Zaralith
...
we could post challenges about starting a world, going to a spot, and then doing something like travel as far north as possible without bench or furnace. Or a make your castles here thing.

Good ideas, I've seen some challenges like that before, like the "404 challenge" (here). That one happens to be really fun, and it feels almost like playing nazi zombies

Cyprezz, you are aware that y and z are swapped in minecraft. Y is the height. Also, when you look at coordinates, the y is from your face, not feet.anyway, like the ideas, we could post challenges about starting a world, going to a spot, and then doing something like travel as far north as possible without bench or furnace. Or a make your castles here thing.

CyprezzThe idea of posting blank worlds has come up a few times in the past. Seeds are so easy to create and use now, should we have a section on the site dedicated to them? You'd post the seed along with at least the spawn point screenshot.

Absolutly Cyprezz P.S. thanks for the feedback on my Pigman

yessssssss please!

I got 0, 67 with seed "Testing". I'm going to make the Seed section. The above data, theories and from reading other threads, it all seems to indicate the spawn point is random but not completely. Having it based on 0,0 and then randomized within a range makes sense. Maybe the range is really a matter of how far sand is from 0,0. Maybe it finds sand within a range and then randomizes from there. Who knows. I guess Notch does

Anyway, I think we should add a Seed section. Here are some thoughts on it.
* Seeds can be added by anyone but won't be "owned" by anyone. I'll track and display stats on seeds but individual members won't earn stats from simply contributing a seed.
* When you add a seed, system will check to make sure it isn't already listed on the site. If it is, you can go to that seed page and contribute photos.
* Anyone can contribute photos to a Seed page. I'll probably limit it to 5-10 per member per seed. Photos will require you to input x,y,z coordinates. Having the actual coordinates will give me the option to develop advanced tools in the future. Such as a proximity search and an interactive Google map with pinned photos.
* Photos will be credited to you but won't affect your stats.
* Projects and servers can be associated to Seeds. For example, if you build something in a world generated using the "Glacier" seed, you can associate your project to it. This way, like how servers currently work, your project can show up on the Seed page. Coordinate fields will be available when you're associating a project to a seed.

That's all I got right now.

Well I'll try to get to that later today... I have a lot of work to be doing, and surprisingly enough, playing minecraft somehow means I won't get any work done if I start playing it. Weird, right?

Edit: Also, I'm going to make a table generator that allows the user to enter in info for the locations and times for when they spawn and will calculate all of the statistics (mean, median, mode(probably not), standard deviation, median abs. deviation, variance, etc.). As I write this, I'm coming up with ideas for having a standard format for people to record data and send them in so somebody (probably me, I don't mind) can process them (via script/program, I'm not doing this by hand). Give me a day or two, I'll work it out.

So i have an idea for the seed section. If there is a seed section, have screenshots taken with debug on (the one that shows your position) or just require the screenshot's X and Z coordinate. Did some testing of my own using the seed Testing and the name Testing. Every single time (3 of them with clock moved forward and hour each time) had me spawning in the same place. I thought it might be the folder/seed combination so i switched to using the seed Random with the same world name. Full range was little over 250 for x difference and almost 400 for Z. I am thinking that the spawn is randomly generated within +-200 of 0. After a quick check above i revise that to +- 500 from Hibbsi's Data. I generated a set of worlds back to back (6) within 5 minutes. They were all the same folder name as the seed except 1, which changed nothing. I rolled the system clock back the 2 hours i had pushed it forward and still nothing. I am thinking as i said above that the spawn is generated by time. gonna try setting clock back a minute and generating world with same time. Quick check of that yielded nothing. Spawn is different, though more samples are needed than two. Anyway, those are my theories. quick check against Hibbsi data shows no direct limit to spawn distance. Spawn location may be limited by a set straight line distance from center that it must be within. Anyway, make a world using the Testing seed and tell me if you get 0,0

Thanks Hibbsi! Your findings will determine the fate of the Seed Section. 450 blocks is a reasonable distance in Minecraft terms. How different were the times between test generations? Would you mind running the test and intentionally changing the date/time on your computer?

Ok, so here's how "pseudo-random" the spawning is for a small sample.

Test conducted on world generated with seed: Test


As far as I can tell, no matter when you create the new world, you'll spawn somewhat (relatively) close to the spawns of a world with the same seed. Yes, 450 blocks isn't what I would call "really close", but it's not really that far.

For each one of these, I went to <0,0,0> to make sure that all of these worlds were the same, and they were.

It might seem like you're nowhere near where you were the last time you used the seed, but check your coordinates and you'll see it's not too far

Some other time, I will see how the actual hash codes from the "seed" string that is input affect the worlds by using strings that have hash codes that are almost identical. To the lab I go

Yah, when I originally tested I spawned only a few blocks away from my first generation. I just tried the same seed again, ~1 hour later, and I have no idea where I am. Which... means you might be able to specify a time to set your computer to along with the seed but that's a step back into a 'pain-in-the-ass' feature / hack / unsupported mod type thing. Will have to test that idea though just to confirm.

Jeb talks a bit about on Mojang's blog:

http://mojang.com/2011/02/23/a-short-demystification-of-the-map-seed/

I bet you are absolutely right.

The spawn is probably an artifact of a random number seeded by system time, just like a world that is not seeded gets the seed from system time. does the spawn move farther, the later you create the world?

I'm trying to figure out how random it really is. It seems to be somehow related to the seed since many are able to walk to the same coordinates. If for the most part it's only a few chunks away, then a seed section makes sense.

I agree with the no seeds because the spawn is random inside the world, even if it is the same world. Either that or have a single featured seed with a find this landmark thing and then change it when enough people have found it.

Unfortunately the spawn location is not included in a Minecraft seed. I don't know how useful it is to create a directory of Minecraft seeds when you spawn randomly in the world. It makes showing off the spawn pointless and actually misleading. Hmm, wut do?

Kind of thought you were saying C-Section. Lawl.

Yeah, this is a great idea.

hmm... Probably a good idea. It would keep the forum thread from becoming 100 pages long and getting bumped all the time. Plus, you know someone is gonna post a seed in the Projects section and start a chain reaction. Just my thoughts