CompuServe Messages

Hi Geoffrey, Why don't you use Modex? Of cause, drawing the pixels is slower, but you can use the multiple bitplane-writes to speed up the Clear-Page-Mechanism up to 4 times. And you can use tripple-buffering (paging) for flicker-free animation. I think it's worth the approach. By the way, I wouldn't use a bresenham-line-algorithm.. I used a smaller algorithm, which is much faster.. I'll post the source here.. ; ============================ Prozeduren =============================== ; DDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDD Proc BuildLineTable near ; ; Aufruf: p1=AX:BX p2=CX:DX ; calculates for all Y-values between Y1 and Y2 the corresponding x-vals ; values will be stored into the arrays minbuf and maxbuf ; DDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDD local x1:word, y1:word, y2:word, x2:word uses si, di cmp cx, ax ; Vergleiche X2 und X1 jz v_line ; Gleich? Vertikale Linie jg nichtdrehen ; das x1 links liegt xchg ax, cx xchg bx, dx nichtdrehen: cmp bx, dx jz h_line ; Horizontale Linie mov [x1], ax mov [y1], bx mov [x2], cx mov [y2], dx xchg ax, cx sub ax, cx xor dx, dx ; HI-Word auf 0 mov cx, [y2] ; Delta-Y berechnen sub cx, [y1] js reverseline ; Delta-Y negativ? ; ——————– Code fr die '\'-Linie ————————– mov di, cx ; DeltaY merken idiv cx mov [byte ptr cs:cm1+2], dl ; Div-Rest in den Code schreiben mov [byte ptr cs:ueberlauf], INC_BX mov bx, [x1] mov si, [y1] ; Offset-Zeiger in den Tabellen add si, si shr dx, 1 inc cx lineloop: StoreXVal bx ; Tabelle aktualisieren add bx, ax ; Steigung addieren cm1: add dx, BYTEDUMMY ; Div-Rest addieren cmp dx, di ; mit DELTAY vergleichen jae ueberlauf ; Rest -> Deltay? add si, 2 ; Zeiger erhhen loop lineloop ret ueberlauf: inc bx sub dx, di ; Rest wiederherstellen add si, 2 loop lineloop ret ; ———————- Code fr die '/'-linie ————————– reverseline: neg cx ; DeltaY negieren mov di, cx ; DeltaY merken idiv cx mov [byte ptr cs:cm1+2], dl ; Div-Rest in den Code schreiben mov [byte ptr cs:ueberlauf], DEC_BX mov bx, [x2] mov si, [y2] ; Offset-Zeiger in den Tabellen add si, si neg ax ; Quotient negieren inc cx ; Zhler erhhen shr dx, 1 jmp lineloop ; normale Line berechnen ; ———————— Horizontale Linie ——————————- h_line: mov si, bx add si, bx StoreXval ax ; X-Werte fr Xmin aktualisieren StoreXval cx ; X-Werte fr Xmax aktualisieren ret ; ———————– Vertikale Linie —————————- v_line: cmp bx, dx ; Vergleiche y2 mit y1 jl v_marke1 xchg bx, dx ; Tauschen, wenn bx>dx v_marke1: mov si, bx add si, bx ; Offset berechnen v_loop: StoreXval ax ; Tabelle aktualisieren add si, 2 inc bx cmp bx, dx jbe v_loop v_lineende: ret Endp I apologize for the bad format and the german comments… if you have problems (I only what to show you how it works) ask me… By the way, I use code-manipulation, you have to define the opcodes for inc_bx ect. Regards, Nils