From frinkofox at yahoo.com  Thu Nov  3 13:38:09 2005
From: frinkofox at yahoo.com (arry)
Date: Thu, 3 Nov 2005 05:38:09 -0800 (PST)
Subject: [cl-gd-devel] Antialiased lines, arcs ... don't work,
	even the test orig pic is wrong... 
Message-ID: <20051103133809.42242.qmail@web54115.mail.yahoo.com>

Hi

I am using CL-GD for over a week now and found out 
that the antialiasing doesn't work at all...
The test, I think was number 4, produces the same
result as its orig picture in the orig folder, 
but the orig picture in the orig folder is also
wrong - there's no antialiasing in it...
And this orig pic comes from the package itself...

The picture on the web though is correct
here http://www.weitz.de/cl-gd/ the pic's name: is anti-aliased-lines.png

I tested this with AlegroCL and CLISP both under windows... with the same results...
Is this a bug?

Thanks for the help
arry




	
		
__________________________________ 
Yahoo! Mail - PC Magazine Editors' Choice 2005 
http://mail.yahoo.com


From edi at agharta.de  Fri Nov  4 01:28:58 2005
From: edi at agharta.de (Edi Weitz)
Date: Fri, 04 Nov 2005 02:28:58 +0100
Subject: [cl-gd-devel] Antialiased lines, arcs ... don't work, even the
	test orig pic is wrong...
In-Reply-To: <uy845dy5n.fsf@agharta.de> (Edi Weitz's message of "Fri, 04 Nov
	2005 02:01:08 +0100")
References: <20051103133809.42242.qmail@web54115.mail.yahoo.com>
	<uy845dy5n.fsf@agharta.de>
Message-ID: <upsphdwv9.fsf@agharta.de>

On Fri, 04 Nov 2005 02:01:08 +0100, Edi Weitz <edi at agharta.de> wrote:

> I should probably change the test and the examples.

But I have to admit that I don't know how.  I can't seem to find a
simple example using gdImageLine where one can really see the effect
of anti-aliasing.  If someone on this list has a nice example to
replace the one in the docs (and the test suite) please send it.

Thanks,
Edi.


From edi at agharta.de  Fri Nov  4 01:01:08 2005
From: edi at agharta.de (Edi Weitz)
Date: Fri, 04 Nov 2005 02:01:08 +0100
Subject: [cl-gd-devel] Antialiased lines, arcs ... don't work, even the
	test orig pic is wrong...
In-Reply-To: <20051103133809.42242.qmail@web54115.mail.yahoo.com> (arry's
	message of "Thu, 3 Nov 2005 05:38:09 -0800 (PST)")
References: <20051103133809.42242.qmail@web54115.mail.yahoo.com>
Message-ID: <uy845dy5n.fsf@agharta.de>

Hi!

On Thu, 3 Nov 2005 05:38:09 -0800 (PST), arry <frinkofox at yahoo.com> wrote:

> I am using CL-GD for over a week now and found out that the
> antialiasing doesn't work at all...  The test, I think was number 4,
> produces the same result as its orig picture in the orig folder, but
> the orig picture in the orig folder is also wrong - there's no
> antialiasing in it...  And this orig pic comes from the package
> itself...
>
> Is this a bug?

Here's the equivalent C code:

  #include "gd.h"

  int main () {
    gdImagePtr im;
    FILE *out;
    int orange, white, red;

    im = gdImageCreate(150, 50);

    orange = gdImageColorAllocate(im, 255, 165, 0);  
    white = gdImageColorAllocate(im, 255, 255, 255);
    red = gdImageColorAllocate(im, 255, 0, 0);  

    gdImageFilledRectangle(im, 50, 0, 99, 49, white);
    gdImageSetThickness(im, 2);
    gdImageLine(im, 5, 10, 145, 10, red);
    gdImageSetAntiAliased(im, red);
    gdImageLine(im, 5, 25, 145, 25, gdAntiAliased);
    gdImageSetAntiAliasedDontBlend(im, red, orange);
    gdImageLine(im, 5, 40, 145, 40, gdAntiAliased);

    out = fopen("/tmp/test.png", "w");
    gdImagePngEx(im, out, 3);
    fclose(out);

    gdImageDestroy(im);
  }

If I run this on my Debian system (with GD being 2.0.33) then diff
says the result is identical to the image produced by CL-GD.  (Note
that the GD documentation for gdImageSetAntiAliasedDontBlend's
signature is wrong, BTW.)

I just re-checked the GD docs and there I find this:

  Drastically faster, less memory-intensive antialiased drawing,
  thanks to Pierre-Alain Joye. This code was imported from the PHP
  "fork" of gd. The API for antialiased drawing has not changed,
  however the implementation has been completely replaced. Antialiased
  line drawing does not support widths other than 1, however this did
  not work properly with the other implementation of antialiasing
  either.

Aha, so this once (kind of) worked in the GD versions I started with
and ceased to work in the meantime.  I should probably change the test
and the examples.

Cheers,
Edi.


From frinkofox at yahoo.com  Thu Nov 10 12:54:50 2005
From: frinkofox at yahoo.com (arry)
Date: Thu, 10 Nov 2005 04:54:50 -0800 (PST)
Subject: [cl-gd-devel] Antialiased lines, arcs ... don't work,
	even the test orig pic is wrong...
In-Reply-To: <upsphdwv9.fsf@agharta.de>
Message-ID: <20051110125450.7323.qmail@web54101.mail.yahoo.com>

Does anybody know whether antialiasing works at all in gd ?

Edi, do you keep this old version of cl-gd or gd where antialiasing still worked? or the
.dll file for windows also...
Everything is jagged without antialiasing, not beautiful, can't draw anything in a proper
manner...

all the best
arry

--- Edi Weitz <edi at agharta.de> wrote:

> On Fri, 04 Nov 2005 02:01:08 +0100, Edi Weitz <edi at agharta.de> wrote:
> 
> > I should probably change the test and the examples.
> 
> But I have to admit that I don't know how.  I can't seem to find a
> simple example using gdImageLine where one can really see the effect
> of anti-aliasing.  If someone on this list has a nice example to
> replace the one in the docs (and the test suite) please send it.
> 
> Thanks,
> Edi.
> 



		
__________________________________ 
Yahoo! FareChase: Search multiple travel sites in one click.
http://farechase.yahoo.com


From edi at agharta.de  Fri Nov 11 22:29:38 2005
From: edi at agharta.de (Edi Weitz)
Date: Fri, 11 Nov 2005 23:29:38 +0100
Subject: [cl-gd-devel] Antialiased lines, arcs ... don't work, even the
	test orig pic is wrong...
In-Reply-To: <20051110125450.7323.qmail@web54101.mail.yahoo.com> (arry's
	message of "Thu, 10 Nov 2005 04:54:50 -0800 (PST)")
References: <20051110125450.7323.qmail@web54101.mail.yahoo.com>
Message-ID: <uek5makdp.fsf@agharta.de>

On Thu, 10 Nov 2005 04:54:50 -0800 (PST), arry <frinkofox at yahoo.com> wrote:

> Does anybody know whether antialiasing works at all in gd ?

I think you should try with a simple C program and ask the GD
maintainers.  This doesn't seem to be a CL-GD issue I'm afraid.

> Edi, do you keep this old version of cl-gd or gd where antialiasing
> still worked? or the .dll file for windows also...

No, sorry, unfortunately not.

Cheers,
Edi.


From frinkofox at yahoo.com  Tue Nov 22 11:54:35 2005
From: frinkofox at yahoo.com (arry)
Date: Tue, 22 Nov 2005 03:54:35 -0800 (PST)
Subject: [cl-gd-devel] just to share
Message-ID: <20051122115435.8412.qmail@web54110.mail.yahoo.com>

I just want to share a problem with cl-gd or maybe gd itself
i don't know...

Has anybody tried to   draw an ellipse which starts  and ends
at a precise angle?
Well, I have and it turns out that the angles that i have 
specified are not drawn exact , they are either bigger or
smaller and SOMETIMES theyy are correct!

is this a problem with gd  or cl-gd?

example
I draw an ellipse that passes through two points
then I want this ellipse to start from exactly one of the two point
aand end at the other one. I calculate the angles with tthe help
of the function (atan) or (asin) using the coordinates of
the center of the ellipse and the two given points...

and the angles drawn don't match with the given points
That's funny!!! ridiculous!!!
There's always 1 or 2 maybe more degrees discrapency 
between the angles drawn and the given points


Anybody encountered the same thing?


		
__________________________________ 
Yahoo! FareChase: Search multiple travel sites in one click.
http://farechase.yahoo.com


From edi at agharta.de  Tue Nov 22 12:38:51 2005
From: edi at agharta.de (Edi Weitz)
Date: Tue, 22 Nov 2005 13:38:51 +0100
Subject: [cl-gd-devel] just to share
In-Reply-To: <20051122115435.8412.qmail@web54110.mail.yahoo.com> (arry's
	message of "Tue, 22 Nov 2005 03:54:35 -0800 (PST)")
References: <20051122115435.8412.qmail@web54110.mail.yahoo.com>
Message-ID: <upsoskgbo.fsf@agharta.de>

On Tue, 22 Nov 2005 03:54:35 -0800 (PST), arry <frinkofox at yahoo.com> wrote:

> I just want to share a problem with cl-gd or maybe gd itself i don't
> know...
>
> Has anybody tried to draw an ellipse which starts and ends at a
> precise angle?  Well, I have and it turns out that the angles that i
> have specified are not drawn exact , they are either bigger or
> smaller and SOMETIMES theyy are correct!
>
> is this a problem with gd  or cl-gd?
>
> example
> I draw an ellipse that passes through two points then I want this
> ellipse to start from exactly one of the two point aand end at the
> other one. I calculate the angles with tthe help of the function
> (atan) or (asin) using the coordinates of the center of the ellipse
> and the two given points...
>
> and the angles drawn don't match with the given points That's
> funny!!! ridiculous!!!  There's always 1 or 2 maybe more degrees
> discrapency between the angles drawn and the given points

I don't have time to investigate this at the moment but if in doubt
I'd first try with a similar C program to see if the problems are in
GD.  You also might want to check if any rounding errors occur within
CL-GD.

Good luck,
Edi.


From frinkofox at yahoo.com  Wed Nov 23 13:32:07 2005
From: frinkofox at yahoo.com (arry)
Date: Wed, 23 Nov 2005 05:32:07 -0800 (PST)
Subject: [cl-gd-devel] just to share
In-Reply-To: <upsoskgbo.fsf@agharta.de>
Message-ID: <20051123133207.79835.qmail@web54111.mail.yahoo.com>

never mind my post I figured myself whats wrong

The angles that the function "draw-arc" accepts are 
actually the angles that the line that draws the ellipse
forms with the x-axes, which is not the angle between
the radius of the ellipse and the x-axes. Hope that is  clear

bye

--- Edi Weitz <edi at agharta.de> wrote:

> On Tue, 22 Nov 2005 03:54:35 -0800 (PST), arry <frinkofox at yahoo.com> wrote:
> 
> > I just want to share a problem with cl-gd or maybe gd itself i don't
> > know...
> >
> > Has anybody tried to draw an ellipse which starts and ends at a
> > precise angle?  Well, I have and it turns out that the angles that i
> > have specified are not drawn exact , they are either bigger or
> > smaller and SOMETIMES theyy are correct!
> >
> > is this a problem with gd  or cl-gd?
> >
> > example
> > I draw an ellipse that passes through two points then I want this
> > ellipse to start from exactly one of the two point aand end at the
> > other one. I calculate the angles with tthe help of the function
> > (atan) or (asin) using the coordinates of the center of the ellipse
> > and the two given points...
> >
> > and the angles drawn don't match with the given points That's
> > funny!!! ridiculous!!!  There's always 1 or 2 maybe more degrees
> > discrapency between the angles drawn and the given points
> 
> I don't have time to investigate this at the moment but if in doubt
> I'd first try with a similar C program to see if the problems are in
> GD.  You also might want to check if any rounding errors occur within
> CL-GD.
> 
> Good luck,
> Edi.
> 



		
__________________________________ 
Yahoo! FareChase: Search multiple travel sites in one click.
http://farechase.yahoo.com