PDA

View Full Version : Zelda Classic 1.92 Beta 150



Dark Nation
04-07-2002, 08:57 AM
Please note: There is a new support file you need for this version. Get it at the bottom (the Beta 150 NewDefault Patch).

What's new in this version.

1. Added a test feature to ZQuest (under the Quest menu). All it does is run Zelda, so you need to save before you use it or you won't see your changes. Also, in the DOS version (haven't tested the Windows version yet), when you come back from testing, the menu is blacked out and mouse movement is jerky. No idea why. :shrug: If I can't get this problem licked soon, I'll take it back out.

2. Psuedo-translucency effects for layers.

http://www.armageddongames.com/darknation/images/blending1.gif
(nam)

Note:
-No, that's not 16 or 24-bit, it's still 8-bit color (using a translucency table).
-I haven't tested it with palette cycling, but I bet it won't work right. Colors that cycle probably won't appear to cycle any time they go under a translucent layer.
-No, the fog isn't scrolling. It's stationary.
-No, that's not an environmental effect, it's a layer. In fact, if you walk under the archway, it will cast a shadow on you.
-No, I don't want you telling anyone about this. This is probably not a production feature and will most likely not be given to the custom quest contest participants as it is still buggy. For instance, when the image is the size of your screen, the colors look a little odd. However, just for your enjoyment, I will probably stick it in beta 150 to get your opinions. I've decided to go ahead and let the custom quest people fiddle with it, too, if Cyclone wants to give them this version. However, I don't want this feature discussed outside of the beta tester or custom quest forums. And don't hint about it, either. Such as "we're currently testing a feature that I can't tell you about". Not until/unless I decide to keep it. To that end, I went ahead and update the beta reports thread so this information doesn't accidently slip out. :D

On another note... In ZQuest, pressing the 'b' button will now rebuild the translucency table. I'm working on making this an automatic thing without slowing down ZQuest too much. You mainly need to do this if you move from a screen with one color set to a screen with another color (either from pressing the F4 key or pressing one of the number keys). This will make the translucency effects look 'right'. To turn on translucency for a layer, go to the layer screen and click the 'Trans:' check box for that layer. Very simple. I've already tried different things with this and they work quite well. I've made the rain in newfirst translucent, the waterfall, some fog (as you saw).

The layer data screen looks different. Don't worry about the extra stuff for now. Just worry about the map, the screen, and the translucency. The other values may get pulled out one day if I don't get around to implementing them.


I'm sure that the bug that previously was corrupting quests that had exactly 256 combos is still around. I think it's a roaming bug, though, so something else may trigger it now. You might want to turn on the auto-backup feature.


Anyway, here are the files:

Zelda Classic 1.92 Beta 150 for DOS (http://www.armageddongames.com/darknation/beta/zc192b150/zc192b150d.exe)
+zelda.exe
+zquest.exe

Zelda Classic 1.92 Beta 150 for Windows (http://www.armageddongames.com/darknation/beta/zc192b150/zc192b150w.exe)
+zelda-w.exe
+zquest-w.exe


Here's the support archives:

Zelda Classic 1.92 Beta 147+ Support Archive (http://www.armageddongames.com/darknation/beta/zc192b147/zc192b147s.exe)
+1st.qst
+2nd.qst
+3rd.qst
+new enemy tiles.pdf
+new.til
+newfirst-fixed.qst
+zelda.dat
+zgp.dat
+zquest.dat

Zelda Classic 1.92 Beta 144+ Windows Support Archive (http://www.armageddongames.com/darknation/beta/zc192b144/zc192b144ws.exe)
+alleg40.dll


You will also need to download this file:

Zelda Classic 1.92 Beta 150 NewDefault patch (http://www.armageddongames.com/darknation/beta/zc192b150/zgp.dat)
+zgp.dat

Overlord
04-07-2002, 12:34 PM
Well... (and I loath to say this as it shows my laziness.. it also happens in 149) I can't get it to run. ZQuest is fine, Zelda gives me an invalid page fault on startup. Windows version. The exact fault?



ZELDA-W caused an invalid page fault in
module ZELDA-W.EXE at 0167:00461035.
Registers:
EAX=00e593a0 CS=0167 EIP=00461035 EFLGS=00210206
EBX=00000000 SS=016f ESP=0080fcc0 EBP=0080fce8
ECX=00000009 DS=016f ESI=37302d34 FS=3c0f
EDX=37302d34 ES=016f EDI=004b7070 GS=0000
Bytes at CS:EIP:
f3 a5 c7 05 84 70 4b 00 00 01 00 00 c7 05 88 70
Stack dump:
0000002a 00e40560 00000022 78031248 ffffffff 0080fd58 000012d7 00000000 00000000 0000f0d9 00e40590 66900c54 00000001 00e40590 78038008 66900ab4


As I said, beta 149 is hit with it too...

EDIT: Ok.. I'm not as lazy as I thought. I have saved quest files in my 149 folder, which means I was working with it. And now every beta after 144 is giving me the error. :shrug:

Dark Dude
04-07-2002, 12:51 PM
Okay, I ran ZQuest Windows beta 150, and it appears as though any combo that's animated and uses that feature that allows you to use the combo on a different cset (by filling in the aqua square) it's messed up.

I checked zelda.exe, it's messed up there, too.

Overlord
04-07-2002, 01:29 PM
Ok, ignore my above statement. I got it working :redface:

Seems for some reason I was using the 144 support archive :shrug:

Dark Nation
04-07-2002, 02:38 PM
Originally posted by overlord147
Ok, ignore my above statement. I got it working :redface:

Seems for some reason I was using the 144 support archive :shrug:

I was just gonna ask about that. People, if you get an error when trying to run this program, double-check the support archives. Download whatever is referenced in the announcement post. :D

Starkist
04-07-2002, 03:02 PM
Originally posted by Dark Dude
Okay, I ran ZQuest Windows beta 150, and it appears as though any combo that's animated and uses that feature that allows you to use the combo on a different cset (by filling in the aqua square) it's messed up.

I checked zelda.exe, it's messed up there, too.

I noticed this as well.

Also I tried translucifying an animated combo and it came out all weird, I could barely see the tiles beneath it but the translucent combo itself was one solid color.

Daarkseid
04-07-2002, 03:27 PM
How complicated is this translucency table, or perhaps how complicated would it be to predict with some accuracy how a transparency will affect other colors? I've been working with the transparency effects, trying to get my colors to look fairly decent, but it winds up drastically changing certain colors... it could just be a shortcoming of the translucency table since it has to work within the 8-bit color boundaries, but it would be nice to know if there were a method to using transparency.

VEL
04-07-2002, 03:29 PM
I noticed that every tile that uses cset2 is messed up (I checked 149 with the same quest and it wasn't like this). It isn't only the animated ones.

The translucency looks very cool!

Dark Nation
04-07-2002, 03:36 PM
Well, the translucency table basically returns the color in the current palette that is the closest to what the two input colors merged would be.

Dark Dude
04-07-2002, 03:50 PM
I was sort of doing this earlier. I was making the quest in ZQUEST then alt+ctrl+esc and opened zelda.exe, but then zelda.exe didn't have any sound, so my question is, when using the test feature, will zelda.exe have sound? Also, this would elminate my need for extra sprite palletes for dark rooms because you could just put a black translucent layer over it to make the sprites darker, right?

Daarkseid
04-07-2002, 04:16 PM
Originally posted by Dark Nation
Well, the translucency table basically returns the color in the current palette that is the closest to what the two input colors merged would be.

Wow, that sounds much more simple than I had initially thought. Thanks.

Dark Nation
04-07-2002, 04:24 PM
Here's the source code for the table populator for all of you programmers out there. No big secrets here. It's just a slightly modified version of Allegro 4.01's create_trans_table function.



/* create_zc_trans_table:
* Constructs a translucency color table for the specified palette. The
* r, g, and b parameters specifiy the solidity of each color component,
* ranging from 0 (totally transparent) to 255 (totally solid). If the
* callback function is not NULL, it will be called 256 times during the
* calculation, allowing you to display a progress indicator.
*/
void create_zc_trans_table(COLOR_MAP *table, AL_CONST PALETTE pal, int r, int g, int b, void (*callback)(int pos)) {
int tmp[768], *q;
int x, y, i, j, k;
unsigned char *p;
RGB c;

for (x=0; x<256; x++) {
tmp[x*3] = pal[x].r * (255-r) / 255;
tmp[x*3+1] = pal[x].g * (255-g) / 255;
tmp[x*3+2] = pal[x].b * (255-b) / 255;
}

if (callback) {
(*callback)(0);
}

for (x=0; x<PAL_SIZE; x++) {
i = pal[x].r * r / 255;
j = pal[x].g * g / 255;
k = pal[x].b * b / 255;

p = table->data[x];
q = tmp;

if (rgb_map) {
for (y=0; y<PAL_SIZE; y++) {
c.r = i + *(q++);
c.g = j + *(q++);
c.b = k + *(q++);
p[y] = rgb_map->data[c.r>>1][c.g>>1][c.b>>1];
}
} else {
for (y=0; y<PAL_SIZE; y++) {
c.r = i + *(q++);
c.g = j + *(q++);
c.b = k + *(q++);
p[y] = bestfit_color(pal, c.r, c.g, c.b);
}
}

if (callback) {
(*callback)(x);
}
}
}

Daarkseid
04-07-2002, 08:03 PM
I'm not absolutely sure this is a bug, but I noticed that when you delete a screen, it will delete all screen layer info(including the extra stuff) but then also checks off each transparency box for all six layers.

I'd think though that if a full wipe of the layer info was to result from screen deletion, that layer transparencies would be unchecked, not checked.

Maybe this has been mentioned, but it is also present in build 151.