Re: A83: Re: line drawing algorithm


[Prev][Next][Index][Thread]

Re: A83: Re: line drawing algorithm




Yikes...I guess I should have looked over Harper's routine a bit more
closely :)  Just for fun I timed the three routines with the 940 lines.
Here's the times (approximate of course, but should be close):
Harper's -- 9 seconds
Mine -- 4.5 seconds
TI's -- 12.5 seconds

I think part of the reason Harper's is slower than mine is that he actually
calculates the slope, which requires (relativly) slow division and more
complicated increments.

-Dan
----- Original Message -----
From: "Sam Heald" <evil_sam@hotmail.com>
To: <assembly-83@lists.ticalc.org>
Sent: Friday, August 25, 2000 12:26 PM
Subject: Re: A83: Re: line drawing algorithm


>
> My 2 cents:
>
>   Having studied the problem thoroughly (~2 minutes), I've concluded that:
> A)Dan failed to properly impliment Harper's routine into his source code.
> B)Harper failed to make his line drawing routine user-friendly, thus
helping
> to incur Dan's blunder.
>
> Dan's mistake:
> He incorrectly defined the variables in his program. He defined them as
> words, not bytes, which is not usually a problem except that Harper had a
> line of code like this "ld (slope+1),a" instead of this "ld (remainder),a"
> In Dan's definitions, "slope+1" does not equal "remainder"
>
> Harper's mistake:
> He should not make the user define all of those variables (5). It is much
> much smarter to use self-modifying code that doesn't require such
> definitions (or temporary space).
>
> Anyway, so Harper's routine works...
> The modified routine for self-mod code is here:
> void.calc.org/line.h
>
> Here's a working version of Dan's source code (converted to ION)
> void.calc.org/lines.zip
>
> QED
>
>
> >dan, try to run the program inside line.zip called freed.83p   its for
sos,
> >and it demonstrates that the routine works.  I don't know why it fails in
> >your program, since i cut and pasted the "loop2" part of your program
into
> >mine and it was successful in drawing the same triangle as iline and your
> >fastline
> >
> >-harper
> >
> >ps. this thread is getting kind of old.
> >
> >At 11:58 PM 8/24/00 -0400, you wrote:
> >
> >>I got the updated routine, but seem to still have the same problem...
> >>I've
> >>reuploaded lines.zip file (http://tcpa.calc.org/members/dan/lines.zip)
> >>with
> >>the updated source and binary for the program to test all the lines
> >>routines.  The screen shots remain the same:
> >>plotline/drawline (harper's routine) -
> >>http://tcpa.calc.org/members/dan/plotline.bmp
> >>fastline (my routine) - http://tcpa.calc.org/members/dan/fastline.bmp
> >>iline (built in routine) - http://tcpa.calc.org/members/dan/iline.bmp
> >>
> >>-Dan
> >>
> >>----- Original Message -----
> >>From: "Harper Maddox" <gte172i@prism.gatech.edu>
> >>To: <assembly-83@lists.ticalc.org>
> >>Sent: Thursday, August 24, 2000 10:50 PM
> >>Subject: A83: line drawing algorithm
> >>
> >>
> >> >
> >> > Ok, I figured out what the problem with the line routine was, and I
> >>fixed
> >>it.
> >> > Once again, it is available at http://void.calc.org/programming
> >> >
> >> > -harper
> >> >
> >> >
> >
> >
>
> ________________________________________________________________________
> Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com
>
>




Follow-Ups: References: