Thank you for answer.
As you said, I tested everything.
I'd recommend using Farty's branch of GPTP which integrates DatExt with GPTP. Newer versions of PyMS have the ability to add extended entries to .dat files which can then be used as normal. FireGraft gets a little wonky with extended unit entries, so if you're additing additional units that need buttonsets, I'd recommend using something like Farty's WIP graft compiler (although I don't have a link to it on hand)
oh, i used this Recently I've been using DatExtend which has been merged with gptp and I didn't mention this in my initial question.
I used DatExtend before being merged into gptp in 2020 and only units, flingy, sprites, images data extension was successful.
However, as I said, there was a problem when running the plugin-applied mod with FireGraft. The program could not be closed normally. And in this state, it seemed to be initialized with the existing data whenever a new save was made. When running, an error saying that the command length was incorrect came out and it could not be executed. In addition, there was a phenomenon that the button set was not applied, and there was a crash when selecting a Zerg egg and Cocoon placed in the map editor.
This is my experience using the DatExtend plugin in 2020.
A while ago I rediscovered DatExtend and found that it was merged with the gptp plugin and tried it out.
I tried increasing all the data to 65535 in datex.cpp as an experiment, and it built successfully, but I haven't noticed any change since. I'm wondering what to do next.
I also found a serious bug while using this gptp plugin. A crash occurs when Corsair and Valkyrie's Air Splash attack hits. Additionally, some spells such as Dark Swarm, Consume, Stasis Field, and Disruption Web are not used or behave erratically. I also found a bug where if I right-clicked a non-enemy building with my unit, it would be judged to follow it. I didn't touch these parts and found out that this is a bug in gptp itself. Because of this, the problem was difficult to find and could not be resolved.
It's a pity that gptp is not available in the remaster. I wonder if the day will come when gptp can be used in the remaster as well.
Still, it's comforting that some of the features of gptp are available as mtl plugins.
Yes, these should work in 1.16. never tried it though.
Together, these three plugins allow to do some quite remarkable things, far beyond that which was ever available in old times. AICE in particular is a powerhouse that allows pretty rad effects to be implemented, going beyond strictly graphics. I am very enthusiastic about it, and there's always something new to do with this plugin. :-)
AISE requires you to use a compatible fork of PYMS->PYAI to edit the aiscript.bin/bwscript.bin files. It fixes some AI issues from unmodded game, and of course adds new commands that significantly expand the Scripting capabilities, and add more opportunities to interface between AI Scripts, triggers and iscript.
I experimented right away. I knew that I just had to change the dll extension plug-in to the qdp extension, and I immediately tested it and it was definitely usable. However, there were some problems with the AICE plug-in. Some texture frames in the game are displayed as if they are lagging. This is a problem that did not exist when applied to the remastered. Other than that, there were no other issues.
It seems that the solution to this is not well known.
Thanks to everyone who responded. However, there are not many other courses for mtl and aise, so I think I will still need more time.
None.