Jonathan Binney

It looks like both the mean and median filters in this repo use `push_back()`; I'd guess that is the most common case.

I should have asked this earlier, but at some point is it worth encouraging people to just use `boost::circular_buffer` directly instead of exposing more and more of its member functions...

I believe this was fixed by #76 .

Wow, that is odd! [ScanToScanFilterChain](https://github.com/ros-perception/laser_filters/blob/65cbefd6e8a4f58b360061f8786d80668505cdb2/src/scan_to_scan_filter_chain.cpp#L110) just calls [FilterChain::update()](https://github.com/ros/filters/blob/5507bc5175cb74e825ee20b9dc19e67ef6a4a022/include/filters/filter_chain.hpp#L195), which just calls each filter in order. Each filter in the chain can do whatever they want to the timestamp; but typically...

Looks like there are three kinds of filter in the chain; here is where each of them update the timestamp in their output scan: - [AngularBoundsFilter](https://github.com/ros-perception/laser_filters/blob/65cbefd6e8a4f58b360061f8786d80668505cdb2/include/laser_filters/angular_bounds_filter.h#L105) - [ScanShadowsFilter](https://github.com/ros-perception/laser_filters/blob/65cbefd6e8a4f58b360061f8786d80668505cdb2/include/laser_filters/scan_shadows_filter.h#L141C8-L141C24) - [BoxFilter](https://github.com/ros-perception/laser_filters/blob/65cbefd6e8a4f58b360061f8786d80668505cdb2/include/laser_filters/box_filter.h#L132)...

The reason for changing the timestamp is that not all readings in a LaserScan message happen at the same time. The laser (typically) spins and takes readings one after another....

Should be fixed by https://github.com/ros-perception/laser_filters/pull/186 , so i'm closing this. Feel free to reopen if it is still a problem.

Looks like you update this PR - thanks for working on this! Is it ready for me to take another look?

Looks good. @jcarlosgm30 thanks for this PR!

@bjsowa sorry i haven't had time to review and merge this! Are able to try out this branch and give some feedback?