Bennyboy

I've finally got around to doing a full release of Thimbleweed Park Explorer, or as its now known Dinky Explorer. Thanks mainly to the hard work of one man reverse engineering machine Jan Frederick it now fully supports ReMI, including images, sounds and decompiling the scripts. It even allows patching of the weird.dink script to do things like enabling debug mode features. I'm looking forward to seeing people create their own patches and sharing them. More details and download here.

I think dinky is the scripting language, for ReMI the engine internally seems to be called WeirdEngine

Its under heavy development at the moment , a lot of ReMI stuff has been added. I added support for dumping the audio a few hours ago. I'm not doing a new official release until I can think of a new name for it. Grumpy Explorer is one possibility, any other suggestions welcomed.

We've been updating TP Explorer to add support for ReMI but its still a work in progress. Proper music dumping will come in the next release, I have it working manually but I done the code to do it automatically yet.

Also the closeups are about 0.5% of the graphics in the game. Surely everyone loves that EGA sunset.

I love this thread and much as it pains me to say it Remi is right, it's deffo #1

SCUMM games contain the interpreter (the exe file) and the resource files (the game data). To find the version info for the interpreter you run the game (in dosbox these days) with an invalid command line parameter like /? and it'll output the version of the interpreter and its build date Interpreter Verson 5.2.02 (Nov 21 1991 08:35:15) Unknown flag: '/?' The game data also has version info that you can access by pressing ctrl+v ingame. For years we've had text files with lists of different interpreter versions but not the version info from the scripts. For the 3 of us who care about things like this its a pain to gather the different information. I've made a program that automatically extracts the version information from a folder of SCUMM games, puts it in a grid and lets you export it to a spreadsheet. For the interpreters only the Dos and Windows versions are parsed but for the game scripts its platform-agnostic and has been tested with Amiga, Atari, FM-Towns and Mac games. Download it here. Here is my list of version information. One oddity I've noticed from the game scripts is that the DOTT demo is named in the game scripts as "Day of the Tentacle BBS Contest DEMO". Anyone know anything about any contest for DOTT?

The art and animation looks great in the trailer. I don't understand how anyone could watch that and not be giddy with excitement and positivity.

Amazing work, I love it.

You monster, I will never be able to unhear this now (well done)

Having a closer look into the wavebank format again (its been years), it looks to me like the loop points are embedded in the wavebank not the soundbank as I said earlier. If that's the case then possible solutions are: Rename the MusicNew wavebank + soundbank to MusicOriginal so the game just uses them. I just had a quick test of this, it didn't work so I changed the internal names inside the wavebank and soundbank to MusicOriginal. This didn't work as there's a checksum in the soundbank, so I fixed that and...it worked. Except of course the wrong tracks play as they are listed in a different order in musicnew and musicoriginal. You could patch the loop points back into the wavebank after you've built it. If you use the unxwb tool with the -l and -v flags you'll be able to see the region offset for each sound. I think that's the start position for looped audio, its a 4 byte value. So you could make a tool to change those region offset values in the new wavebank.

Where does it sound noticeably wrong? I've just rebuilt the bank and I don't notice any issues so far in woodtick. I cant use your save from the first post, every version of SE2 that I have packs their saves into monkey2.bin.

XGS files are settings for the project. The blocksize might not be it, its just a possibility as I know that it can affect loop points.

As Serge says, its not the audio data that contains the loops its the soundbank (xsb file). You can also build soundbanks in XACT, you drag a track from the wavebank into it and build cues and loop points but that will be a hassle. Again as Serge says, you would expect the game to just use the loops from the xsb so I would guess its one of: ADPCM blocksize - this can affect looping as the blocksize affects block alignment. Try lowering Samples per block on the wavebank. Your wavebank has the SYNC_DISABLED and BANK+ENTRYNAMES flags set - the original doesn't. On the wavebank in xact turn off 'friendly names' and turn on 'sync in-game data'. I havent built it to try but these should remove those flags.

I just gave it a quick try in WinUAE and those keys didn't work, but then again I couldn't get debug mode working on it either so I may have had issues with the keymap or a wonky copy of the game. Shift+F9 and Shift+F10 are indeed the keys for recording/playing back a demo. Whether they are active depends on the version of the game. Normally I think you you need a full windex debug interpreter for the recording/playback functionality to be enabled (like the one included with the MI2 talkie prototype). So the pc cd version of MI2 doesn't let you do it, even in debug mode. However that functionality is enabled in the pc floppy version of MI2 without needing to enable debug mode. So I can well imagine that there's a version of the Amiga MI2 interpreter where playback and recording is enabled.