Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Clément]

DT uses creature_interaction_effect_body_transformationst.race which was removed in this commit. It is not something really important, it is only used when cheating to show curses and it should show "werebeast" without giving the race because of the missing offset (there is already a fallback for older versions). I may make a 0.47.03 release anyway if I cannot fix it soon.

[PatrikLundell]

You can find it through a bit of referencing:
- df.global.world.raws.syndromes.all [unit.syndromes.active

• .type] gives you the syndrome

- syndrome.ce

• ._type == df.creature_interaction_effect_body_transformationst is the one you want (weed out inebriation, etc).

- ce.race_str gives the creature_id string for the creature, and ce.caste_str give the string for the caste.
- Searching for the string (ugh!) should let you find the race.

Logically, there ought to be a numeric race id in the syndrome, but it if there is one it hasn't been located. Hacking these strings on my were caused it to transform into a weresquirrel anyway, but without the wound it should have gotten, so somewhere there is unmapped info on what the creature race is.

Edit: Note that the syndrome data is in the active vector even when the were is not transformed (I looked at my were when writing the above).

[hagger]

Experimental memory layouts for 0.47.03:

• Linux 32 bits
• Linux 64 bits
• OSX 32 bits
• OSX 64 bits
• Windows 32 bits
• Windows 64 bits

Not sure if anyone's chimed in on this, but seems fine for Linux 64 bit.

Can dig out my 32 bit hackbook to check linux 32 bit if you like

[PatrikLundell]

Some mapping detective work resulted in the discovery of where were numeric race and caste info was hidden in the syndrome, but the changes haven't been reviewed or approved yet.
Anyway, race is found in a vector that was previously labelled as a pointer, with caste sort of going into the next vector (also previously labelled as a pointer). "Sort of" means the vector is empty for weres, as they're all asexual with only the default caste. To get a non default caste (such as a male), that vector needs an element, presumably with the caste number (that's what I tried), but I guess you're only interested in the race anyway. While turned, the unit's caste and sex fields match the were one, as does the race (of course).
Note that you still have to go via the syndrome index, but don't have to do a string search.

[Clément]

Thanks, I've made the changes accordingly. But it is still not tested, since I don't have a werebeast. Is there a way to easily create one in Fortress mode?

[PatrikLundell]

Here's a dropbox link to my save https://www.dropbox.com/s/2vgq1cii71cjqz5/Newhope.zip?dl=0
 It's from 0.44.12 and I use it in 0.47.12 as a reference. The were is a caged elf found with f4 (and on the "other units" list, past the merchants. The were turns on the 17:th, so it's a bit of thumb twiddling to get there. A message about starvation stops the time at the 15:th, but there may be a mood in between.

[Clément]

New version released: 41.1.4

Changelog:

• added memory layouts for DF 0.47.03
• always read animal health
• fixed transformation syndromes races

Windows builds are also available on DFFD (win32, win64).

[Clément]

I also updated the alternative experimental appimage (built on Ubuntu Xenial with Qt 5.12). If no one complains this will be what I will use in the future.

[Clément]

Experimental memory layouts for 0.47.04:

• Linux 32 bits
• Linux 64 bits
• OSX 32 bits
• OSX 64 bits
• Windows 32 bits
• Windows 64 bits

[ZayZe]

Experimental memory layouts for 0.47.04:

• Linux 32 bits
• Linux 64 bits
• OSX 32 bits
• OSX 64 bits
• Windows 32 bits
• Windows 64 bits

Thanks for keeping this up to date over the years! A must have until the graphics update I'm sure... You made this game playable. Thank you.

[RoboLuddite]

Clément, I've logged in to the forums for the first time in years just to say you're a legend. I can't believe that you're able to pump these out within just a few hours of the release.

[clinodev]

Win64 looks good so far!

[Clément]

Clément, I've logged in to the forums for the first time in years just to say you're a legend. I can't believe that you're able to pump these out within just a few hours of the release.

It's mostly automated. BenLubar's bot is doing most of the work. I just git-pull the new structures and run some script that generate the memory layouts.

[jecowa]

Experimental memory layouts for 0.47.04:

• Linux 32 bits
• Linux 64 bits
• OSX 32 bits
• OSX 64 bits
• Windows 32 bits
• Windows 64 bits

Thank you for memory layouts! 0.47.04 works in osx64

[feelotraveller]

Three things:

1)  Thank you for the quick memory layouts.  Linux 64 bits have worked well for me so far.
2)  I believe that under default colours Options -> Happiness: Very Unhappy (red) and Miserable (orange) should be swapped.
3)  When I select the Canterell font for Tooltips the bold part of the tooltip gets very, um, fuzzy (like it needs dithering?).  It is my preferred font, so while most others don't show this problem - although D050000L seems to suffer worse and URW Bookman/URW Gothic also have non-clear bold text - I wonder if there is a fix possible, either on my part or from the other end?