Open dglazkov opened 1 month ago
Hmm, could it be like a "Run Component" though? Like, let's re-run this part of the run?
I think we should possibly make a distinction between the debuggable state of affairs when the graph is read-only versus not. So:
So it's like "if you want to edit this we need to create a copy of the board here" aka Save As... first then we can treat it like any other board? Or, if we want to a bit more lenient we'd need to create a temporary board in memory, i.e., not on a board server / provider anywhere, but then we run the risk of someone accidentally making changes and then losing their work.
I think we should possibly make a distinction between the debuggable state of affairs when the graph is read-only versus not. So:
- If you open a graph you own you can update edge values, run the board and so on. The runs in this state are malleable inasmuch as you can go back and forth as needed.
- If you open a historical run you can do nothing besides navigate the historical values. If you want to edit edge values and so on in this state we have to take you to something that is like the first state?
That sounds right!
For run boards and other types of boards that can't be edited.
Maybe remove the “run board” for read-only run tabs, too? Usually, these will be historical data, and "run" is kind of weird there.
Hmm, could it be like a "Run Component" though? Like, let's re-run this part of the run?
Related: #3150