-
-
Notifications
You must be signed in to change notification settings - Fork 5
Full rebuild June 2025 bump ros2-distro-mutex to 0.10.0 and build number to 10 #23
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Conversation
d9cad83 to
d864057
Compare
Signed-off-by: wep21 <daisuke.nishimatsu1021@gmail.com>
|
@traversaro I have no idea why ros2-distro-mutex is not rebuild in ci(works fine locally). Any thoughts? |
I am not sure if it s related, but note that the check-patches step is failing, but somehow the CI is not failing (related to RoboStack/ros-jazzy#86). Furthermore, I think we could:
|
After https://github.com/prefix-dev/rattler-build/pull/1666/files we can probably remove many workaround in the testpr job. |
It seems that in 685cd40 the build of additional_recipes was commented out, but I can't remember doing that intentionally. |
|
I re-added the build of additional recipes, and drastically simplified the testpr job, and moving the check patches to its own step seems to have solved RoboStack/ros-jazzy#86 . It would be cool to just use pixi tasks directly in the testpr, but I guess for that we need to add cross-compilation support to the pixi tasks. |
|
@wep21 I did some cleanup, I think some patches are still failing but that seems as the patches are actually outdated (I hope). |
Signed-off-by: wep21 <daisuke.nishimatsu1021@gmail.com>
|
@traversaro How can I avoid this version conflict? |
I think we can merge conda-forge/urdfdom-feedstock#38 . I was hesitant to do that as I was afraid that that release was not intentional, as it was done without any actual breaking change to urdfdom, but after ros/rosdistro#45869 I think for sure it is an actual release. Anyhow, to avoid that the rest of the distros remain stuck with an old urdfdom (that will create a diverge w.r.t. to the version pinned in the conda-forge pinning), I think we need to add a new vinca parameter in the |
Actually I would like to wait for the feedback on ros/urdfdom#220 . To avoid blocking the PR, I implemented RoboStack/vinca#81 . |
The build was unblocked by bc7875f, I also took the occasion to devendor some packages that have much more features in the conda-forge build, that should provide more feature to users and reduce the time spent in the build. |
|
backward-ros does not build on osx-64 - we should really try and unvendor this .. does your new magic help @traversaro? |
Unfortunatly the backward-ros and backward-cpp from conda-forge packages are quite different.
So if we want to de-vendor it, we need to do a good old heavy patching of the package. |
|
I think I "fixed" the issue by basically disabling this feature on osx-64 by not installing the required |
|
Is it possible to rerun only osx-arm64? |
|
I’m not too worried about it; we can fix issues in a follow up PR |
|
Thanks @Tobias-Fischer @wep21 , I would include also the fix RoboStack/ros-jazzy#57 if that is ok for you. |
|
I'm fine for including the fix. Thank you! |
|
It would be cool to also unvendor ompl , but that is probably blocked on conda-forge/ompl-feedstock#53 . |
|
CI is clearly happy, let's merge so we can quickly deal with post-PR failures (if any). |
There is: #24 . |
No description provided.