Ultimate Omicron
- renex
- renex
- America/Sao_Paulo
- Yes
- No
MC-3567
In our server, most of us use a custom jar with Optifine and a language mod to match the server's texture pack. The language mod modifies languages.txt to include another entry that can be selected from the ingame menu.
When using English (en_us), Anvils behave correctly. However, when using the custom language (en_er), Anvils will behave oddly and believe every operation is accompanied to a rename, making all costs excessive.
What happens is:
When you place an item in the Anvil slot, it will be available to pick up on the result slot, even though no operation was specified (repair, mix, rename). Picking up this item from the Anvil result slot will make it vanish and will consume XP equivalent to that of renaming, and also make the Anvil use sound. Upon exiting the Anvil interface, the item will be thrown to the ground and will display a renamed tag. Moreover, an example operation that would normally cost 9L will instead silently remove 16L and make the item display a renamed tag.
Inputting the regular english name of the item will deny the operation. Inputting the language mod's regular name for the item will initially deny, but a fraction later revert and display a possible operation with no cost. Screenshot included.Modding the server jar with the language mod makes no difference.
Rejoining the server with English (en_us) will still display the modded name on renamed items, showing they are in fact renamed.
Strangely enough, this entire issue does not happen on Single Player.One approach to this would be allowing the texture pack to include a custom lang file that both the server and the client can load, accompanied by a toggle on the game's options.
I built a TNT cannon with 80 focused charges. When firing at an angle of 45º i blasted off to y=620, at which point my character acted like if it hit a solid ceiling, sliding sideways and falling. When i used half charge, i blasted off normally and hit a vector of y:1059.
Then i built a larger cannon with 160 charges, at which the full-blast cap was seemingly reduced to y
~=310. Character instantly hits this height and slides parallel to the ground until coming to a halt approximately 500m from the cannon.I attached a scheme of the observed trajectory.
I built a TNT cannon with 80 focused charges. When firing at an angle of 45º i blasted off to approximately y=620, at which point my character acted like if it hit a solid ceiling, sliding sideways and falling. When i used half charge, i blasted off normally and hit a vector of y:1059,5.
Then i built a larger cannon with 160 charges, at which the full-blast cap was seemingly reduced to approximately y=310. Character instantly hits this height and slides parallel to the ground until coming to a halt and falling to the ground approximately 500m away from the cannon.
I attached a scheme of the observed trajectory.
I built a TNT cannon with 80 focused charges. When firing at an angle of 45º i blasted off to approximately y=620, at which point my character acted like if it hit a solid ceiling, sliding sideways and falling. When i used half charge, i blasted off normally and hit a vector of y:1059,5.
Then i built a larger cannon with 160 charges, at which the full-blast cap was seemingly reduced to approximately y=310. Character instantly hits this height and slides parallel to the ground until coming to a halt and falling to the ground approximately 500m away from the cannon.
I attached a scheme of the observed trajectory.
I built a TNT cannon with 80 focused charges. When firing at an angle of 45º i blasted off to approximately y=620, at which point my character acted like if it hit a solid ceiling, sliding sideways and falling. When i used half charge, i blasted off normally and hit a vector of y:1059,5.
Then i built a larger cannon with 160 charges, at which the full-blast cap was seemingly reduced to approximately y=310. Character instantly hits this height and slides parallel to the ground until coming to a halt and falling to the ground approximately 500m away from the cannon.
I attached a scheme of the observed trajectory. On multiplayer, all that happens is that after (latency) milliseconds the player is teleported back to the explosion origin with a speed of zero like if it never happened.
Y coordinate capped during uber TNT launchPositive Y speed discarded upon exiting loaded area
Positive Y speed discarded upon exiting loaded areaY capped during uber cannon launch
scheme.png
Intended behavior. The pressure plate is powering the hopper, which in turn stops taking in items.
After big explosions, the player can sometimes fly off with such a high speed that it will exit the loaded region before the server can load/generate new chunks. Since recent versions, a fix for dropping to the void after teleporting was added that discards the player's Y speed when not inside any chunk. However, this breaks some cannon designs.
A solution would be to only remove negative Y speeds when exiting the loaded region; This lets these
extreme cannons work as before without breaking the previous void fix.After big explosions, the player can sometimes fly off with such a high speed that it will exit the loaded region before the server can load/generate new chunks. Since recent versions, a fix for dropping to the void after teleporting was added that discards the player's Y speed when not inside any chunk. However, this breaks some cannon designs.
A solution would be to only remove negative Y speeds when exiting the loaded region; This lets these vertical speed mechanics work as before without breaking the previous void fix.
Confirmed for 14w30c.
Okay. I was pretty sure it was related to the language mod, though.
I have seen this maybe four times my entire life. It's very rare, but it's there. You'd have to push a million buttons to see this happen.
Happens on smp, too, and it's not a client-side problem because a minecart moved halfway through a powered rail with this shorter pulse.
I had a painting go through an entire mountain and drop on an opening below. Seems like it drops right at the intersection between blocks. IMO paintings fly off too fast, it's like they explode off the wall.
There seems to be a slight delay when loading sounds that haven't played yet, and this happens often when attacking mobs for the first time, and is very annoying.
This used to happen to me, but it stopped happening recently. Vanilla client. Then, out of nowhere, it started happening again, /even connecting to localhost/. It's got nothing to do with the internet connection or the machine, but something that screwed up in the game between 1.2.5 and now. It's pretty random, but when it happens, the server ping is added to exactly 15000. I never experienced this on 1.2.5, ever, and i used it for way longer than 1.3.x and 1.4.x combined. I remember this started on one of the first 1.3 snapshots, but i'm not sure which. Could be an issue in Java 7 too, i remember that i made a special launcher that ran the game with Java6 to fix the sound issues of 1.3, i'll do some testing with it and report back.
—
Nope. Java6 does not affect the bug, even though the entire game takes much longer to load.
http://prntscr.com/lwget
That's caused by many entities violently expulsing each other. You should be glad they don't explode through the walls like they did on previous versions.
He probably downloaded a world named "world" and overwrote his own.
Easily reproducible in 13w01b and any other version :/
Attached crash report during fire, and crash report after firing. Do you want me to attach the map with the loaded cannon? It's 1.9 MB.
The output is working as intended however the short pulse it gives after it's powered is a bug.
That happens because the hopper has a composite hitbox. The item tries to fall on the carved middle, but can't occupy that space due to the new item ejection code so it jumps. Similar behavior can be seen with the lower step of a stair block.
Done
So you basically break the anvil client-side while it falls server-side? Seems normal to me as long as it corrects itself after it's done falling in the server.
I saw this happen intermittently on 1.3.x, really hard to reproduce. But it does happen.
High latency and high packet loss?
You can use Soul Sand to have the cart fall down a cliff.
Many wireless keyboards have trouble combining multiple keys. Try remapping the jump action and see if it still happens.
This bug was introduced on the smp/ssp merge and was then fixed on 1.4.7 with the help of a user. Did not see it happen ever since, and i used to have severe problems with it.
You are most probably melting down your graphics hardware. Try setting the FPS limiter to "Power Saver" to see if the frequency of crashes diminishes. Also check if your drivers are up to date.
Agreed. Air friction is too high and hinders gameplay in various aspects.
Please provide a screenshot of your setup.
Thanks
I gave it another shot just to be sure, was capped at y:372.
This is intended hehavior, it's not a dynamic block. You place it and it stays in position forever.
For now you can use detours and space between the hoppers. Are you using water to transport the items?
Indeed, it's an artifact caused by slight latency. Nothing to worry about, it's what happens when the server notices your client is a little bit late and corrects it.
Maybe adding a little jitter to the entity ejection code?
This behavior makes it possible to count time during the night. So it's a good feature and should stay.
Aw damn, i spent half an hour searching... This search isn't efficient.
That seems good
Thank you.
Thing is; This was supposed to prevent cheating on durability by switching tools. However, when any aspect on the itemstack in changes the damage is reset. This is incorrect, since durability and count changes should not affect the mining of a block. So it's a bug created by a previous fix.
Try disabling Threaded Optimization.
Also check if your video card isn't overheating, Minecraft can generate a lot of load even on the best cards and nVidia cards tend to flicker when they're above 65C. And they pass that barrier way too easily.
Most probably a client-server desync.
This might be an overheat of the graphics card. Try setting the performance to Power Saver and view distance to Tiny and see if it still happens.
Redstone can power the block it is sitting on, or pointing to, and it can be powered from every direction. Stairs are transparent, which means redstone pointing to it won't power it and redstone running on top won't power it either; However, if you point redstone to a stair which has dust on top, you're powering the block below the dust and this transmits power. This is mostly intended, and also a zero-tick diode, super useful.
Might be a video card issue. Also, make sure the stitcher created textures with no blank space. You might need to assist the unstitching of textures for these smaller packs.
Yes, texture sizes need to be powers of two because the video hardware requires it. You can't change the way chips work.
Working with non-power of two on the old system was limited to high-end video cards that correctly mapped noncompliant textures. They were prepared to handle user error. Using textures with incorrect resolutions is user error and usually not handled very well by the graphics card.
What they can do about this instead is add a width and height parameter to the texture pack. You would still use power of two resolutions, but only display a certain part of it. So if you specify your textures are 17x17, you'd use a small fraction of a 32x32 texture while the game maps only the 17x17 section. This adds the functionality you need while also respecting the basic requirements of video cards.
News. I finally understood what happens; These shots have an oblique launch angle, so the player hits unloaded chunks. Due to the new mechanic implemented to avoid teleported players falling into the void, the Y speed is discarded when they exit the loaded area. As soon as the chunks load, the player then falls with gravity.
So i believe this is working as intended
Uploaded the correct trajectory diagram.
The map file attached to
MC-6584can display this effect. After i precised the true cause of the problem in that issue, i made a new issue to better illustrate the error and a possible fix.Disable threaded optimization globally, not just for javaw.exe. Fixed for me.
In 13w07a, directly pointing a comparator on normal mode at a self-resetting BUD causes the BUD to blink. Setting the comparator to subtract mode stops the pulsing.
Forgot to say; Reconfirmed in latest.
Again, depends on the keyboard. Many nonstandard keyboards like laptop or wireless devices have a lower simultaneous key limit, and the spacebar is a key that usually demands more. So you can't usually combine more than 3 keys if one is the spacebar.
If the mouse button is participating in the problem, then you might have either a laptop setup where the keyboard and mouse participate in the same ps2 connection or a wireless relay with the keyboard as middleman.
Just throwing ideas.
It actually looks pretty good. If only it didn't cover a part of the map.
The piston head could very well be modified into a useable block for creative mode, but that's more of a feature request than anything.
The launcher threw me here so I didn't know what to do, sorry.
Confirmed for Radeon HD 5570.
I bet this is too difficult if not impossible to reproduce in latest due to extreme optimizations. Besides, cannons that explode so violently aren't even a concern as they somewhat break the expected game behavior and instantly kill anything with health in a 20 block radius from it.
I will, though, try and recover my cannon map and attempt to recreate this issue. I'll have to rebuild it because a piston timing issue in a recent snapshot destroyed the cannon. Hang on.
e> using a /fill cannon I managed to reproduce the problem with view distance set to 2 and 150 tnt charges.
Issue much more noticeable in latest snapshot's remote multiplayer.
Confirmed for 14w30c.
Did you check your FOV option?
Yep, that's the same issue, even though each report focuses on a different part of the problem. Thanks!
Also happens for me on latest (32b, 32c crashes everyone for me) when there are online players.
This makes zero-maintenance batch-restarters useless, because when I remotely /stop I'm still online and it locks up, requiring me to get to the physical machine.
If command blocks could /stop, this would have a workaround.
A usable workaround, though, would be to have a sniffer program look up a magic word on server log files and kill java when it finds that.
Was this fixed? I am currently using a sniffer to manually kill the server due to being unable to stop from inside.
Hmm, If this is not fixed for release I might have to rely on a separate tool to shutdown the server from inside. I have been using a Game Maker executable to kill java when it finds a passphrase on the server logs to accomplish this. Might write a dedicated tool on a more appropriate platform.
If command blocks could run all commands this would not be a problem too.
Yeah, this is a major problem for remotely managed or zero-config servers. Specially since sometimes the server will overload and then the target machine is saturated, preventing remote connections to stop the server.
Well this is bad. Bumping for visibility.
I will now try to find a sequence of ingame actions that crashes the server, and try to automate that.
Confirmed on latest 1.8.1 pre3 using tnt player cannon.
It's very difficult to reproduce now that chunks load really fast but it's still there.