strands-project/sicks300

reintroduce STRANDS-specific changes and PR upstream

Closed this issue · 3 comments

As discussed on gitter:

  • @BFAlcerda:
    • using the apt-get sicks300 package stops filtering out laser readings that are inside the robot because of the internal pct. Anyone knows what we did to filter it in our fork? i know that mira does the filtering, i've just seen it masking the inside readings a couple of days ago. was our sicks300 package integrated with mira somehow?
  • @cdondrup:
    • There is a parameter reduced_fov which, if true, sets the field of view to 180. Did we do that?
  • @marc-hanheide
    • good question... we have now moved to use the official sicks300 package fork as opposed to have our own version
    • oooh, this looks suspicious: strands-project/scitos_drivers@058bdc9#diff-2055efedbd9df636bf6501a14736094d
    • @cburbridge obviously introduced it and it's missing from the current fork, I suppose
      that commit had a few other changes that makes it deviate from themaster fork quite a bit 😞
    • we could cherry-pick those commits into our fork and then try to create pull requests upstream, but @cburbridge should comment on this I believe
  • @hawesie

I have opened a pull request for changes upstream.
bohlender#3
Then, to use this with our scitos robots a launch file like this should be made:

<launch>
    <arg name="laser" default="/dev/ttyUSB0" />
    <node pkg="sicks300" type="sick300_driver" name="sick300" output="screen">
        <param name="devicename" value="$(arg laser)" type="string"/>
        <param name="field_of_view" value="225" type="double"/>
        <param name="send_transform" value="0"/>
        <remap from="/laserscan" to="/scan"/>
    </node>
</launch>

Do you want me to open a PR against this fork too?

Did you see this: bohlender#4

I think let's wait a bit for him to respond. Those two PRs are non-conflicting and once he accepts them it should all be fine and we discard our fork (i.e. sync it with the upstream)