Alright, I'll forward that to them. Thanks.

Any chance we can get the AI-related discoveries from this thread put into PyAI's documentation?

edit: Attempting to ctrl+z in PyAI does not revert changes successfully. Sometimes the text is deleted, and then the deleted text is pasted back in a bizarre way.

Post has been edited 1 time(s), last time on Jul 17 2017, 10:48 pm by Pr0nogo.

PyMS is expecting installation registry keys to exist?

This should be fixed now.

Should be fixed now.

Yep, I have plans to look into making updates for things from these threads:
Little known iscript/images.dat tricks
.dat file unknowns
Defining AI commands and functions
Information on orders.dat fields

Hmm, I'm not sure how much control I have over that. I'll look into it. Do you have some steps I could follow to reproduce the issue?

It shouldn't require them, but it does try to use them if they exist, it just had faulty logic. Should be fixed now.

Thanks!

Post has been edited 1 time(s), last time on Jul 18 2017, 1:29 pm by poiuy_qwert.

Yes, select any group of AI, delete it, and then hit ctrl+Z a few times.

These instructions are either not detailed enough (what is a "group of AI"?), or do not reproduce the issue consistently enough. I can't figure out the issue if I can't reproduce it. You must give me as much details as possible, every little thing from launching the program to the issue, so that it can be reproduced 100% of the time. It should be something like this completely fake example:
1) Open default files
2) Open the script XXXX
3) Find the first "wait"
4) Delete "it(" from the "wait(1)"
5) Add blah before the "wa1)" so its "blahwa1)"
6) Ctrl+z two times
7) I'm left with is "1)it(" instead of "wait(1)"

Open 'Brood Wars Terran 5 - Town A' in this script
Click on line 9 - 'build(1, Protoss Nexus, 150)'
Write 'xd' after 'Nexus' and 'adadad' before 'Protoss'
Ctrl+z a few times
You are left with 'build(1, adadarotos xd, 150)'
If you continue to ctrl+z, the entire script is erased.

Editing these files with PyDAT is resulting in bizarre changes to the flingy.dat, where Scourge's flingy entry is being set to seemingly random entries. First it was Rhynadon, then it was Power Generator. I've hit 'reload entry' to fix it but there's no telling how long before it happens again. Worth noting is that those two graphic entries are ones that I've replaced in the mpq/exe, but I don't know if that helps figure out the issue.

I am able to reproduce the issue, I'll see what I can do.

This one is hard to reproduce. I seem to remember fixing something like this that Neiv reported... I have an idea that should get rid of these issues for good.

Thanks!

I just encountered an issue with PyTBL where trying to use the search function hard freezes the program with no error message. Where would I go to get diagnostics to send to you?

On another note, that PyAI crash happened again. At first I thought it was related to defining a unit that didn't exist in unitdef.txt (can we get an update to that to include all hero units, special buildings, etc.?) but it happened occasionally even with out doing that. The files were nearly identical to what you were using for the ctrl+z test, though obviously they would have been different if it had saved successfully.

This should be fixed now.

If no error dialog appears and there is nothing in the Logs folder, then the only thing you can do is try and reproduce the issue and/or give me detailed steps to reproduce it myself.

If you can give me an aiscript.bin, bwscript.bin, and a script.txt which will reproduce the issue everytime I import it into them, I'll be able to look into the issue. Until I can reproduce it I'll poke around but i'm just fumblng in the dark and probably wont be able to figure it out
Also, was your bwscript.bin corrupted again this time or was it left alone?

Thanks!

Post has been edited 1 time(s), last time on Jul 19 2017, 4:16 am by poiuy_qwert.

bwscript.bin was corrupted, so I imagine it wouldn't have been too useful. I'm not sure what a feasible testing operation would be since it doesn't seem consistent with any specific action I'm doing, but in case it happens again I'll try to keep backups and remember exactly what I did last, saving after each script change.

It was corrupted again? Are you updated to at least PyMS 1.3.6, the file corruption should be fixed on that version+, if not i'll have to look into that again too...

Yeah thats probably your best bet. Maybe I can make a special version for you that will backup those things automatically on a crash, I'll look into that.

It says PyAI v2.5.1 on my end, the version I've been using all day (and the version it crashed with). It never prompted me for an update so I assume this is the latest version?

Yeah that should have had the fix I'll look at it again.

If I encounter it again I will let you know and do my best to remember to save at every turn as mentioned before. Thanks for looking into it again. This program is why I'm able to make incredibly complex AI.

I can not reproduce the file corruption. I have force it to crash during save, even in the same way as your original crash log, and the atomic writing works in keeping the original file safe. Are you sure the second crash was the same as the first one?

I didn't cross refernece the crashdumps but it also hasn't occurred yet with the latest version of PyAI, so maybe it has in fact been fixed and I just didn't update before.

How about adding a "Open SC Tileset" button into PyTILE? What this button would do is open vanilla/bw tilesets, it should list them to you. This would make it far much simpler than having to extract the tilesets yourself. Either load from MPQ or have them come along wih PyMS

Internal error (no crash) in PyTILE when attempting to add these bitmaps to this tileset. I tried to do it one at a time, and after the error popped up, hitting the magnifying glass icon to look for the bitmaps again did nothing.

Sounds like a good idea, I'll look into it.

Hmm, not sure how that bug snuck in there. I'll fix it when I get home, if you want to you can fix it yourself until then by removing ", graphics" from this line so it is just "def select_path(self):"