ROS2 support: ROS Answers?

@gavanderhoorn if you still feel strongly about the drawbacks of encouraging ROS 2 users to use answers.ros.org, we’d still like to hear them and be open to changing even as we enact the proposed strategy.

In the meantime, we’re going to move forward with this change by:

  • updating our support recommendations in the ROS 2 documentation to suggest users post to answers.ros.org
    • @marguedas I think you were going to do this? (I can help if you send me specific tasks)
  • all of the ROS 2 team members at Open Robotics will subscribe to the ros2, ardent, and r2b3 tags
  • if we decided it’s a good idea, add and subscribe to the “i don’t know which ros distro” tag

Again, we’re moving forward with this change, but are open to reevaluating it or taking new criticism into account as we go.

I’d rather not have a tag that requires moderators to try to correct/guess what the users are doing. If they tag it with “I don’t know” the next thing will be for someone to ask in a comment, “please find out what you’re using by running this command …” And then we all wait for the user to run the command and edit their question. Whereas if they tag it with their best guess, odds are their best guess is actually what they’re using, as that will be what they’re familiar with, and we only have to iterate with those who guess wrong.

If we just add that to the question prompt we can take out a whole user interaction. Also there’s still people who read the home page and if anyone browses and sees something that doesn’t match it can easily be re-tagged by anyone with enough karma.

Asking a person who is asking a question to do a little bit more research before they post will be way more effective than setting up a mechanism which will require several extra interactions to tag things before the question can effectively be answered.

I don’t feel strongly about it actually.

The clean cut that ROS2 was/is supposed to be would perhaps make a separate forum a possibility - it seems like it’s mostly newcomers to the community that are really using it. I can see why you’d be hesitant to do that though (duplicate infrastructure, community fragmentation, etc).

From my personal experience I know that finding Q&As relevant to people’s problems is already impossible/hard enough (especially for newcomers) judging from the nr of (near) duplicates (that or they just don’t search). Adding a few more ROS distributions will probably not significantly change this situation.

The tags have been created on answers and the helper text of the tag textbox specifies that tagging the rosdistro is mandatory.
I don’t think there is a way to make the tags mutually exclusive (or if it’s even desired as questions can span several distros). Hopefully most people know what rosdistro means and won’t just tag ros1 and ros2 everywhere. If it appears to not be the case we can consider ways to make it clearer for users.

Yup mostly some grep’ing in the issue templates of our repos and the ROS 2 wiki, I don’t think there is much more to do. I’ll do it as soon as I have a spare cycle.

Thanks again @gavanderhoorn for the feedback and the incredible work you do on answers. As mentioned in previous comments feel free to bring up anything that comes up after the transition and we will try to adapt the strategy accordingly.

I think we need an explanation of what the distro tags mean. There have been a couple tagged with r2b3 unnecessarily, and someone just asked a question tagged with both kinect and ros2 which is, as far as I can see, not related to ROS 2. Is it possible to put a list below the tag field?

1 Like

I updated it to list the active rosdistros explicitly. I was hoping not to have to list them all and have one more place to update. But hopefully this will guide people more accurately.

image

So far it seems to be working.

Yeah, though a few of the corner cases have come up already, e.g. someone picked r2b3 at random because it was required and they didn’t know which to pick, and I believe @gavanderhoorn had one question which he marked Kinetic somewhat arbitrarily because it didn’t really apply to any one ROS distribution.

But aside from a few cases like this, I believe it is working as intended.