You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Diffing with the initial notebook state can be significantly bigger than diffing with the "previous" state, especially if there is a file input somewhere. In this proposal we would send both the current C and the previous P state values and caclulate the diff between P and C, then send this diff back to the client.
Given that state Pmust be already calculated by the same server, it could be memoized already somewhere in the server, to avoid duplicate calculations (given no rand(), of course)
Note that PlutoSliderServer currenty always diffs to the initial state of each notebook.
The text was updated successfully, but these errors were encountered:
Diffing with the initial notebook state can be significantly bigger than diffing with the "previous" state, especially if there is a file input somewhere. In this proposal we would send both the current
C
and the previousP
state values and caclulate the diff betweenP
andC
, then send this diff back to the client.Given that state
P
must be already calculated by the same server, it could be memoized already somewhere in the server, to avoid duplicate calculations (given norand()
, of course)Note that PlutoSliderServer currenty always diffs to the initial state of each notebook.
The text was updated successfully, but these errors were encountered: