On Sun, 12 Oct 2008, Szak�ts Viktor wrote:
Hi Viktor,
> I'm referring to hb_WinToDosError(). Maybe it's
> wrong, but it's accepting an ULONG, when called,
> a value received from GetLastError() is passed
> to it (casted to USHORT, which instantly causes
> loss of significant digits), and hb_WinTo
Correction: values with bit 29 set are reserved
for app error codes.
On 2008.10.12., at 17:09, Szakáts Viktor wrote:
Hi Przemek,
I'm referring to hb_WinToDosError(). Maybe it's
wrong, but it's accepting an ULONG, when called,
a value received from GetLastError() is passed
to it (casted to USHO
Hi Przemek,
I'm referring to hb_WinToDosError(). Maybe it's
wrong, but it's accepting an ULONG, when called,
a value received from GetLastError() is passed
to it (casted to USHORT, which instantly causes
loss of significant digits), and hb_WinToDosError()
then returns an int, which is later store
-Messaggio Originale-
Da: "Phil Barnett" <[EMAIL PROTECTED]>
A:
Data invio: domenica 12 ottobre 2008 16.07
Oggetto: Re: [Harbour] Error code type
If you think about
void main()
where you can't return an error code compared to
int main()
where you can return an
On Sunday 12 October 2008 06:48:57 am Przemyslaw Czerpak wrote:
> Windows uses DOS error codes counted from 0 and significant
> are only 2 low bytes. AFAIR upper bytes are usd in some subsystems
> as bitfields.
Yes on significant 2 low bytes.
Not true on starting at 0. You can return negative er
On Sun, 12 Oct 2008, Szak�ts Viktor wrote:
Hi Viktor,
> OS error codes are currently stored as USHORT.
> Windows uses DWORD, there is a function which
> maps Windows error code to Harbour one.
To be precise: to DOS error. Anyhow such translation
cannot be well done in other systems. Windows uses
Hi all,
OS error codes are currently stored as USHORT.
Windows uses DWORD, there is a function which
maps Windows error code to Harbour one. Digits
are most probably lost here.
Shouldn't we change our error code types to ULONG?
Brgds,
Viktor
___
Har
