Random Chitchat 2012-2016

[xoft]

It's unbelievable, really, how the project has grown in three years into a behemoth:
October 2011: 23k lines of code
June 2014: 140k lines of code
Source: Ohloh - https://www.ohloh.net/p/mc-server

[NiLSPACE]

That's amazing

[bearbin]

Wow, especially as we removed some of the libraries into Submodules.

[xoft]

I believe I set Ohloh up so that it doesn't count the libraries anyway - they mostly aren't our code, after all. So the numbers exclude the libraries in both timeframes.

[bearbin]

Oh ok then

[worktycho]

I just discovered, if you take the default generator setup and set heigen to flat the server never starts up because it can't find a spawn point because the whole world is under water.

[worktycho]

@xoft, I'm trying to implement batching on the lighting as the start-up overhead on the GPU is significant. The problem is allocating the buffers to store the data in. Should I allocate the data dynamically each batch or used fixed maximum batches and allocate the buffer at the start? These are big buffers (half a MB per item in the batch so potentially 10 - 100 MB depending on batch size).

[xoft]

I don't really have enough information to give you a qualified answer. Use common sense - if the code would be allocating and deallocating buffers on the order of hundreds per second, then use a fixed number of buffers preallocated.

[worktycho]

More likely once per second. The reason I haven't given more information is I don't know myself. The frequency depends on the size. the bigger the buffers the less frequent.

[NiLSPACE]

I got the laptop for my next school It's the Lenovo Ideapad Z710-00183. I get 336 ch/s with the pre-build MCServer on default settings.