Print

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

[Cobbler89]

Have you switched to c++11/14? How much did it make a coding easier?

I don't know about the 14 one, but if I remember, 11 had the auto keyword, which I thought was very useful.  I didn't find anything to do with the other 11 features that I recall, like the lambda stuff.

C++14 mostly works a few kinks out of C++11 -- little fixes and enhancements to the stuff that didn't even exist before. If you like auto, one thing of note is that in C++14 auto works for function return types (not to be confused with writing auto before the function name in order to write the return type after, which was already in C++11 for similar yet different reasons). I also recommend, even in C++11, becoming passingly familiar with decltype, which can be used to get auto-like effects (of tying a type to a variable rather than repeating the type explicitly) in some situations where auto wouldn't work.

I'd also recommend a familiarity with at least the concept of smart pointers (which were added in C++11), for the same reason I'd recommend a familiarity with at least the concept of container classes -- they both make it easier to avoid a lot of the dangers of pointers and dynamic memory. In particular, the shared_ptr + weak_ptr combo can help avoid dangling references (although you have to code the weak_ptr uses for the case where the thing they refer to goes away, but they make that pretty easy).

That's my two cents' worth on the nearly universally applicable, anyway; make of it what you will. ;^)

As for lambdas, as far as I can tell they're mostly to help support functional programming composability styles -- sort of thing you'd find useful to no end if you came from a Haskell background, but it's an entirely different mindset with its own pros and cons...

...Unicode...

Switching gears, I wonder whether an issue with the whole text tileset limit has less to do with a specific character set and more to do with the fact that it's mapped to a specific character set instead of either mapping to specific tiles (more like the graphics pack method, but for the things that are currently character-based tiles) or else to tiles 1 through n in a tileset of at least n tiles, either way requiring (for good user experience; I doubt it could be program-validated) that the tileset provided make sense with/match up to whatever is mapped to them in the raws and init files. Unicode, of course, would provide a way to achieve the latter sort of mapping, with the added bonus that if the raws/inits were written in Unicode the tiles used could still be specified as letters/symbols, but the disadvantage that Unicode has a lot of extra complication because it's designed for universal text handling, not game spritesheets. Also just my two cents' worth, feel free to take it or to leave it. ;^)

[Mel_Vixen]

You made toady go "he he he" 4 times, you monsters! Dont you know that each time he does that a dorf goes into a morbbid mood? For the safety of our Mining and metalworking brothers dont give that man any more ideas! XD


Thanks for all the answers Toady. You mentioned that we might get the occasional Human elf or goblin on our doorsteps, will your fort be a possible destination for refugees?

Also hooray for fleshed out animalpeople.

[Astarch]

Have you switched to c++11/14? How much did it make a coding easier?

I don't know about the 14 one, but if I remember, 11 had the auto keyword, which I thought was very useful.  I didn't find anything to do with the other 11 features that I recall, like the lambda stuff.

14 was mostly minor cleanup, changes to constexpr etc...

The simplest way lambdas are useful is that you can use the functions in <algorithms> without needing to add tons of boilerplate.

c++11 has a few other features that don't really necessitate a massive rewrite, the explicit nullptr to replace NULL, range based for loops:

Code: [Select]

for (auto& thing : someContainer){foo(thing);}static assert if you do any metaprogramming, technically not needing to put a space between each > when nesting templates is a c++11 feature that was added really early. Stuff that is nice but might require rewriting old code is move constructors, smart pointers, and initializer lists.

Have you switched to c++11/14? How much did it make a coding easier?

I don't know about the 14 one, but if I remember, 11 had the auto keyword, which I thought was very useful.  I didn't find anything to do with the other 11 features that I recall, like the lambda stuff.

C++14 mostly works a few kinks out of C++11 -- little fixes and enhancements to the stuff that didn't even exist before. If you like auto, one thing of note is that in C++14 auto works for function return types (not to be confused with writing auto before the function name in order to write the return type after, which was already in C++11 for similar yet different reasons). I also recommend, even in C++11, becoming passingly familiar with decltype, which can be used to get auto-like effects (of tying a type to a variable rather than repeating the type explicitly) in some situations where auto wouldn't work.

C++11 auto/decltype function signatures are a huge pain in the ass and probably aren't worth it if you're not doing a lot of generic programming, which I suspect is not going on in DF. C++14 auto function signatures are nice, but I don't know if Toady is using new enough compilers for them.

As for lambdas, as far as I can tell they're mostly to help support functional programming composability styles -- sort of thing you'd find useful to no end if you came from a Haskell background, but it's an entirely different mindset with its own pros and cons...

Toady's an algebraist isn't he? I bet he'd enjoy Haskell.

[Putnam]

How do invaders find their way to a fort now, do they have to path over other sites to get to the fort?

They walk over the land on a path avoiding water

Does this mean that they avoid rivers? If so, will they simply give up if the fort is surrounded by rivers, or will they just go right over?

[Cruxador]

Are temple effects on immigration on the table for the tavern release? What about pilgrimages? By immigration effects, I mean things like getting more worshippers of a certain god if you have a big temple to it, or more devout worshipers if you have awesome temples in general. It seems like it would be more crucial for start scenarios where making a temple to a god is your primary goal, but on the other hand it could be a relatively low-hanging fruit. Pilgrimages on the other hand, fit quite perfectly with the "temporary visitor" idea being dealt with in the tavern release, but if they're to be relic-oriented like medieval Christian pilgrimages, it might make more sense to wait until some more artifact stuff is done.

[Majoras123]

Quote from: Pseudopuppet

Does the new entertainment stuff apply to adventurers? Will I be able to play as a bard soon?

Yep, next release you should be able to make some sort of living that way, to the extent that means anything.  Failure should also be possible.  They can't throw you in prison yet, and they don't have apples sitting around to throw at you, but maybe there'll be some way they can express disapproval when you offend them.

Will there ever be magical songs to control your enemies and bolster your allies ala D&D Bards?

[smjjames]

Will there ever be magical songs to control your enemies and bolster your allies ala D&D Bards?

You can have music to bolster or give courage to your allies without magic even being involved. As for the controlling part, I guess some psychological intimidation might be involved, but really, the magic is really far off.

A thought of my own:
With shrines (I guess those would be small temples), will ancestor worship be in?

[Vivalas]

I think some of this was in your transcripted DF Talk, but if it's not to early to ask, I want to clarify a bit on it..


Will military squad orders be updated along with the new justice system. For example, will the kill order be replaced with an attack order, where you can set the terms of engagement for your troops. The options would be something like, non-lethal, lethal, and no-quarter, so that you could target your own dwarfs without loyalty cascades and maybe attempt to capture invaders like goblins or kobolds. I was thinking being able to quell riots and beat dwarves would be fun along the lines of running a totalitarian fort.

Will there be prison zones, so you could have rooms designated for prisoners, like holding cells and cells. Would this be considered as part of the prison colony scenario? Will we be able to shackle, torture, and restrain prisoners based on civ ethics?

This ties into the justice system, but if it's too early to answer then okay

[Cobbler89]

Have you switched to c++11/14? How much did it make a coding easier?

I don't know about the 14 one, but if I remember, 11 had the auto keyword, which I thought was very useful.  I didn't find anything to do with the other 11 features that I recall, like the lambda stuff.

As for lambdas, as far as I can tell they're mostly to help support functional programming composability styles -- sort of thing you'd find useful to no end if you came from a Haskell background, but it's an entirely different mindset with its own pros and cons...

Toady's an algebraist isn't he? I bet he'd enjoy Haskell.

Now that you mention it, that's quite possible. I was thinking more about programming in general than about Toady's background in mathematics.

[Paaaad]

Will the new temples in fort mode tie into desecration and curses (in the initial release)?

With the addition of temples will the "defile the temple" night-creature creating event now be able to happen in our own fortresses?

He he he.

...I'm going to take that as a 'Yes'.

[LHLF]

Thanks, Toady! The new release sounds awesome 

I don't know if this has been asked, but won't there be a special skill for creating instruments? Making an instrument is really complex, not something that any craftsman can just do. Besides, we have special labors and workshops just for making soap and ash, so why not? hahaha

[Knight Otu]

Thanks, Toady!

A thought of my own:
With shrines (I guess those would be small temples), will ancestor worship be in?

I doubt it. Given the myth generator Toady mentioned, it's possible that it comes in with the Artifacts section or the Starting Scenarios section, though.

Are temple effects on immigration on the table for the tavern release? What about pilgrimages?

As you mentioned, it would almost certainly be part of the Starting Scenarios. For the tavern releases... I guess it depends on how flexible the code is for attracting tavern visitors. It could well be the case that Toady just needs to add a few things for it to work.

How do invaders find their way to a fort now, do they have to path over other sites to get to the fort?

They walk over the land on a path avoiding water

Does this mean that they avoid rivers? If so, will they simply give up if the fort is surrounded by rivers, or will they just go right over?

I think Toady means oceans and lakes here - the wandering groups in adventure mode don't particularly care about rivers that I've seen, and from my understanding, they follow the same general code.

[Urist Tilaturist]

Eventually there should be boats, pontoons and river fording. Crossing water is harder than crossing land, but it is in no way insurmountable.

[mifki]

Thx for the quick answer.

However that link is already more than four years old. Many new features, many new items (like all the tree tiles) were added in the meantime. "Hitting the wall" as Toady says appears due to that for sure closer than back then. Any change in that attitude? Also I am not referring to graphics, which for sure will cause major and long time dedicated work. Just extending it from 256 tiles to some more. Additional tiles could be derived from e.g. UNICODE. I am no specialist regarding copyright. But as most of the UNICODE-tiles are from various recent/historic alphabets, I can't believe they are more restricted than ASCII.

Interesting thing about moving to Unicode is that it would give access to a lot of different characters that can perfectly fit the game, and for free. While still staying in text mode for quicker development without all the complications of graphical tiles that Toady mentioned.

[Knight Otu]

Thx for the quick answer.

However that link is already more than four years old. Many new features, many new items (like all the tree tiles) were added in the meantime. "Hitting the wall" as Toady says appears due to that for sure closer than back then. Any change in that attitude? Also I am not referring to graphics, which for sure will cause major and long time dedicated work. Just extending it from 256 tiles to some more. Additional tiles could be derived from e.g. UNICODE. I am no specialist regarding copyright. But as most of the UNICODE-tiles are from various recent/historic alphabets, I can't believe they are more restricted than ASCII.

Interesting thing about moving to Unicode is that it would give access to a lot of different characters that can perfectly fit the game, and for free. While still staying in text mode for quicker development without all the complications of graphical tiles that Toady mentioned.

Of course, the relevant display tiles would still need to be created (not all of them can be used, of course). Then again, if you go Unicode, you can essentially go freeform. Got a few hundred tiles the game'll never use on its own, such as control characters that would just be blanks? Characters that would be too large or detailed to fit the default resolution? Pack 'em with game-relevant tiles instead.