Nicron #97 16 июля 1998

Программирование

Programming - the withdrawal of sprites.

          Open letter
            for coders

m Editorial

Traditional note: the views of the writer and editors can
not coincide. ;-)

            *

(C) Spencer Winset


  Hello, readers!
I hope that you at least a little chop in the codes and 
certainly not I dream that you love to take strokes.


In fact, I encourage you to help in the codes, namely, it will
the banality of the withdrawal of so-called sprites. IN 
EX-USSR, of course, professionals in this field rife. Well, 
okay, after much speculation and fast "Krakow" (?) branded games

I have not seen the expected diversity of algorithms output, 
although closer to the October 1987 trend of avoiding Komad 
type: LDIR, LDDR, LDD and LDI, however, certain facts are 
manifestations of these teams still occur (after all, they are 
automatic). 


   In the nineties progress is evident: 7 LD A, (DE)
                    7 XOR (HL)
                    7 LD (DE), A
                    4 INC E
                    4 INC L
                 -------- TOTAL: 29 cycles per byte.


   And finally, on the threshold of the second
 Goals no longer afraid
 use a "stack": 10 POP DE
                     7 LD A, (HL)
                     4 XOR E
                     7 LD (HL), A
                     4 INC L
                     7 LD A, (HL)
                     4 XOR D
                     7 LD (HL), A
                     4 INC L
                  -------- TOTAL: 54 / 2 = 27 cycles per byte.


   As can be seen, the sprite without a mask and arbitrary size 
- as they say, the poet's dream. And in all seriousness, who 
predlozhet conceptually new and faster algorithm for removing 
the sprite, write, but I've had a better bottle of tasty cold 
beer! 


   The text is a consequence of the lack of more accessible 
sources of information about the current programming on the 
Spectrum assembly language. Many (especially beginners), there 
are such problems, some broken and abandoned, but most sit down 
hard. (In the Diamond group, REAL on "emmulyatore" sitting 
exactly the same and FIL / ANTARES, and yet a lot of people, 
and a very long time). 

Personally: Roman Petrov / Star goup / RDS / RPSG, if you are 
able, write that the project "_Scorched Earth_", which is much 
better than "Tank War". Coincidence and the fact long ago 
noted: Project Diamond group "_FOUR RASCALS_", is also better 
than "Tank War" all ethical and moral directions. Action - 
needed by all! 


   Once I have ventilated air to create a library of algorithms 
and the various procedures developed over the past century, but 
no enthusiasm has been noted. What is it - self-interest? "No, 
I think "green" in our sandbox almost did not come their 
attraction for light multimedia monitors, leaving only 
Dinosaurs coding, domestic bison demostroeniya, Hints which are 
not needed. Do not exist encoders middle class: or You do not 
know anything, or do you know (know how) all, in both cases is 
usually not do anything you want. And the truth - and have fun, 
and the library has helped, would have all seven novice RST, 
faster to do the first steps to new projects that will not let 
spektrumistam 90 bored in retirement vacation.


        P.S. And maybe that coder was "thinking"
             he must himself to reinvent the wheel.

                    Spencer Winset / Diamond group
                    ZXNet: 500:95 / 550.8

            *