Print

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

[NW_Kohaku]

... this probably belongs in a dedicated topic at this point...

Damn, I kind of went long, and made another giant rant.  I guess it's fine if I went and made a topic for the thing, then, though...

Well, it's this thread.

EDIT: Oh, crap, while writing it, I got ninja'd 4 times...

[Granite26]

I think if you're writing something that takes two hours

, Ninja isn't the term...  Maybe hared...


______

Result: pathfinding running n^2 times for n jobs, instead of n times.
[I assumed that number of jobs is liner vs number of dwarves, but "where n is quantify of dwarves" can be changed to "where n is quantify of jobs"]

You didn't even try to optimize.  You're checking every dwarf and every task, rather than being able to exclude some tasks based on criteria (say, "only tasks within 5 horizontal tiles of the A->B path").  You're trying to imply that this new functionality would be performed in the least efficient and most brute force way imaginable.

I don't think there's an optimization that'll make an order of magnitude difference.  Even culling on distance is still going to send you pathfinding through the fort.  Say what, you pathfind to where you're going, then look everywhere within X (ACTUAL) spaces of where you're going for an object that's to within X squares of the path further on?  Then you've got to check to make sure you can get there only going 5-10 steps out of your way?  (no walls, etc)  Then you've got to redo that every few steps, just to make sure your plan still works?

Yeah, ugly, even with optimizations.

You don't check through walls...
....You know what, never mind.
I'm not even going to try debating this with you, as clearly any solution is not good enough in your eyes.

I'm not saying there's not a good enough solution, just that saying that I can't think of any optimizations. 

How do you know it's through a wall until you try to pathfind to it?  I mean, step 1: draw a straight line, but even then, shouldn't they walk around a pillar to get there?  How far should they walk around?  That's your path distance (I used 5 - 10) that you have to check before deciding it's not close enough to on the way. 

Sure, you aren't checking the distance for every other (hauling) job, but you are checking the distance to the ideal path (trivial math), twice and then checking two length 10 paths.  You're doing this for (up to) all jobs on the path.

[NW_Kohaku]

I think if you're writing something that takes two hours
, Ninja isn't the term...  Maybe hared...

... I did a lot of thread searching and reading and linking and wrote a godawful amount of text, OK? 

I'm not saying there's not a good enough solution, just that saying that I can't think of any optimizations. 

How do you know it's through a wall until you try to pathfind to it?  I mean, step 1: draw a straight line, but even then, shouldn't they walk around a pillar to get there?  How far should they walk around?  That's your path distance (I used 5 - 10) that you have to check before deciding it's not close enough to on the way. 

Sure, you aren't checking the distance for every other (hauling) job, but you are checking the distance to the ideal path (trivial math), twice and then checking two length 10 paths.  You're doing this for (up to) all jobs on the path.

Didn't I just create a topic to keep this discussion out of this thread? 

I even mentioned this exact thing, and how much complexity it has.

[Draco18s]

You're pretending that you can't look at the map when trying to determine if something is "5 squares away."

Lets say this is our map.  * are the path, T is a Task Object.

.*......
.*......
.*..#...
.*..#T..
.*..#...
.*..#...
.*......
.*......

If we count diagonals as 1 space, then T is 5 away from the path (go around the wall on the top).  So we want this to be included, but we don't want to run an A* check on it (and by definition, every other task on the map).

What can we do?

Well, we do this: find every square adjacent to the path, and every square adjacent to that, etc. until we have 5 deep and note every task that falls inside that area.  Essentially we're doing a limited distance flood-fill.


1*12345.
1*12345.
1*12#45.
1*12#5..
1*12#5..
1*12#45.
1*12345.
1*12345.

Voila.  And if the object is in another room not direclty accessible, then the item isn't categorized:


1*12345.
1*12345.
1*12####
1*12#T..
1*12#...
1*12####
1*12345.
1*12345.

Now you have a list of every task that is inside some boundary distance of the dwarf's desired path.  Some tasks have a starting point and an ending point ("fetch quests" and for those you need to check the starting point and the ending point with the requirement that the start location is closer to the dwarf's starting location (and vice versa) which you can do through normal distance calculations (no need to do a path-find distance, as we already know that both points are within 5 of the existing path).

[Granite26]

I'm 80% convinced.

[NW_Kohaku]

Draco... seriously... thread just for this topic... no need to carry on the discussion in two different threads at the same time... please?

[Kogut]

Draco: I replied here: http://www.bay12forums.com/smf/index.php?topic=78413.msg2025177#msg2025177

[Puzzlemaker]

Has the eternal suggestions basically been dropped?

[Granite26]

I think the truth is closer to 'digested'.  There's enough stuff on it for 3-4 years of development.  Many of the items have been added to the dev schedule.

Anything new would go on the end of that list, and probably after another arc (Right now, the schedule is Caravan arc mixed with the vote stuff near the end or as it makes sense)

Given that, do you think there should be an evergreen document for people to say what they want a year or so from now?  Or is it smarter to let him work through the Caravan arc and the bug fixes, and take stock of where we are then?

[Kogut]

Has the eternal suggestions basically been dropped?

Yes - mainly due fact that DF has ideas for 50 developers for 20 years.

[SombreNote]

http://www.bay12forums.com/smf/index.php?topic=80121.0

I love all these ideas pertaining to monster generation, and monster behaviors, it is just that pretty much none of the ideas people have come up with will work until hospitals are efficient and effective, and training military is not such a pain.

I would love DF to be a game where people are able to get into more scraps, but as of now, it is just so deadly and irritating most of the time. "Urist McSusceptible Stubbed his toe and died of in infection". One big improvement would be to take a page out of other game systems, and make getting dropped easy, but dying of wounds hard. This would be kind of like going negative in D&D. It would serve the function of letting dwarves get beat down, but you would have the chance to nurse them back to health, sometime later. Of course there would be many beasts that are pretty much sure to kill you out right (e.g. demons, magma. etc.), but the vast majority of bad guys (e.g. gobs, trolls, and FB) might be perfectly content moving on when they knock out an enemy for good. Who knows why, it might just be because they want to keep the meat fresh before they eat it, and they might not have time for a while.

This system works really well in a lot of games, but as of now in DF when you are dropped by a bad guy, you are 99% dead. It makes sense for the dwarven genre to have dwarves take a really long time to die. And this will give feather emphasis in the functions of hospitals.

[Draco18s]

This is a thread about the voting system, not the voting system itself.

If you want that idea actually IN the voting system you're going to have to add it.  And in any case, this thread is not the place to discuss the merits of a suggestion.

[Silverionmox]

Has the eternal suggestions basically been dropped?

Yes - mainly due fact that DF has ideas for 50 developers for 20 years.

One would expect a selection mechanism to become more important then, not less.

[Kogut]

Has the eternal suggestions basically been dropped?

Yes - mainly due fact that DF has ideas for 50 developers for 20 years.

One would expect a selection mechanism to become more important then, not less.

We have one and this mechanism is called Toady ("Hm, I would like to add many types of grass").

[Khym Chanur]

Isn't "Message control" (#230) currently implemented in 0.30?

EDIT: Also, "Fix Cooking Exploit" (#225), "Allow cycling through site finder results" (#208), "Doctors/Hospitals" (#57).