Release 100 and 101 introduced a few breaking issues for a mixture of players. This post is to update everyone on the status of the fixes.
The issues are:
-
ATI driver issue: The game fails with a black (or white) screen during startup. Here. Here. Here. Here. This appears to be a driver issue. This means that code, shaders and assets that work happily on other graphics cards and other drivers, for some reason, fail on some ATI devices. These are a huge pain to work around. Ideally the drivers would work as expected. These issues are extremely difficult to reproduce without the specific device and driver combinations that are failing. We’re still working on this - and it’s our #1 priority.
You’re hitting this issue if your Oort Online log ends with:
consoleobject.cpp(272): (turbulenz::ConsoleObject::LogCB) CONSOLE LOG: Using standalone worker[object Object] consoleobject.cpp(272): (turbulenz::ConsoleObject::LogCB) CONSOLE LOG: MeshWorkerManager onMessage WORKER_CREATED consoleobject.cpp(272): (turbulenz::ConsoleObject::LogCB) CONSOLE LOG: WORKER_CREATED -
Missing servers: Some servers are still offline. Here. Here. Here. Here. Here. Here. Updating the game often comes in two parts: 1. updating the game application itself, and 2. updating the saved games (basically the worlds). It’s easiest to just throw away the data and ask players to “start again”. But this isn’t a great option for voxel MMOs! So instead we write migrations from the old data format to the new data format.
The migrations can sometimes be complicated depending on what has changed. For example, say we remove a block type, we need to go through every world and every chunk searching for that block and replace it with an alternative. But the migrations are extra complicated when we have 50 servers to migrate and many GB of data. Migrating each world took 3+ hours of processing.
(This isn’t a sob story - just trying to give some insight into what we’re doing behind the scenes.)
We originally thought that the failed migrations would be simple to fix - but we actually found it quite hard to work out where the problems were - particularly because the migrations were taking so long to run. Again, we hope to have all the world live tomorrow. (Where have we heard that before?)
-
SSL sign up issue: Some users reported failures after purchasing the game when attempting to create accounts. Here. Again, we can’t reproduce the issue locally, so we’ve tweaked and updated our SSL certificates. Fingers crossed this will resolve the issue.
-
Game Inventory(?) crashes: Some users reported crashes when opening the game. Their Oort Online logs listed inventory flavoured error messages and failures. This issue has been found and fixed, and will be deployed tomorrow.
These appear to be the main issues. Sorry everything didn’t work as expected after the update. All the issues all ultimately a result of playing an Early Access game that is still in active development. We’re working to get everything fixed ASAP.
If you have any questions, let me know.
And the results of vote 3 say i’m not alone.
for us.) When this happens the game generates additional chunks as a flat plan. The problem is that the migration process also needs to take into account the additional chunks. There are so many additional chunks that the migration process is taking over 10 hours. (We really need to get the wrapping worlds in to stop this from happening.)