Testing with two PPC machines. Mac mini -- 9200, and iBook g4 -- 9550. Both are 1.42 GHz, 1 gig memory. 9550 has 64 MB vram (smaller than the 128 normal minimum); 9200's vram I do not know.
12w30b bug reports
1. Attempting to open a LAN game, to test with my second PPC machine. Turned out that I did not change my username, so machine 2 logged in, same name; this caused the player on the main computer to be logged out, which stopped the server. Note that I was logged in from the other machine at the time the server stopped.
Not sure if this is wanted or not wanted.
2. Server does not properly identify/broadcast itself. Does not choose the interface correctly. Does not let you choose the interface.
INFO: Saving chunks for level '12w30b flat'/[email protected]
Started on 192.168.2.1:49187
+ Stopped mine30b
Kleiman-ibook:~ michael$ ifconfig en0 en1
ifconfig: en1: bad value
Kleiman-ibook:~ michael$ ifconfig en1
en1: flags=8863<UP,BROADCAST,SMART,RUNNING,SIMPLEX,MULTICAST> mtu 1500
inet 192.168.0.30 netmask 0xffffff00 broadcast 192.168.0.255
media: autoselect status: active
supported media: autoselect
Kleiman-ibook:~ michael$ ifconfig en0
en0: flags=8863<UP,BROADCAST,SMART,RUNNING,SIMPLEX,MULTICAST> mtu 1500
inet 192.168.2.1 netmask 0xffffff00 broadcast 192.168.2.255
inet 169.254.110.33 netmask 0xffff0000 broadcast 169.254.255.255
media: autoselect (100baseTX <full-duplex,flow-control>) status: active
supported media: none autoselect 10baseT/UTP <half-duplex> 10baseT/UTP <full-duplex> 10baseT/UTP <full-duplex,flow-control> 10baseT/UTP <full-duplex,hw-loopback> 100baseTX <half-duplex> 100baseTX <full-duplex> 100baseTX <full-duplex,flow-control> 100baseTX <full-duplex,hw-loopback>
Kleiman-ibook:~ michael$ # Note that the wifi does not have access to the world?Kleiman-ibook:~ michael$ fg
The server could be reached from the machine on the cable, but was not detected from the machine on the cable.
The server could be detected from the wifi, but could not be reached from the wi-fi.
3. Restarting with a new player name gave me a start at spawn, next to the bonus chest, with some of my inventory contents. But I'm not where I turned the LAN on; my inventory is just what I had when I raided the bonus chest, and none of what I did after that point, etc.
Equally, my inventory is not blank. Achievements are reset (new name), but inventory is not empty (player data saved as single-player world, but from some point in the middle of playing).
4. Two players next to each other do not see each other!
They did see each other once they started moving down the race track.
5. Speed of players is not accurately controlled.
The 9200 machine normally is too slow to play single player, or as client to a full speed machine. But as client to another PPC, it plays just fine. And, it's faster than the host.
Flat world, running past an NPC village, it maintained about 15-25 fps, while the host was around 8-12 fps. Speed was about equal until the village, and the slower guest machine (only running client) went past the host in the village (when the host's frame rate dropped.).
6. Diggy diggy diggy.
Both machine were giving a line of gravel and a shovel.
Test 1: Diamond shovel.
Answer: Not even close. Host was about 2/3rds the speed of the guest.
Initial thought: Each machine decides on it's own if it has spent enough time to break the block, and then tells the server "I think I broke it". The host does a frame update less often, and so takes more time before it thinks it starts, spends longer before it thinks it broke, etc, and the server never gives it any corrections.
Test 2: Hands.
Guest is doing 18-23 fps. Host is doing 3-5 fps. Host is about 95-98% of the speed of the guest.
Note that if it takes the machines longer to say "I think I broke", then the slowdown doesn't bother the host as much.
7. Non-flat world.
Same world was started on PPC machine and i7/4+4 macbook pro machine. Both worlds waited until the PPC machine had appeared to load the level, then "/time set 0".
Not surprisingly, the PPC machine could not keep up -- the world clock advanced slower (GUI doing 2-3 fps on normal versus 28-35 on far).
Reset both worlds back to time zero, open to LAN.
Now both game clocks (measure: watch the sun move) go at the same rate.
NB: Major slowdown on PPC machine at sunset. Only at sunset.
NB: BOTH systems opened the server on the ethernet cable; neither on the wifi.
8. A sample of the f3 complaints on the PPC:
Something's taking too long! 'root' took aprox 879.523 ms
Something's taking too long! 'root.tick' took aprox 409.401 ms
Something's taking too long! 'root' took aprox 526.776 ms
Something's taking too long! 'root.tick' took aprox 289.788 ms
Something's taking too long! 'root.gameRenderer.level' took aprox 128.301 ms
Something's taking too long! 'root.gameRenderer' took aprox 150.096 ms
Something's taking too long! 'root' took aprox 511.852 ms
Something's taking too long! 'root.tick' took aprox 325.301 ms
Something's taking too long! 'root' took aprox 544.122 ms
Something's taking too long! 'root.tick' took aprox 394.208 ms
Something's taking too long! 'root.render.display' took aprox 128.051 ms
Something's taking too long! 'root.render' took aprox 187.71 ms
Something's taking too long! 'root.gameRenderer.level.terrain.occ' took aprox 119.741 ms
Something's taking too long! 'root.gameRenderer.level.terrain' took aprox 124.585 ms
Something's taking too long! 'root.gameRenderer.level' took aprox 141.669 ms
Something's taking too long! 'root.gameRenderer' took aprox 186.861 ms
Something's taking too long! 'root' took aprox 782.065 ms
Something's taking too long! 'root.tick.level.entities.regular' took aprox 104.164 ms
Something's taking too long! 'root.tick.level.entities' took aprox 107.88 ms
Something's taking too long! 'root.tick.level' took aprox 110.909 ms
Something's taking too long! 'root.tick.textures' took aprox 171.864 ms
Something's taking too long! 'root.tick' took aprox 692.676 ms
Something's taking too long! 'root' took aprox 778.41 ms
Conclusion the first: As a client to a PPC host, PPC machines are just fine.
To test: i7 vs PPC connecting to a different PPC.
PPC connecting to an i7. (tested father down -- works great!)
9. Surprise surprise! Slower machine is faster!
Letting the 9200 (mac mini) host, and the 9550 (iBook g4) play guest, gave a surprise: The two machines were almost the same speed. The 9200, despite having a worse graphic card, was doing about the same fps -- shift-F3 showed lower C: counts for the same scene. Consistently.
In other words, with a weaker graphic card, it removed more stuff, displayed less, and had better performance. That's new here -- 1.1 did not, and performed so badly I haven't used that machine for minecraft since. My hope here was that with a G4 host, running at slower G4 speeds, another G4 could act as client.
And in a big surprise, I can get both machines performing acceptable -- by having the slower one do less on the display.
Now, can I tell my 9550 machine to do less as well?
Note: Test was on peaceful, in the ocean, with an E count from 0 to 25. No hostile mobs, and only squid.
NB: Sky on the 9200 machine has strange shading in what looks like a chunk above me, and the 4 chunks in any direction. But it's not chunk aligned -- as I move, the shading issues in the sky move with me (no apparent change).
NB 2: Have yet to manage to get all 3 machines talking to the same LAN world.
10. Final test: both machines talking to an i7 normal multiplayer server.
Well, I am impressed. Both machines are doing in the 10-15 FPS on normal, with sheep and a spider visible. That means that LAN parties have gone from "a nice idea" to "very playable" -- even with hostiles around. That's before optifine or anything else.
This is a very nice addition/surprise.
And, I'm still seeing a higher FPS on the lower-end video card :-).
11. Colors do not match.
Biome is a swamp. The 9200's grass looks normal, not swampy.
Moving elsewhere ...
(a few hours of playing with calibration later, the 9200's monitor is a lost cause)