Qbasicnews.com

Hi:

In winbase.bi the structure DEVMODE has a bug in the size of one of its members.

'dmBitsPerPel As Short << GOK Wrong data type
dmBitsPerPel As Integer



Also a note about Missing Declarations FreeBASIC v0.13

The following declarations are missing from FreeBASIC v0.13. I needed to add them to get OpenGL code to work with FreeBASIC, so I think they are generally useful additions to FreeBASIC.

'' Declare missing GL functions
DECLARE FUNCTION wglMakeCurrent ALIAS "wglMakeCurrent" (BYVAL hdc AS LONG, BYVAL hglrc AS LONG) AS LONG
DECLARE FUNCTION wglDeleteContext ALIAS "wglDeleteContext" (BYVAL hglrc AS LONG) AS LONG
DECLARE FUNCTION wglCreateContext ALIAS "wglCreateContext" (BYVAL hdc AS LONG) AS LONG

'' Declare missing Windows functions
DECLARE SUB WaitMessage ALIAS "WaitMessage" () 'should be in user32.bi

Thanks

Garvan

0.13b? wtf? it's not even OUT yet...!

He's talking about the CVS I think. But I thought you couldn't compile it with .12 :o .

Hi:

I had another problem I reported in a different post on the 17th, and I download the updated version from the CVS on source forge and built it myself just to make sure the bug was not already fixed before making the report.

It is not easy to compile v0.13 with v0.12, but it is possible. You need to do it in 2 steps.

Garvan

Cool, so what would that be?

It seems to me that there is another bug with 0v13b with the Command$ function.

Dumbledore provided his compiled version of 0v13b here:
http://qh2.qbtk.com/337-d
With this file I was able to compile even a newer version.

FreeBasic is coming along nicely:
just dwnloaded the newest files from CVS and the changelog says:

• [added]
  - long integer (64-bit) data types (signed and unsigned), all operations done inline but for division and modulo (v1c)
  

and readme.txt says:

• - The LONGINT and ULONGINT data types are 64-bit wide.
  

:bounce:
Suppose the INT64 type in Winbase.bi is now obsolete.

Vic & Lillo keep up the good work, you guys rule!
:king: :king:

Alright, the field type was fixed and the new functions were added, changes are in the CVS (it may take up to 5 hours to sf.net update the anonymous access).

About the 64-bit integers, yeah i completed adding them today (literal longint's couldn't be used with the code you probably have, take care), a couple of thousands of lines of code were needed to get them to work inline. They are probably faster than LONG's were in QB and most libs i'm porting like MySQL, libxml and gsl need 64-bit ints to work, so..

I may add the currency type for fun, actually.. forget.. it's not fun.

The compiler's source is "impossible" to be built using 0.12 now, as it's using longint's internally..

huh? worked for me...
(shutting up... :wink: )

edit: and your example works too, so it's not some freaky bug thingy or something... (18446744073709551615 is a really big number :roll: )