Different Branches
25 views
Skip to first unread message
Eddie
Jul 9, 2012, 4:03:16 AM7/9/12
to sc2r...@googlegroups.com
Hi I'm new to sc2reader but I'd like to help with development on github.
Are other branches more active? It seems the s2gs branch in particular seems to have a lot of development. Why isn't that branch just merged with this one?
Graylin Kim
Jul 9, 2012, 11:23:26 AM7/9/12
to sc2r...@googlegroups.com
Hi Eddie,
The s2gs branch is the branch currently under development. It hasn't been merged to master because s2gs parsing, the main addition, wasn't reliable yet. I wouldn't want to guess at the success rate but I know that it errors out on a good number of s2gs files. We could always just put a warning label on the GameSummary resource and merge but I've been hesitant to do that.
Prillan has added a couple additional patches to improve this but there are still issues. The primary issue is with the ReplayBuffer.read_data_struct method. Some aspects of blizzards byte structure aren't interpreted properly. I think with a large enough set of s2gs files the byte structure parsing could be sufficiently improved.
The new_data branch hasn't been getting much attention but it is an effort to switch over all the data to json files that either get loaded on the fly or pre-processed into data.py files. The current task in that branch is primarily checking all the auto-generated json files for accuracy and fixing all the mistakes.
If you want to get started with any of this I'd be happy to help and answer any questions.
~Graylin
The s2gs branch is the branch currently under development. It hasn't been merged to master because s2gs parsing, the main addition, wasn't reliable yet. I wouldn't want to guess at the success rate but I know that it errors out on a good number of s2gs files. We could always just put a warning label on the GameSummary resource and merge but I've been hesitant to do that.
Prillan has added a couple additional patches to improve this but there are still issues. The primary issue is with the ReplayBuffer.read_data_struct method. Some aspects of blizzards byte structure aren't interpreted properly. I think with a large enough set of s2gs files the byte structure parsing could be sufficiently improved.
The new_data branch hasn't been getting much attention but it is an effort to switch over all the data to json files that either get loaded on the fly or pre-processed into data.py files. The current task in that branch is primarily checking all the auto-generated json files for accuracy and fixing all the mistakes.
If you want to get started with any of this I'd be happy to help and answer any questions.
~Graylin
Reply all
Reply to author
Forward
0 new messages