SCI32 Templates

[Kawa]

I have indeed Played Around.

I've messed with views, mostly, to correct a bug where color #255 was considered remappable (it's not) and ended up skipped over. This also affected backgrounds as they share a format with view cels and thus share a render function. And since scaleSignal had most of its bits unused I uplifted that to add some more remap effects like "replace every non-skip pixel with #0", "draw them as #253", or "draw them as if #253 is drawn on top", which would usually darken them, SCUMM-style.

I've been working on color table remapping like in TSN, too, for a bit.

But compiling on Windows? Well, no. And no multiplayer.

[Dhel]

I have indeed Played Around.

I've messed with views, mostly, to correct a bug where color #255 was considered remappable (it's not) and ended up skipped over. This also affected backgrounds as they share a format with view cels and thus share a render function. And since scaleSignal had most of its bits unused I uplifted that to add some more remap effects like "replace every non-skip pixel with #0", "draw them as #253", or "draw them as if #253 is drawn on top", which would usually darken them, SCUMM-style.

I've been working on color table remapping like in TSN, too, for a bit.

But compiling on Windows? Well, no. And no multiplayer.

Interesting. You are doing this with Omar's sci32 source? Do you have it compiling in DOS? Has it worked in recompiling for any games yet?

[Dhel]

I have indeed Played Around.

I've messed with views, mostly, to correct a bug where color #255 was considered remappable (it's not) and ended up skipped over. This also affected backgrounds as they share a format with view cels and thus share a render function. And since scaleSignal had most of its bits unused I uplifted that to add some more remap effects like "replace every non-skip pixel with #0", "draw them as #253", or "draw them as if #253 is drawn on top", which would usually darken them, SCUMM-style.

I've been working on color table remapping like in TSN, too, for a bit.

But compiling on Windows? Well, no. And no multiplayer.

What do you mean by you are working on color table remapping like in TSN?

[Kawa]

I'm not sure if it was The Sierra Network or some later incarnation, but it had an avatar creator that included the ability to change colors. Except instead of having the exact same hairstyles several times over for different hair colors, they just had the one cel for each hairstyle and such, and remapped it on the fly. I think the DrawCel call was extended to take an Array pointer.

So what I've been doing first is porting SCI32's Arrays to SCI11+.

Edit: found it was the LSCI scripts that use a DrawCel kernel that takes a clut argument instead of scale factors, and a Clut of ByteArray, which in turn is just an Array with its type preset.

[Kawa]

Interesting. You are doing this with Omar's sci32 source? Do you have it compiling in DOS? Has it worked in recompiling for any games yet?

Missed this post. Everything I just described is based on Omer's SCI16 source, version 1.001.100.

I tried to recompile SCI32, but that wouldn't run anything and the code had implied int all over so no semi-modern compiler would take it as-is.

[Dhel]

I'm not sure if it was The Sierra Network or some later incarnation, but it had an avatar creator that included the ability to change colors. Except instead of having the exact same hairstyles several times over for different hair colors, they just had the one cel for each hairstyle and such, and remapped it on the fly. I think the DrawCel call was extended to take an Array pointer.

So what I've been doing first is porting SCI32's Arrays to SCI11+.

Edit: found it was the LSCI scripts that use a DrawCel kernel that takes a clut argument instead of scale factors, and a Clut of ByteArray, which in turn is just an Array with its type preset.

Ahh yes i am very familiar with this. This is how the Realm manages its characters as well, and uses a clut system for remapping. I could probably help give you some insight into this.

[Dhel]

Interesting. You are doing this with Omar's sci32 source? Do you have it compiling in DOS? Has it worked in recompiling for any games yet?

Missed this post. Everything I just described is based on Omer's SCI16 source, version 1.001.100.

I tried to recompile SCI32, but that wouldn't run anything and the code had implied int all over so no semi-modern compiler would take it as-is.

Ahh ok I see in your sci11 repo now. Thats cool I might play around with that.

[Kawa]

Ahh yes i am very familiar with this. This is how the Realm manages its characters as well, and uses a clut system for remapping. I could probably help give you some insight into this.

No need, I've studied it well enough. The hard part would be to actually do the remapping thing.

[Collector]

Any interest in some INN source?

[Dhel]

Any interest in some INN source?

Definitely!

[Collector]

Here is the interpreter and some system stuff.

[Dhel]

Here is the interpreter and some system stuff.

Thank you! I am excited to check it out!

[Dhel]

A little update with the progress of my FotoSCIhop fork. Right now I have the restructuring finished for both view and picture files. I have a modify properties window that allow to modify pretty much every aspect of the view file. This even allows for adding and subtracting loops and cells, and support for editing 10 link points. Right now it is pretty raw, as I plan to clean up the ui a little more before releasing. But I am already having some fun modifying view files to see what everything does exactly. I have also implemented link point drawing so I can we finally see what those vectors are doing and how they work. The pic files need a little more work, but I was able to finish the main restructuring today. I have some ideas to allow for full cell and priority editing directly in fotoscihop.

[Dhel]

Here is the interpreter and some system stuff.

Have you been able to build this yet by any chance?

[Kawa]

Have you been able to build this yet by any chance?

This is the SCI16 interpreter code and system scripts, and LSCI system scripts but no interpreter. There is source code for the video drivers and such though, which includes a cel renderer that takes remap tables (cluts).

So for me, the answer is "most of it, yes".