Unpin time button uses current time, and should use plot settings time #80

Open
DanBerrios opened this Issue Dec 11, 2012 · 3 comments

Comments

Projects
None yet
3 participants
@DanBerrios
Contributor

DanBerrios commented Dec 11, 2012

No description provided.

@VWoeltjen

This comment has been minimized.

Show comment
Hide comment
@VWoeltjen

VWoeltjen Jan 3, 2013

Contributor

Just to clarify, this is undesirable when no data is flowing (or for historical/predictive data)? It seems like this would be correct behavior for live data.

In cases when data isn't flowing, should this restore the time from plot settings, or should it simply keep the current time axis settings (which may be panned/zoomed)? My assumption would be the latter (since the former is handled with corner reset buttons) but I am not sure.

Contributor

VWoeltjen commented Jan 3, 2013

Just to clarify, this is undesirable when no data is flowing (or for historical/predictive data)? It seems like this would be correct behavior for live data.

In cases when data isn't flowing, should this restore the time from plot settings, or should it simply keep the current time axis settings (which may be panned/zoomed)? My assumption would be the latter (since the former is handled with corner reset buttons) but I am not sure.

@VWoeltjen

This comment has been minimized.

Show comment
Hide comment
@VWoeltjen

VWoeltjen Jan 4, 2013

Contributor

Discussed this with Tom & Madelyn to work out the various cases - for streaming data, the goal should be to keep "now" (as defined by the telemetry point's time service) on the plot once it's un-pinned, without changing span (and changing as little else as possible).

To this end, three things should be tried until "now" falls within the axis:

  • First, leave time axis as it appears when unpinned
  • Second, restore maximum value of time axis to what it was when first pinned (and min = max - span)
  • Third, set maximum value to now+padding*span (and min = max - span)

Note that min/max/span in these cases refer to the state of the time axis as shown, which may be distinct from any persisted settings.

Non-streaming data (predictive, historical, generated, etc) does not need to do anything when unpinned.

The above is designed with Jump mode in mind, but per Tom, this is okay for Scrunch mode as well (since panning+Scrunch is a very atypical use case, and it's not clear what a user would expect to happen anyway)

Contributor

VWoeltjen commented Jan 4, 2013

Discussed this with Tom & Madelyn to work out the various cases - for streaming data, the goal should be to keep "now" (as defined by the telemetry point's time service) on the plot once it's un-pinned, without changing span (and changing as little else as possible).

To this end, three things should be tried until "now" falls within the axis:

  • First, leave time axis as it appears when unpinned
  • Second, restore maximum value of time axis to what it was when first pinned (and min = max - span)
  • Third, set maximum value to now+padding*span (and min = max - span)

Note that min/max/span in these cases refer to the state of the time axis as shown, which may be distinct from any persisted settings.

Non-streaming data (predictive, historical, generated, etc) does not need to do anything when unpinned.

The above is designed with Jump mode in mind, but per Tom, this is okay for Scrunch mode as well (since panning+Scrunch is a very atypical use case, and it's not clear what a user would expect to happen anyway)

@DanBerrios

This comment has been minimized.

Show comment
Hide comment
@DanBerrios

DanBerrios Jan 4, 2013

Contributor

OK..but if there no live data, it's most desirable to ignore the "now" value as far as changing the time axis values when unpinned.... Do we need a control for that?

Contributor

DanBerrios commented Jan 4, 2013

OK..but if there no live data, it's most desirable to ignore the "now" value as far as changing the time axis values when unpinned.... Do we need a control for that?

@ghost ghost assigned VWoeltjen Aug 2, 2013

@VWoeltjen VWoeltjen assigned JieD and unassigned VWoeltjen Apr 24, 2014

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment