Messages

1

AGI Development Tools / Re: Any AGI engine upgrade projects? Looking for better graphics.

« on: May 12, 2025, 12:38:37 AM »

It was based on Carl Muckenhoupt's SCI Decoder

Great, now we have a name to blame for the default palette being wrong in basically everything since.

Well, I spotted his name in a Telltale production (I forget which one). So he is/was active for a long time after that. And there were other bugs too, for example his routines choke on pics that contain the monochrome opcodes we've discussed earlier.

2

SCI Development Tools / Re: SCI Companion V3 - alpha build notes/bugs/feature requests

« on: May 04, 2025, 05:30:17 PM »

Am I stupid for not being able to tell the difference (apart from the fact that the two images are cropped slightly differently)?

3

AGI Community How To's & Tutorials / Re: Better dialogue in AGI

« on: April 29, 2025, 07:41:09 PM »

This would be easy in SCI (see for example the CatDate demo by Kawa where the intro presents its dialogue very differently to the main game). I think AGI allows you to print text given text mode coordinates (meaning it works in multiples of 8 pixels or so), but I'm not sure.

4

AGI Development Tools / Re: Any AGI engine upgrade projects? Looking for better graphics.

« on: April 28, 2025, 03:13:37 PM »

Ah yes, AutoLISP. And as for SCI being hard due to the different OO concepts, I remember writing the first class dumper way back in 1998 (I still have the source!). It allowed me to deduce the logic of the class hierarchy pretty quickly, and I was entranced.

It was based on Carl Muckenhoupt's SCI Decoder, and I ported his decompression routines to Pascal. The one thing that caused me trouble was that Turbo Pascal doesn't have an unsigned long type, so I had to write those parts in asm.

5

SCI Development Tools / Message file comments

« on: April 28, 2025, 03:09:12 PM »

I've known about these for a while, but they are oddly disconnected from the rest of the message file, and so I couldn't figure out how to link message and comment (plus the comment text seemed to contain junk, which turned out to be the timestamp and an ID which isn't the usual N/V/C/S tuple). But I doubled down and figured it out. The ones from GK1 are especially informative. They even include timestamps (not shown in the message editor), so we can know what the developers were doing during crunch time. Example (room 210, the bookshop):

Code: [Select]

    noun = 11
    verb = 0
    case = 57
    seq = 12
    talker = 8
    offset = 25795
    refNoun = 0
    refVerb = 0
    refCond = 0
    refSeq = 0
    string = "Like I said.  Anyway, we shouldn\'t stand out here and gab; someone on the street might see us.  Let\'s go in back and talk." (total 124)
    comment = Container:
        text = POST:  Mosely turns and walks back into studio.  Gabriel follows automatically.  Fade to 220. (total 93)
        timestamp = 1993-11-11 16:37:56
        msgnum = 744


6

SCI Community How To's & Tutorials / Re: Thoughts on the LSL4 babes

« on: April 20, 2025, 06:46:28 PM »

The format of an image in a Picture resource is identical to a View's cel. Which means it uses RLE.

Yup, but this is not very efficient for a (dithered) gradient. 1 pixel of color x, 1 pixel of color y, etc.

7

SCI Community How To's & Tutorials / Re: Thoughts on the LSL4 babes

« on: April 19, 2025, 02:10:33 PM »

I think it was SCI01, but it's been 25 years since I worked on this...

8

SCI Community How To's & Tutorials / Re: Thoughts on the LSL4 babes

« on: April 17, 2025, 10:11:05 AM »

You'll notice if you look into it that this is one of the few non-vector backgrounds in the game. Cos there's no way you're doing that gradient in vectors.

Sure, all they had to do was create a GradientFill opcode and... oh, never mind.

9

SCI Development Tools / Re: New resource type discovered in SQ3 Mac

« on: April 10, 2025, 07:51:16 AM »

If it was an art test, why would you create a new Mac-specific resource type for it, on a game port at that?

10

SCI Development Tools / Re: New resource type discovered in SQ3 Mac

« on: April 09, 2025, 10:54:18 PM »

There are no black and white pics as far as I can tell. That's why I wondered about the upscaling. But I'll post an example of that black-and-white-but-not-quite thing I talked about. The grabber is in black and white, but the strobe light was left untouched:

11

SCI Development Tools / Re: New resource type discovered in SQ3 Mac

« on: April 09, 2025, 03:57:04 PM »

Yeah, but I was thinking that a mixture of upscaled views and pics with redrawn ones would look bad. If nothing else because of the differences in art style that come from having more pixels (see Roger's eyes in the above for example). And those pics... the SCI flood-fill algorithm is rather sensitive, so just upscaling them could spell trouble.

12

SCI Development Tools / Re: New resource type discovered in SQ3 Mac

« on: April 09, 2025, 03:04:18 PM »

(continued)

What I did here was extract the 0x8f resources (which scummvm thinks are message resources), rename them to view.* and fix the type tag so that the system knows they are views. Then they load without problems in SCI Companion (a few resources fail to load; I haven't investigated why). They contain a palette section, as SCI01 views do (and which QfG2 uses a lot for special effects).

13

SCI Development Tools / New resource type discovered in SQ3 Mac

« on: April 09, 2025, 02:55:52 PM »

So I was looking at Sluicebox's newly released source, and discovered this innocuous comment:

Code: [Select]

                   // HACK: sq3-mac and hoyle2-mac have resources of type 0x8f (message)
                    // which are of course not message resources, so version detection throws.
                    // i don't know what they are, neither does scummvm.
                    // until i update MessageReader to detect and reject these, whatever they are,
                    // i'm just skipping attempting to parse messages from these two text games.
Intrigued, I took a look and I found that they are alternate views for some of the graphics, mostly but not quite black and white. About half of the views have corresponding "altviews"; so I'm guessing there must be some sort of conversion mechanism for the remainder. Now, as Kawa once said about 2-color graphics, they are a bit meh - but unlike the monochrome pics we discussed then, these views are often bigger than the originals. Compare:

14

SCI Community How To's & Tutorials / Re: Thoughts on the LSL4 babes

« on: April 09, 2025, 11:50:21 AM »

It was way worse in other games. Early on, Sierra had this tool - written in Prolog - that would lay out the resource volumes on each disk with combinatorial optimization. They didn't carry it forward from AGI to SCI for some reason.

15

Everything-Else / Re: The modern definition of "AGI"

« on: April 08, 2025, 03:39:12 PM »

I noted the same thing here: https://sciprogramming.com/community/index.php?topic=2060.msg74857#msg74857