Conversation
|
What is the proposed roll out strategy for this breaking change? |
|
I have opened pull requests against every instance of this template in ros2.repos. |
re: ament/ament_lint#209 Signed-off-by: Tully Foote <tfoote@osrfoundation.org>
|
@tfoote, are we still interested in getting this in? Can I do anything to help? |
|
Pulls: #209 |
|
Pulls: #209, ros-visualization/rqt_bag#205, ros-visualization/rqt_reconfigure#157, ros-perception/image_common#382, ros-visualization/interactive_markers#118, ros2/geometry2#855 |
tfoote
left a comment
There was a problem hiding this comment.
Thanks for iterating on this and fixing the other places it's cropped up. The changes look good to me, pending CI, which looks like it's having issues right now.
Related with ament/ament_lint#209 --------- Signed-off-by: Alejandro Hernandez Cordero <ahcorde@gmail.com> Co-authored-by: Tully Foote <tfoote@osrfoundation.org>
Flagged in #109 this template is not a valid template for the BSD license.
The correct license was added in #205
The original author doesn't know where it came from: ros2/geometry2#222 (comment)
This will cause a lot of warnings/errors. But anyone who's using this template for legacy code has modified the license, naming it something that doesn't exist. It doesn't match with legacy code as flagged in ros2/geometry2#222 And anyone who has written new code is not using the standard BSD 3 clause license that we would want because they're trying to pass our linter which wouldn't pass with that in the past. The subtle differences will cause nightmares for due diligence efforts.
The longer we go the worse the problem will be. We could add a backwards compatibility period for a while but there's no way that we'll know that everyone has fixed things, and in that time more people will likely follow the incorrect template.