DavidW

Which one?

See the readme, but basically: at your own risk.

Indeed. Plus, I did an install check on classic BG2 (though not BGT) before releasing v34, and it installed cleanly.

It's serendipity actually - I'm pretty sure you'd have had the same bug with SR+SCS, which I didn't know about (I didn't test on SR this time around) and I think I've now fixed it. I don't think there's any tests I immediately need done - I did tests with BG2/TUTU (though not BGT) and got a clean install, and I'm also going to release 34.1 in an hour or two once it's passed its install checks, so it probably makes sense to wait for that.

@Bartimaeus Any chance there's a stray copy of 'dvimhere.mrk' lying around in your override folder? There shouldn't be, given that mod load, but I can't otherwise see how the problem you're describing could happen. (It's a consequence of the 'friendly_fiends' function being run early enough in the install order that it tries to do AI before detectable spells and the difficulty system are in place - that only happens when demivrgvs=1, which in turn is set according to whether 'dvimhere.mrk' exists. In any case, this is showing up a minor compatibility problem between SCS and SR.

This isn't true, actually. It's only on a failed save that enemies detect immunities.

I'm going to need someone to confirm that we get these parse errors on a BGT install without the anomalous ini setting, before i look at it more closely.

OK, if this is general to BGT (and not specific to Salk's ini choice) I'll try to address it. (I didn't run BGT install checks this time around.)

Thanks! - but as Roberciiik pointed out above, I forgot to traify dw#carsa.d and carsa_add.d. I'll do so in 34.1 (probably released later today); I assume it makes sense to wait till you've had a chance to do those lines too before including the updated translation?

I haven’t a clue.

Yes... that's probably right. I'm not sure that can be helped; I'll think about it.

It's basically the same issue as with normal modding: overwrites need to come before patches. Most self-described 'UI mods' replace the whole UI. Mods with UI tweaks normally do it with REPLACE_TEXTUALLY.

When you ran batch mode, you should have been asked about 'A' too - 'reinstall/uninstall/no change/quit'. What did you answer?

Doing this will break SCS's difficulty system, for instance.

Yes, I can reproduce that - problem with v34 that crept in and somehow missed my SoD test install. Wait till tomorrow (it's not easily hotfixable).

OK, so just to check I'm not being stupid, if I create a text file called stratagems.sh, whose contents are 'weinstall %MOD_FOLDER% --uninstall --force-install-list blah'; put that text file in weidu_external/batch; typed 'sh weidu_external/batch/stratagems' (/ rather than \, yes?) at the command line the result would be the same as entering 'weinstall %MOD_FOLDER% --uninstall --force-install-list blah' at the command line? & this is Linux, not OSX, yes?

I think you're using WEIDU v246.

I'm afraid I triaged that one as too much of an edge case. The parse errors are probably harmless.

Damn; you're right. I meant to do that before release and forgot (and it doesn't show up as a problem when testing). I need to do a 34.1 release over the weekend anyway, to fix a couple of 2.6 issues I missed first time round. I'll try to traify those two files then too.

OK, I'll give that a go. What's the command-line instruction to run an sh file?

Batch mode doesn't affect the interaction with the ini: it fine-tunes options exactly the same way as if you install the usual way.

The reason I'm not supporting other OSs is primarily that I'm completely ignorant of what their equivalent of .bat file are and how to format them (and secondarily that I don't have a non-Windows testing environment). If someone wants to give me idiot-proofed instructions on how to feed a text file with setup-stratagems --uninstall --force-install-list blah to WEIDU in OSX or Linux, I don't mind trying to implement it.

I don't know about 'should', but it has lots of advantages. (The point is if you know enough about WEIDU to mess with the innards of a tp2, you can find your own fast-install solution.)

Thank you - that's a very clean repro case. I'll see if I can reproduce it myself.

But if the trap went off, you can record that fact via the trap script. (I think you're right about Disarmed() in the area script, but you might want to play around with where it does work.)