Print

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

[stoyang]

Are there any scripts or features in DFHack that you think would be good candidates to integrate into DF itself, should time allow?

[Criperum]

Rimworld which is basically DF for casuals, it has a mod that adds pretty decent multiplayer experience without issues.

This, you mean? (https://steamcommunity.com/sharedfiles/filedetails/?id=2606448745) Edit: trying to cut down on confusion caused by you leaving that one out, if so.

Yes, this one

[dikbutdagrate]

1. As I recall from the myth demos, deities and primordial forces are given spheres by looking at what objects with obvious spheres (like the sky, elemental planes, certain creatures, divine law, etc.) they created during mythgen. So if a deity created the ocean, they might become associated with OCEAN or maybe WATER. However, some spheres would seem required for a more cohesive myth narrative. Like the TRICKERY of a trickster god, or a deity of LAW not running around wrecking things. Especially if mythgen doesn't cease completely at the point where historygen starts (or the player is able to enter the world), but continues to run in parallel to history (albeit slow). At what point would these sphere-connections be put into place? At the end of (normal?) mythgen where historygen starts? When something with an obvious sphere connection is created?

2. Many cultures have considered mundane animals to have magical properties, such as scarabs or black cats. Can we expect any of this with mythgen, or will it rather be the separate fantastic beasts, like the unicorn, who are granted powers?

3. Any thoughs on sacred animals/plants and mythgen?

Ha, well somebody is gettin' antsy about the Myth & Magic release.

I'm sure Tarn will answer your questions to the best of his ability, but I seem to recall from interviews that Tarn is still pretty decidedly undecided, even now, with how the variation of gods, spheres, and biomes are still all going to tie into one another come the Myth & Magic release. 

I imagine he'll attempt to kick that can down the road for as long as possible, at least until enough time frees up so he can meditate on it, you know, like under a waterfall for 4 months, until he settles on something.

_{edit: Brevity}

[Buttery_Mess]

With all the changes to UI, has it made it technically easier to translate Dwarf Fortress to other languages? Are we likely to see localisation of Dwarf Fortress, perhaps by KitFox or another third party? Is it technically feasible to abstract the interface to the point where fans could do the translations themselves? If not yet, is it possible or likely, perhaps something Putnam could take the lead on?

[A_Curious_Cat]

With all the changes to UI, has it made it technically easier to translate Dwarf Fortress to other languages? Are we likely to see localisation of Dwarf Fortress, perhaps by KitFox or another third party? Is it technically feasible to abstract the interface to the point where fans could do the translations themselves? If not yet, is it possible or likely, perhaps something Putnam could take the lead on?

I've asked a similar question before, and Toady gave an answer here.  Of course, this was before Putnam was hired, and I expect she'd be the one to implement it as I don't think Toady One would consider it worth his time to work on it himself (i.e. it's not a "must-have" core game feature), assuming it gets done.  I also suggest that you read the documentation for GNU Gettext (especially the parts about the "Programmer's view (Putnam,s and possibly Toady One's view)) in order to get an idea of what would be involved.

To answer a concern that Toady One and others have had about proceduraly generated text (particularly in mods), I would suggest pushing as much text out to the RAWs as possible.  Basically, the RAWs would contain the untranslated text that gets fed into the gettext() function.  In the case of the game, I would expect that third parties would do the work of translating once the basic framework was set up.  As for mods, mod author's or third parties would be responsible for translating mods.

Edit:

Also, the main reason I've been harping on about GNU gettext, is because it is (as far as I know) the only such solution that works on Windows, MacOS, and Linux.  There are other solutions, but (as far as I know) they only work on one or two of those operating systems.  Also, I would suggest that Toady One ask Kitfox's lawyers to go over the licensing before he committed to using it...

[lethosor]

Given that we are about to switch to sdl2 and at the same time sdl2 is going into maintenance mode, are there any wishlist items that you want to see from sdl3? And on the same line of questions, can we move to sdl3 more quickly once it is released?

Worth noting the difference between "maintenance mode" and "end of life". SDL2 will still minor releases for bugfixes from time to time. As far as I know, an end-of-life date hasn't been set yet. In contrast, SDL 1.2 has not seen any releases since 1.2.15 was released in 2012. (There actually are occasional changes made to it still, including as recently as last month (!), but you would have to compile SDL from source to use them.)

Also, the changes between SDL2 and SDL3 have a pretty minimal impact on DF. SDL2 significantly improved support for a lot of modern platforms (e.g. SDL 1.2 apps barely even run on new versions of macOS), and dropped support for old platforms/hardware (some that DF never would have run on). In contrast, SDL3 is more of a developer-focused release. It fixes a fair number of inconsistencies in the API, and adds some new features as well, but nothing that DF would immediately benefit from.

So while SDL3 might be nice to have, I would probably encourage Bay12 to hold off on upgrading just for the sake of upgrading. I think SDL2 will fit DF's needs for some time. Of course, that could change in the future. The upgrade path from SDL2 to SDL3 looks easier to me than from SDL 1.2 to SDL 2, and there are even some scripts provided by the SDL authors that can help with migrating to SDL 3, but it's still an upgrade that will take time and has some risk of breaking things.

[oninoshiko]

With the multi-threaded improvements to the line-of-sight code going live, is there any plans to look into threading other parts of the code? Pathing and stone temp come to mind for me.

Putnam has actually published some findings from profiling the game - this thread goes into some detail about how pathfinding isn't a major contributor to speedups. So even if it is relatively easier to multithread than other parts of the game, the payoff would be pretty minimal.

I hadn't looked at that thread in a while, and didn't realize the numbers where rechecked and found to be even less then initially thought. Still, heat-checks (#6) was the other thing I asked about. and re-reading it just, really, changes the question to strike pathing and replace it with relationship checks during watching performances. What is being threaded isn't really the question, so much as is there things in the works to make better use of this 8-core/16-thread processor that I'm not sure I would have believed could be bought for a non-governmental budget when I started using computers.

In any event, it was worth reviewing that thread, there was a lot of new information in it since I last looked at it.

[Fell]

Saves can be very large so I don't think it can be a default option given how long it would take to sync saves.  It might be possible to make it optional but the fast way of setting up a sync folder doesn't seem to do that.  We're open to doing it but it might require a small workshop-like push to get it to function.

I'm a developer of We Are Football on Steam, our save files can get pretty large (> 100 MB) as well, and the game stores a history auto saves which can easily grow into gigabytes. Enabling Cloud Sync was pretty painless for us! It was just a Steamworks setting and required no changes to the game executable. The storage space can be configured by the developer and the theoretical maximum is 100 GB or 10000 files. It's pretty reliable.

I would just enable it and see how it goes.

[SeeSchloss]

With the Linux port compiling now, will we get an aarch64 build for Linux, or just aarch64 for macOS and x86_64 for Linux?

[Eusebio Ptolomeu]

I already asked something similar here, but I thought maybe it would be better to do so here. It's basically the same thing:

Is there a roadmap being followed?
or
What would be the best way to keep track of the development of Dwarf Fortress?

[Shonai_Dweller]

I already asked something similar here, but I thought maybe it would be better to do so here. It's basically the same thing:

Is there a roadmap being followed?
or
What would be the best way to keep track of the development of Dwarf Fortress?

The version number of the game indicates the percentage of features implemented from a list they have of "features for game to be considered 1.0".

Of course they go off-track at times, but the past few years have been pretty much following the plan. Last time things took a diversion as I recall was in implementing villainous networks, turned out the world wasn't varied enough for networks to work properly so guilds, religious strife, love, jealousy, envy, mercenary groups and various noble positions were added.

Oh, and then someone said "let's add graphics and release on Steam!" recently, that was quite a bit off-map (they were always there, but assumed to be much later on).

[Egan_BW]

What would be the best way to keep track of the development of Dwarf Fortress?

There is a dev log at http://www.bay12games.com/dwarves/ if you want somewhere you can check for updates.
And at http://www.bay12games.com/dwarves/dev.html is a list of planned features and whether they're implemented or not. Looks like it hasn't been updated for a couple years now, though, as the steam release was being worked on instead of the usual new features.

[Deno]

This may have been asked before but currently, Legendsviewer doesn't work with the steam version as some of the data required can't be exported. Are there plans to bring back the required exporting options?

[Clatch]

What news of reinstating keyboard controls? (I'm really struggling with the inefficiency of the mouse for large designations, plus the loss of macros)

Macros are in there in some form, though maybe they don't play nice with the designation cursor - I thought I tested that part.  In any case, there isn't specific news, but Putnam has been revamping interface stuff, including standardize some list widgets and such that should really help when we get there.

Toady mentioned something about playing nice with the designation cursor. Can someone explain a bit more about that?

I've spent too many years pounding the keyboard.    My wrist aches pretty badly messing with a mouse even for a few minutes.  Am I reading into this too much, or is there a way to control the cursor with the keyboard after initiating an event with an available macro?

[Putnam]

Given that we are about to switch to sdl2 and at the same time sdl2 is going into maintenance mode, are there any wishlist items that you want to see from sdl3? And on the same line of questions, can we move to sdl3 more quickly once it is released?

Moving on ASAP is planned on my part, what sort of features might be useful will remain to be seen--the game already has an entire action sets system made bespoke for it, for example

With the multi-threaded improvements to the line-of-sight code going live, is there any plans to look into threading other parts of the code? Pathing and stone temp come to mind for me.

Putnam has actually published some findings from profiling the game - this thread goes into some detail about how pathfinding isn't a major contributor to speedups. So even if it is relatively easier to multithread than other parts of the game, the payoff would be pretty minimal.

I hadn't looked at that thread in a while, and didn't realize the numbers where rechecked and found to be even less then initially thought. Still, heat-checks (#6) was the other thing I asked about. and re-reading it just, really, changes the question to strike pathing and replace it with relationship checks during watching performances. What is being threaded isn't really the question, so much as is there things in the works to make better use of this 8-core/16-thread processor that I'm not sure I would have believed could be bought for a non-governmental budget when I started using computers.

In any event, it was worth reviewing that thread, there was a lot of new information in it since I last looked at it.

I made those relationship checks 10x as fast by just introducing a little caching, turning the checks from O(logn) to amortized O(1)--multithreading is an O(1) speedup, so improvements in algorithms will win just about 100% of the time, and this is a great example of that