-
Notifications
You must be signed in to change notification settings - Fork 82
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Updates to support mod pack development #207
Conversation
Thank you for your effort, but could you please break this down into more concise PRs? Also, have you tested the performance impact for larger atlases for the change from ints to Biomes? |
How would I go about doing performance testing? |
I'd start with a big map and look if there is significant lag occurring during rendering or export as png. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Actually, the patchset isn't that big. I have a few issues though.
EDIT: Submitted this as your issues post appeared. It's worth noting that I started by removing reliance on the deprecated FMLLog, among other things. How should I handle this in the PRs? Performance shouldn't be too greatly impacted by the change to Biomes. The only function that is called regularly that has been altered is BiomeTextureMap#getTextureSet. We have to do one conversion from ID to Biome per tile drawn. As the underlying map is a hash map, this should be speedy. It's client side only, so I'll be looking for a hit to FPS rather than TPS. In all my testing so far, I have not observed this. |
On the subject of performance. Frame rate, as well as frame time consistency as monitored using F3 debug screen, is indistinguishable between the two versions. Tested in a variety of situations, including; flying generating new chunks with atlas overlay shown and hidden, flying loading old chunks from disk, standing in various different biomes with and without rain, with and without nearby villages. Also tested was rapid panning and zooming in the GUI. This causes frame drops in both versions. During all of these tests, average TPS remained comparable, and only going below 20TPS when generating new terrain. Hardware used (for reference) i5-5287U laptop low power processor with Intel 6100 IGP. Is there anything further? |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I know, this is nitpicking now, but these small things add up in a large code base.
CI fails due to mapping change breaking the screenshot folder existence check. I need to rebase. Back soon :( |
Experiments with using resource locations internally. This probably won't stay in. Revert "Experiments with using resource locations internally. This probably won't stay in." This reverts commit ef27806 Possible fix - allow addition/removal of biomes after mod added to pack in old worlds. Bug fix Change biome categorisation so deserts are hot Support for more than 256 biomes in biome detector. Add swamp biome type Must do texture map load in init, after mods have registered their biomes Match formatting Fix test? Handle biome removal and malformed strings in BiomeTextureConfig without crashing (whoops!) Address issues. Fix pseudo biomes - these didn't work in a Biome keyed map. Use LogManager directly, instead of FMLPreInitializationEvent's wrapping Whoops, I did it again. Fixed the test. Logger can never be null now, right? Nits picked
Done, I think |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
In relation to #161 I have achieved the aims of that pull request but did not go as far as TehNut suggested would be necessary. |
Hi,
This pull request includes the following major changes:
As a result of these changes, the version number of BiomeTextureConfig has been bumped to 2.