Latest News Article

[Marko413]

[quote name='MegaTrain' timestamp='1360256964' post='20785258']
No, that's not what that means. It has to do with mods/hacks that want to make "falling sand" look like other blocks (lava flying out of a volcano, for example). This shouldn't impact vanilla minecraft at all.

Some more detail on what that means in this thread:
[url="http://www.minecraftforum.net/topic/1679605-snapshot-13w06a-is-out-apparently-just-bug-fixes/"]http://www.minecraft...just-bug-fixes/[/url]

including an awesome example of how that can be used:

[media]http://www.youtube.com/watch?v=ZOAx7NsivYY[/media]
[/quote]

Ahh okay, I get it.  Thanks everyone for clarifying that

[Julian_GameMaker]

Quote

Baby mob collision boxes are too large.

I'm glad this was fixed, now I can make a more effective cow farm (make it so the young run out of the pin so I don't mix the cows up)

[updraded]

why wasn't the server jar posted?

[Sambaru]

Nice.
I'm glad they are fixing bugs, I hope they do at least a few weeks more of fixes, so 1.5 gets released March 1st.

[MatuliYoshi]

Yeah, finally village fix!
Well done.

[RobinHood999]

Does this mean that the comparator thing is back?

[Blumi18]

Cant wait to try the new snapshot!!

[citricsquid]

lfmxgg368, on 07 February 2013 - 05:02 PM, said:

And the http://assets.minecr...raft_server.jar link was not posted

My bad, the widget has been updated to include all the relevant Snapshot links.

updraded, on 07 February 2013 - 06:29 PM, said:

why wasn't the server jar posted?

My mistake, the widget has been updated to display all the relevant links.

[Vagossssssssss]

If you have a problem with the download you can always go to the mojang webstie

[RingMaster123456]

i think it will be out in 10 days!!

[Darrok]

So the Scoreboard is actually part of the new game mode "Capture the Flag"? http://www.minecraft...apture_the_Flag

[5thHorseman]

MatuliYoshi, on 07 February 2013 - 06:32 PM, said:

Yeah, finally village fix!
Well done.

I'm very excited about this one in particular. Finally, my villagers will abandon that one single house and actually use the entire village!

(Or at least most of it)

[koopachris]

So they added...a bunch of fixes to semi-rare bugs. Ok, that's still nice.
But what features did they add? Oh...a new feature that can't even be used
in vanilla Minecraft...yay...

[murp11]

koopachris, on 07 February 2013 - 07:20 PM, said:

So they added...a bunch of fixes to semi-rare bugs. Ok, that's still nice.
But what features did they add? Oh...a new feature that can't even be used
in vanilla Minecraft...yay...

they won't add any new features. They remove all bugs to get ready with the 1.5 the nect features will come for 1.6

[Azurewhitewolf]

PonyDude186, on 07 February 2013 - 05:44 PM, said:

Yay! they brought back the comparator rapid fire feature!

No they did not.  I just tested that theory.

[RobinHood999]

Azurewhitewolf, on 07 February 2013 - 07:33 PM, said:

No they did not.  I just tested that theory.

NO! Thats terrible! I at least hope they add some sort of BUD switch/Ticker block now.....

[JellyOpera]

FranEGL, on 07 February 2013 - 05:19 PM, said:

In my opinion they should keep fixing bugs and when the game is almost bug-free they should start adding more stuff.

Yeah, but they'd get bored.

Also, get that Curse Client BS out of my face.

[PonyDude186]

Azurewhitewolf, on 07 February 2013 - 07:33 PM, said:

No they did not.  I just tested that theory.

Are you sure, I tested it again and works fine. If the front torch of the comparator is lit, is doesn't do the rapid fire thing. If the front torch is off, then it will do the thing correctly. Try it again with this technique.

[MegaTrain]

5thHorseman, on 07 February 2013 - 07:20 PM, said:

I'm very excited about this one in particular. Finally, my villagers will abandon that one single house and actually use the entire village!

(Or at least most of it)

Huh? I don't think the "villagers crowd into a single house" issue was fixed, unless it wasn't in the original list. The only fixes that seemed relevant to villages/villagers were:

• Baby mob collision boxes are too large.
  
• Village lamp post floating torches
  
• Villager trade menu missing name

[5thHorseman]

MegaTrain, on 07 February 2013 - 08:25 PM, said:

Huh? I don't think the "villagers crowd into a single house" issue was fixed, unless it wasn't in the original list. The fixes relevant to villagers were:

It wasn't fixed, but it was substantially improved. From the site:

Quote

After trying multiple different changes in the code, I couldn't get it even near "good", but at least I managed to make the villagers spread more evenly around the village, and somewhat in the houses, too. There could still occasionally be 3-5 villagers in the same house, but with around 15 villagers, at least 5-8 houses were occupied each night. So, an improvement, but not fully fixed.
Also, if a house has more than one door, the house has a higher chance to get crowding, even with the changes. The other door(s) lure the villagers out away from the door they are supposed to restrict.
The biggest problem to solve this issue somewhat easily is the door restriction operation. Basically, it does not work. There is no point in trying to work out a well working algorithm that takes occupation into account when that occupation data is mostly useless.
The changes
1) Changed the villagers' "home area" to be based on the bounding box of village's doors (instead of mass center). This matches better the needed area of movement, though the current way has the minor benefit of bringing the villagers slightly closer together during daytime - sort of looks like "community behaviour".
2) Increased the radius factor from 0.6 to 0.9. The 0.6 for the current mass center -based solution does make sense as the most distant door from that mass center can indeed be quite far away, and larger factor could allow villagers to wander quite far "out" in the opposite direction. Unfortunately, that will also limit their chances to move to that distant house that is in the village. With bounding box center -based "home area", the radius can be higher without that problem.
3) Changed the way villagers choose which door they'd like to go to. In current code the distance is given unnecessarily large disadvantage weight (squared and in addition anything beyond 16 blocks was given an extra about x30 distance which in practice means fully ignored), and when near enough, it has no effect at all. I changed the distance weight to linear, multiplied by just 4, and added the restriction value (for whatever rare effect it has). Also, I made it keep 3 best choices and pick randomly one of them in the end. The randomness tries to diminish the effects of the remaining badly working things.
4) Removed any extra doors from multi-door houses. Not a nice change, and not absolutely needed.
1 and 2 together spread the villagers quite nicely around the village. This already decreased the crowding effect a lot, but local hotspots would still be a problem.
3 and 4 reduce the local hotspots a bit. Not that well, but at least better than the current code.
What would still be needed
The above changes are merely a minor fine-tuning to reduce the worst of the issue.
Full fix would require quite a total rethinking/design for the way a door or house is marked occupied. To me, it seems that it would be best to simply bind each villager to a single door. Easy when generating the village (the code can just bind the generated villager to the house/door it was generated for). When spawning a new villager or when a door gets invalid/destroyed, there would need to be new code that looks for least bound doors and picks one of those, to spread the villagers evenly.
Then, when villager wants to get inside, it would prefer its "home door" unless it is really far away (like more than 50 blocks away?), in which case it could just pick a random valid door nearby for that night.

EDIT:

I just tested this in my test world where I have a little spawn area and iron golem farm, and it works in that they don't crowd into one area any more. They do however crowd into *two* areas...