Single Operation about punchclock

[rinkeb]

+1
Would like to see this as well. Have to maintain a separate status flag now.

[anaisbetts]

So, instead of serializing operations with the same key, you want to drop them on the floor, right?

[rinkeb]

Indeed. This is (at least for me) about repetitive, reentrant synchronization operations where it does not matter when one or more are skipped.

[prasannavl]

I haven't looked into the implementation yet. But it'd make sense if the priority is validated first to make sure only lower or equal priority is dropped, and higher priority gets replaced - Presuming of course, the existing behavior remain untouched, and so that this feature is explicit is in its nature using a different method name perhaps.

[anaisbetts]

You could get a version of what you want by changing the Concat here to a Switch, but that means that an existing operation with the same key would be cancelled, and the latest one would take precedence. Probably not what you want.

[megakid]

@paulcbetts I am also very interested in this feature. Does anyone here have an implementation for this?

[ghuntley]

@prasannavl @Rintje @megakid we are preparing for a new release of punchclock (upgrade to netstandard v1.0 and implement CI) - if you're interested in taking lead position (with oversight/mentorship from Paul) in implementing this please let us know.

[giusepe]

I close this issue cause it seems a bit out the scope to me and since no one mentioned about it in the last 2 (3?) years, there's no reason to keep it open.
If anyone is willing to take this further and try to implement it in a way that won't break what we currently have , I'm more than glad to look into it.