[Ns-developers] upcoming ns-3 merge plan

[Tom Henderson]

Hi all,

We have a lot of parallel development efforts ongoing, and I would like 
to suggest that we start to move several things to ns-3-dev.

1) Python:  Gustavo has been progressing the Python bindings based on 
feedback from Mathieu, and they now seem to be ready to merge and to 
start to get some user experience with them, so we'd like to suggest 
merging them ASAP

2) NSC:  there are several integration issues that were previously 
mentioned on the list (64 bit build and build integration into ns-3, 
handling multiple interfaces correctly, socket errors and errno) but 
once these are addressed, we should probably start to merge Florian's code

3) IPv6:  we should start to merge self-contained portions of the IPv6 
tree, such as starting with IPv6 address and device changes.  There is 
still the overall IPv4/v6 integration task that Sebastien is taking a 
look at.

4) Gustavo's bridge repository, once the outstanding questions are 
resolved, should be mergeable soon

5) ns-3-emu:  we had some discussion yesterday about a merge plan, which 
Craig can summarize in a separate message

6) tags:  we likely will need to update the packet tags API and 
implementation, which I'll reply to in the separate thread, and we 
should try to get those fixes done by the end of the month or early next 
month

7) IPv4:  There are several issues relating to the existing API and how 
it will line up with IPv6, which I'll summarize in a separate message, 
and these items will be a priority to merge before the ns-3.2 release.

8) statistics:  We discussed with Joe the possibility of merging his 
statistics framework into src/contrib for now, again with the goal of 
getting it in front of users.  Joe has already written nice 
documentation about this on the wiki, so I think that all that remains 
for merging is to integrate it with the ns-3 build system

If anyone wants to get something else into the near-term merge queue, 
please suggest on the list.  We want to coordinate and distribute the 
merge pain sensibly.  I'm thinking that end of August might be a good 
target for the next release.

- Tom