2013/8/2 Till Heuschmann <address@hidden>
Please not another branch, it will be hard enough to sync the existing branches. I do not think that there is such a big bug in r2004 decoding (the existing segfault can be fixed with the correct struct alignment!).
Ok, if not a big deal, better to keep it together. But it do not hurt to isolate the problem for a first release, what about compiling switches?
Why force a release? Are there any benefits?
Yes, a tar ball is a good thing: some stable stuff to deal with. None may use a library which "never" get released. What's the problem to get an alpha release?
To get closer to a release this should be fixed:
- memory management
Yes, I agree, and I may try to help.
- object management (store the objects in a rb-tree or a hash map)
This is more like an API task, no?
You are very welcome to join libredwg again! I think about joining you.
Ok, nice, could you point what task could be done in parallel with the GSoC branches for now? I'm still lost, maybe I need just keep on understanding the code, ok?