Messages

1

Everything-Else / Re: I give up

« on: October 07, 2026, 11:19:03 PM »

I haven't been a part of this forum as long as a lot of people here, but I would like to say that the members, especially Kawa, were invaluable in helping piece together what I know about SCI and AGI. Honestly, the interactions weren't always pleasant or didn't always solve the problems I was struggled with, but they did point me in the right direction.

I'm not a programmer by profession, but working with these games has taught me things I didn't think it was possible for me to learn. I now feel fortunate that I was able to experience that journey in my own way, while having direct access to many of the amazing people who did the RE work to make that even possible.

I've used LLMs recently on some non-SCI projects in an effort to learn more about a technology that, as much as I dislike it, probably isn't going away. In addition to several other complaints, I really feel that its use took away a lot of what I love about programming as a hobby.

It's not right for me to gatekeep, and if LLMs make game design more accessible to people who don't want to invest thousands of hours learning a dead interpreter scripting language, that's probably a good thing.

However, I'm very empathetic to Kawa's point. There have always been alternatives; Adventure Game Studio is the first to spring to mind. Not that AGS is wrong or bad, but it's not SCI/AGI, and agentic LLMs feels the same. I didn't join this forum to create a point and click game, I joined it to learn how to create something myself in SCI/AGI.

If LLMs are driving people like Kawa away from the community I'll have to say that it's a net loss.

2

SCI Development Tools / Re: LSL4 Demo

« on: September 15, 2026, 02:57:19 PM »

The art and animation looks really good

3

SCI Syntax Help / Re: Script disposal question

« on: July 31, 2026, 12:02:01 AM »

(script dispose:) also has the side effect of resetting some properties, so that the script can be reused later.

This seems counter intuitive, as I'd expect you'd want every property reset if previously disposed

4

SCI Syntax Help / Re: Script disposal question

« on: July 30, 2026, 03:25:10 PM »

Great, thank you for the information!

5

SCI Syntax Help / Re: Script disposal question

« on: July 30, 2026, 02:03:12 PM »

So my question would still be does (self dispose: delete:) do anything different than (self dispose:) ?

6

SCI Syntax Help / Re: Script disposal question

« on: July 30, 2026, 01:16:01 PM »

But use-after-free is a common source of issues in many languages and you should never rely on it.

I wouldn't dare, but to confirm, it's just sitting around waiting to be overwritten after dispose is called, right? Would delete: do anything different here or at least clear the pointer to the disposed object so it's no longer accessible? Just curious

7

SCI Syntax Help / Script disposal question

« on: July 30, 2026, 12:54:02 PM »

It seems that if you dispose a script in SCI, you can still check its properties afterward and get the expected values from before disposal. It's not until room change that it appears to really flush the script entirely. This feels a bit confusing because if it's not freeing up the heap before room change, what's the point?

Is it just marked disposed and still accessible until overwritten by new objects? It would be helpful to understand how that's working under the hood. TY

8

The Games and other Sierra Adventure stuff / Re: Rare AGI Games on standard PC DOS

« on: June 26, 2026, 10:30:31 PM »

BTW, the AGI Wiki is no longer being updated.

I think this is why Threepwang is feeling a bit confused. The AGI wiki lists 1.0U 2.272 as from 1984, which seems to be a mistake. KQ1 1.0U 2.272 as Lance mentioned is from Nov 1986 and logically a DOS game, right? 

It's listed as 84 on the Sierra wiki as well

9

SCI Development Tools / Re: Looping Audio in SCI1.1?

« on: April 28, 2026, 06:57:22 PM »

(using unused numbers, as Sound and Audio assets share a namespace).

I believe that despite seeming to share the same resource numbers you can differentiate between sound and audio by prepending a 1 before the number to use digital audio.

For example, in LB2 most songs are MIDI, but "The Archaeologist Song" has both a MIDI version and an audio version with lyrics. So you can call the MIDI version with 312 or the digital version with 1312.

10

SCI Development Tools / Re: Converting Pics from SCI1 to SCI1.1 format

« on: April 15, 2026, 10:33:40 AM »

Isn't that the expected behavior? There should only be one palette available per room. Loading a pic and view with conflicting palettes is going to overwrite the previous colors.

That's why you can't transition between rooms with conflicting palettes without distortion or fading to black. This was often the case with EcoQuest where they would use checkerboard transitions between rooms with incompatible palettes and it looked terrible.

I believe it's also the reason most actors keep to the first 16 (or 32) colors so they can move between all rooms without issue. The rest of the colors are then used for pics or per room objects.

As an aside, with scummvm you can do some palette tricks, like overriding the static SCI0 palette. You're still limited to 16 colors, but you can assign them to any RGB value, allowing for a monochromatic effect with a dozen shades of a single color.

11

SCI Syntax Help / Re: QfGIVCD view corruption when editing 936.v56

« on: April 13, 2026, 12:34:56 PM »

I fixed a bug related to that many years ago (pull request #1467 if you're interested).

In which repo? Not seeing the PR in the IceFall or Kawa githubs.

12

SCI Syntax Help / Re: QfGIVCD view corruption when editing 936.v56

« on: April 09, 2026, 05:04:53 PM »

Seems to be. I've been working with ashlancer's qfg4-enhanced patches recently and everything appears to compile and work in game correctly. I've not encountered any issues and, so far, the translators using the patches haven't reported any problems. Probably why I didn't realize it was sci2.1 - it's working too well except for the views, which still seem to break randomly in different ways.

Edit: There are some issues I forgot about. KString (0x83) and KArray (0x82) will not compile in SCICompanion. KString should use the same hex value as String, so we can skirt that easily. KArray is only used twice in QfG4CD (array.sc and polyedit.sc), but that might require manual hex editing, because Array is a class in QfG4. None of the exported patches in Ash's repo contain KString or KArray, so I haven't actually dealt with them yet. Also pretty sure I've had issues compiling Phant and LSL7 scripts before, but too long ago to remember. QfG4CD seem mostly good, but since I'm only compiling/testing what Ash patched there certainly could be issues I haven't encountered.

I'm looking at Enrico Rolfi's Tradusci, but 2.4 was hosted in google drive, which blocked the downloads because of 'infection'; which is complete nonsense because a google 'scan' of .scr files always finds an infection, until you rename the extension and the virus magically disappears. /rant

Enrico removed the .exe from the source code for 2.4 this morning, so I'll try building it after work, but I understand it's not so much a view editor as a tool to replace cels of the same size with externally edited bitmaps. It will be interesting to see what its limitations are.

As for the original 936.v56 file I was working with in SCICompanion, changing a pixel adds 10k of bytes in four repeating blocks (view has 4 loops) to the end of the file. I'd like figure out what that extra data is supposed to be and if any loop or cel entries actually point to it. Also need to compare other views known to corrupt when saved, but time is in short supply at the moment.

13

SCI Syntax Help / Re: QfGIVCD view corruption when editing 936.v56

« on: April 08, 2026, 10:38:14 PM »

I finally looked up QfGIVCD on the sciwiki. Turns out it's SCI2.1, not SCI1.1 like I assumed.

I'm guessing SCICompanion just isn't able edit those views properly. It's not just 936, probably most the views have issue of some kind when edited. The corruption is expressed in seemingly random ways too based on the few I tested this afternoon.

PhotoSCIchop can modify these views with a bit of effort, but I can't really expect SC to.

SCI2 is also annoying that the Display kernel function was removed, so it's not easy to just print a string of text without a bmp background anymore. Oh, well

14

SCI Syntax Help / QfGIVCD view corruption when editing 936.v56

« on: April 04, 2026, 12:35:50 AM »

Here's a weird one.

If you make any changes to QfG4CD view 936, some of the cels get corrupted (see top three buttons in attached image). Doesn't matter if you change a pixel, alter the palette, export every cel to bmp and reimport them, move an origin point, etc.. Saving any change breaks the view in game. Adding new loops and moving cels around makes the corruption manifest differently, but in erratic ways based on my testing so far.

If you're wondering who cares about modifying some dumb shop buttons, the answer is, of course, translators.

I'm hesitant to blame this on a SCICompanion bug, because I'm told PhotoSCIhop complains about the view palette, though I haven't tested that myself.

15

SCI Development Tools / Re: Script issues with QFG1VGA

« on: February 26, 2026, 11:20:21 AM »

Are you compiling with a nightly Kawa build or the SCICompanion from the website?

If EO's decompile has any asm blocks in the script your working on, that's a likely source for recompile bugs. That can be replaced using Sluices repo as reference.

You could also post on the Spanish or French translation boards. I'm sure one of them has dealt with this already

Edit: invent.sc does have asm. That's a likely culprit