> But maybe the two same errors in both code cancel each other out.
Yes, they would.
> Could you please give a reproducer?
Maybe not. I realized I hadn't pulled for a while, so I did a pull and now
"make all" has errors with the Simple Components. The first I saw was fixed
long ago, before GNAT 13 was released, IIRC: In
deps/simple_components/atomic-access/ada/gnat-sockets-server.adb, procedure
Do_Connect, 'Unchecked_Access needs to be changed to 'Unrestricted_Access. If I
fix that, then there are other errors that are not so obvious. It's as if it's
a much earlier version of the Simple Components. But that is another issue.
I'll look into seeing if I can compile 2.1a and use it to make a reproducer.
But basically, I obtained a Pixel_Data_Type from a Canvas/Context and then
wrote a BMP file of the image. The resulting image was garbled. For a
sufficiently small image you could write a P3 PPM file more easily if you want
to try it out.
All this is based on the usage of X and Y as indices into a Pixel_Data_Type in
procedure Data (Image_Data_Type; Pixel_Data_Type) and function
String_To_Pixel_Data, where you find D (X, Y), for example. There's no
indication in the spec of Gnoga.Types of the meaning of the two dimensions, but
there should be.
--Jeff Carter"Now go away or I shall taunt you a second time."Monty Python and
the Holy Gral
On Thursday, November 16, 2023 at 10:25:39 PM GMT+1, Blady
<[email protected]> wrote:
Hello Jeff,
> Le 14 nov. 2023 à 18:35, Jeffrey R. Carter via Gnoga-list
> <[email protected]> a écrit :
>
> Comments ts based on the 1.7 branch.
>
> There appears to be an error in function String_To_Pixel_Data, based on the
> string returned from function Data (Image_Data_Type) return String: the Y
> loop needs to be the outer loop.
Well, one chance out of two to be ... wrong.
However, I recall having tested the Data procedure and function against
Javascript in canva_test.adb.
But maybe the two same errors in both code cancel each other out.
Could you please give a reproducer?
> In addition, the 4 function calls in the aggregate rely on the order of
> evaluation, which is not specified by the language. While GNAT is unlikely to
> change this, it's safer to assign the results of the calls to variables and
> use them in the aggregate.
>
> Finally, the initialization of S has no effect.
Yes it's typical coding style of David.
I guess it isn't the only one in Gnoga.
You know Ada better than me but I wonder what was David motivation.
> Assuming that the string supplied to procedure Data (Image_Data_Type; String)
> has the same format, then the change to the loops also applies to procedure
> Data (Image_Data_Type; Pixel_Data_Type).
Sure.
Regards, Pascal.
https://blady.pagesperso-orange.fr
_______________________________________________
Gnoga-list mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/gnoga-list
