Print

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

[The Script]

Any chance of seeing 64-bit versions of DF in the near future?

[Lolfail0009]

Any chance of seeing 64-bit versions of DF in the near future?

From what I've read, noooooooooooooooooooooooooooooooooooooope.

[Footkerchief]

Any chance of seeing 64-bit versions of DF in the near future?

Timeline is uncertain as usual, but Toady has phrased it as a question of 'when', not 'if' (my emphasis):

Are there currently any plans to make Dwarf Fortress large-address-aware? A mod is currently available to do so, and the response has been largely positive in that it allows history to work for a significantly greater time. It would be great to see that make its way into default DF.

I have no idea what this would do both with my stuff and with some of my libraries.  I'm not eager to introduce a bunch of strange bugs.  When we go to 64 bits I'll have to deal with it.

[CaptainArchmage]

Any chance of seeing 64-bit versions of DF in the near future?

When the memory use gets so large 64-bit is essential to keep the game going, it will need to go in. Likely this memory barrier will be hit during future worldgens unless serious optimisations have taken place.

[Ramaraunt]

This was asked already, but can you give us an updated guestamate on the release date?

[Dwarf4Explosives]

Can trees can be made out of more than one material?

To clarify, can the center of a multi-tile plant consist of a different material than the outer parts?

[Knight Otu]

This was asked already, but can you give us an updated guestamate on the release date?

Every attempt at an estimate has gone wrong. There's no reason to think further estimates are closer to the truth.

Can trees can be made out of more than one material?

To clarify, can the center of a multi-tile plant consist of a different material than the outer parts?

Given the raws as they stand in the devlog, no. There is only one tag to apply the main tree material, so there is no bark yet, for example.

[Footkerchief]

But you can of course have different materials for the growths:

Code: [Select]

[PLANT:BEET] beta vulgaris
[...]
[USE_MATERIAL_TEMPLATE:LEAF:STRUCTURAL_PLANT_TEMPLATE]
[STATE_COLOR:ALL:GREEN]
[DISPLAY_COLOR:2:0:0]
[USE_MATERIAL_TEMPLATE:FLOWER:STRUCTURAL_PLANT_TEMPLATE]
[STATE_COLOR:ALL:GREEN]
[DISPLAY_COLOR:2:0:0]
[...]
[GROWTH:LEAVES]
[GROWTH_NAME:beet leaf:beet leaves]
[GROWTH_ITEM:PLANT_GROWTH:NONE:LOCAL_PLANT_MAT:LEAF]
[GROWTH_DENSITY:1000]
[GROWTH_PRINT:0:7:2:0:0:ALL:1]
[GROWTH:FLOWERS]
[GROWTH_NAME:beet flower spike:STP]
[GROWTH_ITEM:PLANT_GROWTH:NONE:LOCAL_PLANT_MAT:FLOWER]
[GROWTH_DENSITY:1000]
[GROWTH_TIMING:60000:119999]
[GROWTH_PRINT:5:5:2:0:0:60000:119999:2]


[Parisbre56]

Can trees can be made out of more than one material?

To clarify, can the center of a multi-tile plant consist of a different material than the outer parts?

I'm not sure whether or not it can be done, but if it can, it is probably set somewhere in those templates:

Code: [Select]

[USE_MATERIAL_TEMPLATE:STRUCTURAL:STRUCTURAL_PLANT_TEMPLATE] [BASIC_MAT:LOCAL_PLANT_MAT:STRUCTURAL] [USE_MATERIAL_TEMPLATE:WOOD:WOOD_TEMPLATE]Better green your question so that Toady sees it.

[Snaake]

But you can of course have different materials for the growths:
...

Which got me thinking, it might be possible to add e.g. cork in a bit of a roundabout way, by adding "bark clumps" as either "fruit" or maybe some other type of growth. Then it could either be harvestable or it could have the GROWTH_DROPS_OFF(_NO_CLOUD) tag.

[wierd]

pedanticism:

LAA is not the same thing as 64bit.

Spoiler (click to show/hide)

LAA means that the application can use a full 32bit memory address space all to itself (4gb), when normally in 32bit world it would have to share a 32bit space with the OS and pals. (so, usually around 2gb, with another 2gb reserved for virtual memory address pools (global), and various memory address based operations/services/drivers.)

64bit native means that the application can (theoretically anyway) allocate memory from the global 64bit memory space. (16TB max, iirc.)  Currently, no cpus on the (consumer) market are able to address the FULL 64bit memory space, at least none that I am aware of. (The RAM controllers in the CPUs just arent up to it right now, and rightly so-- consumer boards would need insane numbers of DIMM slots to have 16TB of RAM installed, using current density modules.) However, the CPU should be able to address the extended area with virtual memory, if it really came down to it. On something like Linux, one could get some REAL enhancements with 64bit versions of DF and ZRAM swap. (Then DF would technically have had parts of it pushed into "swap", but still also be in RAM, just compressed with zlib. DF's memory structures are highly repetitious, so they should compress REALLY well), allowing more to fit in the same physical ram.)

64bit native DF on linux would rock something fierce.

[Inarius]

This was asked already, but can you give us an updated guestamate on the release date?

The reply is "no". It will be released when all the things will be implemented + bug resolution.

But it's probably before the 1th of July. Unless it's not.

[CaptainArchmage]

How fast is world generation right now? Is it slower or faster than in 0.31.11? Talking about “time to generate 1000 years of history on a medium map”.

I’ve noticed that on small maps (the step below medium), when you generate long histories the place gets cluttered out with tombs and necromancer towers. I tried to generate 10,000 years of history on a small map, and after 1500 years the map was getting so completely cluttered out with towers and tombs I aborted. It would be great to have longer histories possible once more - at least to get the nice 1050 years default history - but really to do so, the worldgen needs to run faster.

[Putnam]

This was asked already, but can you give us an updated guestamate on the release date?

The reply is "no". It will be released when all the things will be implemented + bug resolution.

But it's probably before the 1th of July. Unless it's not.

I give it 15th of April (SUPER OPTIMISTIC) to 1st of September (FLOODING HAPPENED OH CRAP)

[Valtam]

This was asked already, but can you give us an updated guestamate on the release date?

The reply is "no". It will be released when all the things will be implemented + bug resolution.

But it's probably before the 1th of July. Unless it's not.

I give it 15th of April (SUPER OPTIMISTIC) to 1st of September (FLOODING HAPPENED OH CRAP)

15th of April happens to be next tuesday, so that's slightly impossible even under high optimism standards. I would stretch my aim to the 26th of April, just because. Everything under the 1st of September sounds all right.