FiveTech Software tech support forums

... test it and report any problem. Thanks. Andi //------------------------ Andi YES .. In build 9444 it appears that all your changes work GREAT .. Math between DateTime variables work as well as writing the value of DateTime back to SQL tables work as well. Many Thanks Rick Lipkin

... would hope for the sake of 'clipper' compatibility the xHarbour folks can make their new T valtype work. So far I have tested two commits and the math and comparison operators now work. Unfortunately, when the value writes back to the Ado datetime field the table stores 12:00 am only :cry: I think ...

... .. xHarbour "NOW" considers this expression to be valtype T... prior to valtype D. The problem occurs where you do some math or comparison between dDate as in : nDiff := Date() - dDate or If Date() < dDate .... As you can see the datatype mismatch can be a potential ...

... = (unsigned char ) 0x80 >> loc; if ( hb_parl( 3 ) ) *ptr = *ptr | mask; else *ptr = *ptr & ~mask; } } hb_retl( res ); } #include "math.h" #include "stddef.h" #include <stdlib.h> #include <stdio.h> #ifndef HUGE_VAL # define HUGE_VAL HUGE #endif /*HUGE_VAL*/ ...

... + KEY_ENUMERATE_SUB_KEYS +  KEY_NOTIFY +  KEY_CREATE_LINK) & (~SYNCHRONIZE)) For example KEY_ALL_ACCESS is 0xF003F, which is 983103, not 63 : Math calc. explained at http://www.the-mindseye.co.uk/DllBinder/Registry_access_comments.shtml Hi, Davide

Thanks for the reply Rao.

I think age is catching up to me

Resolved the issue easily by just temporarily increasing SET DECIMALS value

Internal value has full accuracy. What is displayed may be rounded off to 2 decimal places.

msginfo( 2957.45 / 9 ) shows 328.605556

We can use the result for further processing without losing accuracy.

Using windows calculator, 2957.45 / 9 = 328.60555555555555555555555555556. In my program I got 328.61 not the expected result of 328.605555...

What am I missing here? Using xHarbour+BCC55

TIA

p.s. Just tested in xbscript.exe by simple means of "? 2957.45/9". Got 328.61 here too

... reference to `fmod' ./lib/libhbvm.a(hvm.o): In function `hb_vmPower': hvm.c:(.text+0x560d): undefined reference to `pow' ./lib/libhbrtl.a(math.o): In function `HB_FUN_EXP': math.c:(.text+0x7dd): undefined reference to `exp' ./lib/libhbrtl.a(math.o): In function `HB_FUN_LOG': math.c:(.text+0x912): ...

... http://www.scrabble.org.au/words/fours.htm http://www.scrabble.org.au/words/5jxqz.htm http://www.math.toronto.edu/jjchew/scrabble/lists/common-6.html http://www.math.toronto.edu/jjchew/scrabble/lists/common-8.html ...

... source\rtl\isprint.c: source\rtl\itemseri.c: source\rtl\langapi.c: source\rtl\left.c: source\rtl\len.c: source\rtl\lennum.c: source\rtl\math.c: source\rtl\maxrow.c: source\rtl\memofile.c: source\rtl\minmax.c: source\rtl\mlcfunc.c: source\rtl\mod.c: source\rtl\mouse53.c: source\rtl\mouseapi.c: ...

... source\rtl\isprint.c: source\rtl\itemseri.c: source\rtl\langapi.c: source\rtl\left.c: source\rtl\len.c: source\rtl\lennum.c: source\rtl\math.c: source\rtl\maxrow.c: source\rtl\memofile.c: source\rtl\minmax.c: source\rtl\mlcfunc.c: source\rtl\mod.c: source\rtl\mouse53.c: source\rtl\mouseapi.c: ...

Ésta es la parte de mi prg donde me sale el error.... ---------------------- #pragma BEGINDUMP #include "windows.h" #include "math.h" #include "hbapi.h" HB_FUNC( TANG ) //Aggiunta funzione aTan { hb_retnd((double) tan((double)hb_parnd(1) )); } HB_FUNC( ATN ) { hb_retnd((double) ...

Ciao Enrico. Ho risolto il problema aTan() in serendo queste righe: #pragma BEGINDUMP #include "windows.h" #include "math.h" #include "hbapi.h" HB_FUNC( ATN ) { hb_retnd((double) atan((double)hb_parnd(1))); } #pragma ENDDUMP Mi rimane il problema della finestra ...

... using negative numbers for points (and we must use points instead of pitch for equal-spaced fonts like Courier). In this case, the "-" sign isn't mathematical -- it doesn't mean (as math rules would suggest) that your point size gets smaller as the integer gets bigger. Just the reverse. The bigger ...