1.8 "The Bountiful Update" Released

[Bigpak]

Ugh this got even more scary to be honest, now cauldron is discontinued atm until conflicts are resolved. I really wish this didn't happen

What do you guys think is going to happen? Will we still be able to have bukkit/cauldron? Will this conflict ever be resolved in a way that isn't going to completely screw everyone over?

 

[Densely]

gimme 1.8 FTBATLTECHNIC COMBOPACKSNAOW! in all seriousness, i don't see a difference between 1.7 perfromane and 1.8 (1.8 does fix that chunk thing)

 

[Omicron]

Yes, but try performing the exact same test on a computer with worse specs, and it will go down into the single digits insanely early on.

Of course. I did disclaim that I am testing on a high-end machine and that my numbers aren't representative of everyone else's. The point was not to be representative, but rather to compare performance of all three Minecraft versions in the exact same scenario on the exact same map on the exact same computer. This tells us something about the differences between the software versions themselves, independent (to some degree) of the hardware they're being run on. The important numbers are really the 16 block view distances, as those were the ones that are present in all versions (including 1.6.4, despite the fact that a bug actually prevents it from happening).

And honestly, I see that 32 block view distance as more of a "future proofing". It's not really, honestly playable for the largest group of players just yet (though it does make for impressive screenshots). Perhaps Mojang has a plan for moving further aspects of computation out of the main world thread, or for reducing the performance cost of view distance further in the future. Anything they can do to lessen the single core CPU bottleneck will significantly help framerates here. If fps can get up to perhaps 60 on a machine like mine, then a large section of players will be able to get somewhat playable framerates too for view distances of 24+. I recommend trying 24, incidentally. I got 55 fps there, and it represents a 50% improvement over the previous maximum.

All with the caveat, of course, that this was pure vanilla. Modded Minecraft tends to be significantly more demanding on the hardware. But hey, maybe if a player will ultimately be able to go from 12 to 16 in 1.8, that's an improvement too.

 

[Hambeau]

gimme 1.8 FTBATLTECHNIC COMBOPACKSNAOW! in all seriousness, i don't see a difference between 1.7 perfromane and 1.8 (1.8 does fix that chunk thing)

First 1.8 Mod?

Fluff, but it still counts...

 

[midi_sec]

First 1.8 Mod?

chatlog. always chatlog.

 

[asb3pe]

Ugh this got even more scary to be honest, now cauldron is discontinued atm until conflicts are resolved. I really wish this didn't happen

What do you guys think is going to happen? Will we still be able to have bukkit/cauldron? Will this conflict ever be resolved in a way that isn't going to completely screw everyone over?

Just read the Bukkit board, all the contributors seem to have realized its over for good and they're all saying their goodbyes. It's pretty sad actually, and I have nothing to do whatsoever with servers or Bukkit! Altho of course I did play on random servers from time to time.

If you want to read: http://forums.bukkit.org/threads/an-independent-goodbye.310086/

As usual, I found it via reddit... comments are often worth browsing: http://www.reddit.com/r/Minecraft/comments/2fo4ud/bukkit_an_independent_goodbye/

Gotta add one more - I don't think any overview of the server plugin situation would be complete without reading this reddit/twitter as well, from Lex Manos (Forge)...
http://www.reddit.com/r/feedthebeast/comments/2fn3dk/forge_18_to_support_vanilla_clients_and_will/

Keep the faith!

 

[midi_sec]

Keep the faith!

That feeling when you realize that everybody involved in a particular situation is acting as irrational and childish as the other...kinda like that. The one, and only thing I have faith in regarding this situation is that just when you think it's over, It's not.

 

[midi_sec]

In reading that thread, the many before that, and various tweets... If I wasn't laughing right now, I'd probably cry.

I find it humorous that everywhere I look people have this idea that mojang's acquisition of bukkit was some shady back room deal. Actually, In contributor EdGruberman's goodbye post he references knowledge of Mojang's ownership long before any of these events were forseeable. Also, I think this person's goodbye post is as accurate and "real" of a reason for these events transpiring as you're going to find.

Anything else is just this:

As for Wolv, he's not some kind of copyleft activist, all he's done is present an even more unpalatable solution to Mojang than merely letting the Bukkit project die. They want Bukkit dead, and for it to be painful to Mojang to pick back up, nothing more. Sucks for us, good for...somebody

 

[ShneekeyTheLost]

So develop Forge-based admin tools to replace bukkit?

Honestly, never really saw the point in Bukkit, other than to make modder's lives more difficult by having to write a whole new mod just to be compatible with Bukkit. Back in the old days of ML+MLMP, you had to do that if you wanted your mod multiplayer. There's a reason Forge Modloader got so popular with the modders and all-but-replaced ML+MLMP.

 

[midi_sec]

So develop Forge-based admin tools to replace bukkit?

Well... Sort of. I don't like this solution though, personally.

https://twitter.com/LexManos/status/508198938629771264

However, I have announced. 1) 1.8 Forge will support vanilla clients {details TBD} 2) If Bukkit dies, Forge will get a YUAL project

 

[DZCreeper]

Insert excellent performance tests here.

Your numbers solidly back up my belief that Minecraft heavily favors Nvidia graphics cards somehow. On my old machine that I now use for media center computer with a GTX 560Ti I am seeing about the same FPS as you for the 1.8 tests. It makes sense because the 560Ti is still a powerful card. The major discrepancy is that the processor is a Phenom II x4 965 at 4.0Ghz. Your Intel chip should be blowing my old rig out of the water if performance is indeed processor bound. If you would be willing to send me the same world you used for testing, I will post numbers.

 

[Bigpak]

I myself have been backing up my servers (Local servers that I only play on with my cousin) and attempting to make sure I have everything I need to just be able to play on 1.6.4 while everything works out as I am of no use to anyone.

I honestly just hope this all works out and everything is alright. I am quite interested to see what will happen with the dmca and mojang however.

 

[mcalpha]

So develop Forge-based admin tools to replace bukkit?

Something like this. Sponge will be a plugin API on top of Forge that provides a similar version-independent abstraction as Bukkit did, and vanilla clients can connect to servers using it.

http://spongepowered.org/ (click link to read FAQ)
https://github.com/SpongePowered/Sponge (looks like cpw is interested in making it succeed)

 

[Omicron]

Your numbers solidly back up my belief that Minecraft heavily favors Nvidia graphics cards somehow. On my old machine that I now use for media center computer with a GTX 560Ti I am seeing about the same FPS as you for the 1.8 tests. It makes sense because the 560Ti is still a powerful card. The major discrepancy is that the processor is a Phenom II x4 965 at 4.0Ghz. Your Intel chip should be blowing my old rig out of the water if performance is indeed processor bound. If you would be willing to send me the same world you used for testing, I will post numbers.

I'm not sure a different video card would have created much different results here, considering my Radeon was half asleep while testing. It's possible that Minecraft favors the Nvidia driver, causing less CPU overhead, but I'm not sure that would even run in the same thread. Perhaps Minecraft simply favors the Phenom's CPU architecture, or is written in a way that makes it nearly architecture-agnostic. I'm not going to call "badly optimized", but perhaps it simply gets similar performance clock for clock on most CPUs, while being unable to profit from most architecture-specific optimizations. In that case, there would be little performance difference between a fixed 4 GHz Phenom and a 3.4 GHz Haswell that turbos up to somewhere between 4 and 4.2 GHz depending on current load and temperature. Similar clocks, similar fps.

EDIT: Keep in mind that java version and windowed/fullscreen mode differences and game resolution may also affect performance.

I'm going to send you a world download, but you'll have to be patient until tonight. I'm currently at work and won't be done for quite a while yet.

 

[midi_sec]

I'm going to send you a world download, but you'll have to be patient until tonight. I'm currently at work and won't be done for quite a while yet.

I'm an old school overclocker, and from my seat of the pants experiments and experience I can say that some architectures are just plain better at performing certain calculations than others. This can be seen all over on benchmarks; there's the stand out test results that just doesn't quite fit with the rest of the data from previous benchmarks.

I'm very interested to see the results, but for now I sleep.

 

[Golrith]

In my modded 1.7.10 world I'm getting between 40 to 120 FPS, even during world gen. But I do get splutters with large objects being generated, such as Better Dungeons dungeons and Natura Redwood trees. Since I'm used to playing up to 10 FPS on my old rig in earlier versions, this is amazing to me
What is really annoying at the moment are the random freezes when working with GUIs, which I think is an NEI issue. That's a bigger issue atm than any performance problems.