Armel Asselin: > - having a separate notification for SC_START_ACTION is not necessary. > Rather than this the SC_START_ACTION should/can be melt into the first true > action that happens for a transaction > - BeginUndoAction must break coalescing (what it already does, no?)
If there was an explicit break coalescing call, say BreakUndoSequence, would you use it? I've been trying to find a good name and that's about the best so far but maybe someone has a better term. > It means that: > - BeginUndoAction code should remain as earlier code > - I must find a way to detect during the InsertString/DeleteChar stuff that > the action is the first one of a transaction (either because at top level > and not coalesced or inside a transaction and the first), and send > appropriately the SC_START_ACTION notification flag in the action > notification Isn't that already done in your patch with the implicitBegin? I don't like implicitBegin as a name and something like startSequence would be better as startAction is already used. Neil _______________________________________________ Scintilla-interest mailing list [email protected] http://mailman.lyra.org/mailman/listinfo/scintilla-interest
