  FList.Delete(Index);
end;

This is a pattern I use all the time. It works well.

Matt
 


-
This message was sent using Mistral WebMail.
http://www.mistral.co.uk/


___
fpc-pascal maillist  -  [EMAIL PROTECTED]
http://lists.freepascal.org/mailman/listinfo/fpc-pascal

Re: [fpc-pascal]FPDoc reference manual?

[Michael Van Canneyt]

On Fri, 24 Jan 2003, Preben Mikael Bohn wrote:

> Hi all, I have just tried to compile FPDoc, and it went fine except that I had
> to change the makefile to add a few more include-dirs than were in it as
> default. Anyway, I am now looking at the reference manual located at
> 
>http://www.freepascal.org/cgi-bin/viewcvs.cgi/*checkout*/projects/fpdoc/doc/fpdoc.tex?rev=1.2&content-type=text/plain
> however I am not able to compile it with latex due to missing preambles, etc.
> Does anyone have a ps, pdf, etc -version available, or could tell me how to
> compile it?

The needed packages are in the docs/packages repository in CVS.

But I can send a .pdf version if you want.

Michael.

___
fpc-pascal maillist  -  [EMAIL PROTECTED]
http://lists.freepascal.org/mailman/listinfo/fpc-pascal

Re: [fpc-pascal]SetUid; PARANOIA

[Anton Tichawa]

hello, csaba!

here you are:


// ...
uses
  Linux;
// ...
function SetUID(UID: Longint): Longint;
var
  Regs: SysCallRegs;
begin
  Regs.Reg2 := UID;
  SetUID := SysCall(SysCall_nr_getuid, Regs);
end;
//...


written by me some months ago.

a question to the fpc development team: i'm slowly getting paranoid, because 
access to hw, i/o and system seems to vanish as the years pass. what is the 
exact reason that setuid has not been implemented or wrapped in fpc?

anton.


On Friday 24 January 2003 14:36, Balázs Csaba wrote:
> Hello masters!
>
> I am new on maillist. I have got a question. A cannot find function called
> setuid() like in c++. In the source of compiler/units a have found a
> section where the linux like function declared. There is a contant called
> "syscall_nr_setuid", but it is NOT used in any function. A can declare a
> setuid function but i do not know which registers are user int this
> function.
> How can I switch to a low-level user (not root) after a fork()
> Can somebody help me?
>
> Regards!
> Balázs Csaba
>
> ___
> fpc-pascal maillist  -  [EMAIL PROTECTED]
> http://lists.freepascal.org/mailman/listinfo/fpc-pascal


--

"Adas Methode war, wie sich zeigen wird, Tagträume in offenbar korrekte 
Berechnungen einzuweben."

Doris Langley Moore: Ada, Countess of Lovelace (London 1977).

--

Anton Tichawa
Volkertstrasse 19 / 20
A-1020 Wien
phone: +43 1 218 97 59
mobil: +43 664 52 07 907
email: [EMAIL PROTECTED]

--
___
fpc-pascal maillist  -  [EMAIL PROTECTED]
http://lists.freepascal.org/mailman/listinfo/fpc-pascal

Re: [fpc-pascal]SetUid

[Marco van de Voort]

> 
> I am new on maillist. I have got a question. A cannot find function called
> setuid() like in c++.

> In the source of compiler/units a have found a section
> where the linux like function declared. There is a contant called
> "syscall_nr_setuid", but it is NOT used in any function. A can declare a
> setuid function but i do not know which registers are user int this
> function.

First argument in reg2 and so on. (for linux), e.g.


type uid_t=longint;

Function Sys_SetUid(uid:uid_t);
var
  regs : SysCallregs;
begin
  regs.reg2:=uid;
  Sys_setuid:=SysCall(SysCall_nr_setuid,regs);
end;


___
fpc-pascal maillist  -  [EMAIL PROTECTED]
http://lists.freepascal.org/mailman/listinfo/fpc-pascal

Re: [fpc-pascal]SetUid

[Marco van de Voort]

> I am new on maillist. I have got a question. A cannot find function called
> setuid() like in c++. In the source of compiler/units a have found a section
> where the linux like function declared. There is a contant called
> "syscall_nr_setuid", but it is NOT used in any function. A can declare a
> setuid function but i do not know which registers are user int this
> function.
> How can I switch to a low-level user (not root) after a fork()
> Can somebody help me?

Note that that last one is untested, since I don't have Linux myself, and
don't have root rights on the remote system I do some linux testing now and then.




___
fpc-pascal maillist  -  [EMAIL PROTECTED]
http://lists.freepascal.org/mailman/listinfo/fpc-pascal

Re: [fpc-pascal]SetUid; PARANOIA

[Michael Van Canneyt]

On Fri, 24 Jan 2003, Anton Tichawa wrote:

> hello, csaba!
>
> here you are:
>
> 
> // ...
> uses
>   Linux;
> // ...
> function SetUID(UID: Longint): Longint;
> var
>   Regs: SysCallRegs;
> begin
>   Regs.Reg2 := UID;
>   SetUID := SysCall(SysCall_nr_getuid, Regs);
> end;
> //...
> 
>
> written by me some months ago.
>
> a question to the fpc development team: i'm slowly getting paranoid, because
> access to hw, i/o and system seems to vanish as the years pass. what is the
> exact reason that setuid has not been implemented or wrapped in fpc?

No particular reason. It wasn't needed at the time of the linux port
It is (or should be) implemented in the new 1.1 series.

Michael.

___
fpc-pascal maillist  -  [EMAIL PROTECTED]
http://lists.freepascal.org/mailman/listinfo/fpc-pascal

Re: [fpc-pascal]SetUid; PARANOIA

[Marco van de Voort]

> here you are:
> 
> 
> // ...
> uses
>   Linux;
> // ...
> function SetUID(UID: Longint): Longint;
> var
>   Regs: SysCallRegs;
> begin
>   Regs.Reg2 := UID;
>   SetUID := SysCall(SysCall_nr_getuid, Regs);
> end;
> //...
> 
> 
> written by me some months ago.
> 
> a question to the fpc development team: i'm slowly getting paranoid, because 
> access to hw, i/o and system seems to vanish as the years pass. what is the 
> exact reason that setuid has not been implemented or wrapped in fpc?

None. There pretty much was an initial FPC port to Linux to get everything working,
and after that functions were added on demand only. Specially because it changes 
often, and
there is/was no decent way to autogenerate them.

I'm working on improving this in 1.1, and also separate the more portable
from the more unportable functions there.
___
fpc-pascal maillist  -  [EMAIL PROTECTED]
http://lists.freepascal.org/mailman/listinfo/fpc-pascal

Re: [fpc-pascal]SetUid; PARANOIA

[Marco van de Voort]

> On Fri, 24 Jan 2003, Anton Tichawa wrote:
> 
> >   Regs.Reg2 := UID;
> >   SetUID := SysCall(SysCall_nr_getuid, Regs);
> > end;
> > //...
> >
> > written by me some months ago.
> >
> > a question to the fpc development team: i'm slowly getting paranoid, because
> > access to hw, i/o and system seems to vanish as the years pass. what is the
> > exact reason that setuid has not been implemented or wrapped in fpc?
> 
> No particular reason. It wasn't needed at the time of the linux port
> It is (or should be) implemented in the new 1.1 series.

It is, unit baseunix (which also works for *BSD).

___
fpc-pascal maillist  -  [EMAIL PROTECTED]
http://lists.freepascal.org/mailman/listinfo/fpc-pascal

Re: [fpc-pascal]SetUid; PARANOIA

[Anton Tichawa]

hello, list!

On Friday 24 January 2003 15:22, Michael Van Canneyt wrote:

>
> No particular reason. It wasn't needed at the time of the linux port
> It is (or should be) implemented in the new 1.1 series.
>

thanks .. this makes me feel better. but isn't it strange that google finds 
not a single document containing syscall_nr_setuid? sorry if off-topic.

my example of setuid contains a typing error: the argument 
'syscall_nr_getuid' should read 'syscall_nr_setuid'.

anton.


--

"Adas Methode war, wie sich zeigen wird, Tagträume in offenbar korrekte 
Berechnungen einzuweben."

Doris Langley Moore: Ada, Countess of Lovelace (London 1977).

--

Anton Tichawa
Volkertstrasse 19 / 20
A-1020 Wien
mobil: +43 664 52 07 907
email: [EMAIL PROTECTED]

--
___
fpc-pascal maillist  -  [EMAIL PROTECTED]
http://lists.freepascal.org/mailman/listinfo/fpc-pascal

Re: [fpc-pascal]FPDoc reference manual?

[Preben Mikael Bohn]

Hi Michael

> The needed packages are in the docs/packages repository in CVS.
> But I can send a .pdf version if you want.

That would be great! :-)

Best regards Preben

-- 
Don't take life too seriously. You'll never get out of it alive.


___
fpc-pascal maillist  -  [EMAIL PROTECTED]
http://lists.freepascal.org/mailman/listinfo/fpc-pascal

[fpc-pascal]Execution speed

[Mark Emerson]

Now that the list seems to be working again...I'll try posting this again:

I am wondering why our new PC is not executing our fpc-compiled program
very much faster than the old one. It was really quite a disappointment:

Old PC: Laptop, Intel PII, 300 MHz, 64 MB. Execution times: 8:30, 2:30 (min:sec)

New PC: Desktop, AMD Duron, 1.6 GHz, 128 MB. Execution times: 5:15, 1:15

The new PC ought to be 5 times faster (1600 MHz / 300 MHz, right? Of
course the speed of the memory is also a factor) but it's not even twice
as fast.

The execution time pairs are determined from three time stamps that
occur during one run of the program. The sequence is as follows:

* Stamp 1
-Initialize (5-10 secs reading/processing from HD)
-Process 1 (5-9 mins)
* Stamp 2
-Process 2 (1-3 mins)
* Stamp 3

Both machines are running Win98 Second Edition (could Windows 98 be
preventing the faster machine from running at full capacity? Or perhaps
it's because fpc runs in a DOS window, and the DOS mode is forcing it to
run slow?)

The program is very processor intensive. Only about 4MB of memory space
is used.

During runtime, we are doing less than 400 kb of read/write combined to
the HD. We put about 10 lines of text on the DOS screen to show
progress. So I can't imagine the I/O could be slowing us down.

I tried compiling with the two different target platforms, but it didn't
make a difference. Stackchecking is on, but it was on on both computers.

I also tried a few different bios settings (the computer has ready-made
bios configurations for "Optimal" and "Best Performance" (?) as well as
the factory default I started with.) But the compile times were the same
regardless of the bios settings.

Any thoughts?

Mark
___
fpc-pascal maillist  -  [EMAIL PROTECTED]
http://lists.freepascal.org/mailman/listinfo/fpc-pascal

RE: [fpc-pascal]Execution speed

[Lee, John]

Goes w/o saying that program perf depends on many things, but typically if
one runs a processor intensive (no i/o) benchmark eg an fft (fp) or a large
integer calculation on processors of different speeds the time is about
proportional to the proc speed and possibly the memory speed (a bit) if the
program/data in the loop is larger than the cache size for a given
processor. The AMD & pII are somewhat different in the way they work
internally but the factor would still be around the ration of the clock
speeds. I've done this with fpc code on PCs from 100M- 1G+ & it works, it
doesn't depend to any significant extent on the o/s.  

It looks as if there is more tinme spend in disk io than you think (assuming
that there are no sleeps or other time waits within the program), and that
will clearly limit the speed improvement since the disks are probably not n
times faster on the later PC, probably only 1-2 times?. Remember that any
call to an i/o procedure eg as read/write statement may result in one or
more o/s calls and disk i/o as well as many millions of lines of code & this
could take many many milliseconds eg a disk seek is usually 2msec, a read
write can be 10s of msec, compared with the the time taken by a x:=x+23; or
y:=cos(x); or whatever type lines which may only take a few 100 nsecs or
cedrtainly < a microsec on a 1.6G processor   
 
My suggestion is that you do some more timing of the code. I use clock, &
put lots of timing statemenmts in (with some sort of test flag or compiler
conditional) which gives microsec accuracy from Tomas Schatzl's cpu unit,
for checking this sort of thing, or you can probably use profil? to show
where program spends most of it's time.

NB it 100% certainly isn't BIOS or win se! 

HTH 

Regards John

-Original Message-
From: Mark Emerson [mailto:[EMAIL PROTECTED]]
Sent: Friday, January 24, 2003 15:39
To: fpc-post
Subject: [fpc-pascal]Execution speed


Now that the list seems to be working again...I'll try posting this again:

I am wondering why our new PC is not executing our fpc-compiled program
very much faster than the old one. It was really quite a disappointment:

Old PC: Laptop, Intel PII, 300 MHz, 64 MB. Execution times: 8:30, 2:30
(min:sec)

New PC: Desktop, AMD Duron, 1.6 GHz, 128 MB. Execution times: 5:15, 1:15

The new PC ought to be 5 times faster (1600 MHz / 300 MHz, right? Of
course the speed of the memory is also a factor) but it's not even twice
as fast.

The execution time pairs are determined from three time stamps that
occur during one run of the program. The sequence is as follows:

* Stamp 1
-Initialize (5-10 secs reading/processing from HD)
-Process 1 (5-9 mins)
* Stamp 2
-Process 2 (1-3 mins)
* Stamp 3

Both machines are running Win98 Second Edition (could Windows 98 be
preventing the faster machine from running at full capacity? Or perhaps
it's because fpc runs in a DOS window, and the DOS mode is forcing it to
run slow?)

The program is very processor intensive. Only about 4MB of memory space
is used.

During runtime, we are doing less than 400 kb of read/write combined to
the HD. We put about 10 lines of text on the DOS screen to show
progress. So I can't imagine the I/O could be slowing us down.

I tried compiling with the two different target platforms, but it didn't
make a difference. Stackchecking is on, but it was on on both computers.

I also tried a few different bios settings (the computer has ready-made
bios configurations for "Optimal" and "Best Performance" (?) as well as
the factory default I started with.) But the compile times were the same
regardless of the bios settings.

Any thoughts?

Mark
___
fpc-pascal maillist  -  [EMAIL PROTECTED]
http://lists.freepascal.org/mailman/listinfo/fpc-pascal

This e-mail and any attachment is for authorised use by the intended recipient(s) 
only.  It may contain proprietary material, confidential information and/or be subject 
to legal privilege.  It should not be copied, disclosed to, retained or used by, any 
other party.  If you are not an intended recipient then please promptly delete this 
e-mail and any attachment and all copies and inform the sender.  Thank you.
___
fpc-pascal maillist  -  [EMAIL PROTECTED]
http://lists.freepascal.org/mailman/listinfo/fpc-pascal

[fpc-pascal]FPC 1.0.6 and open arrays

[Jilani Khaldi]

Hi All,

Does FPC support open arrays?
I get an error compiling something like this:

type
  m = array of integer;

but it works fine if I declare it inside a procedure or a function:

procedure OpenArray(var m: array of integer);

Thanks!
--
Jilani Khaldi
http://www.jilani.net




___
fpc-pascal maillist  -  [EMAIL PROTECTED]
http://lists.freepascal.org/mailman/listinfo/fpc-pascal

[fpc-pascal]Run Time Error codes

[Eduardo Morras]

Hello:

Does anyone knows what means the next error codes: 998, 313. And 
where find information about the applications exit code?

I'm under win32 platform, fpc 1.0.6 using gdb as debugger

TIA


Alien.org contacted...waiting for Ripley

___
fpc-pascal maillist  -  [EMAIL PROTECTED]
http://lists.freepascal.org/mailman/listinfo/fpc-pascal