Re: [PATCH/RFC] using DispatchResult everywhere

2010-01-22 Thread Jean-Marc Lasgouttes
Tommaso Cucinotta writes: > I also proposed to add a DispatchResult everywhere (I wanted an LFUN > returning a match position in a search, without any GUI-related op). > In my proposed fix, I had added a DispatchResult pointer within the > FuncRequest class itself, so as to avoid changing each and

Re: [PATCH/RFC] using DispatchResult everywhere

2010-01-19 Thread Tommaso Cucinotta
Pavel Sanda wrote: Jean-Marc Lasgouttes wrote: What seems to be a good solution is to define a new LyXFunc::dispatch(FuncRequest, DispatchResult) that just returns what needs to be done, and add change the normal LyXFunc::dispatch to be a wrapper around it that does the screen updates. This i

Re: [PATCH/RFC] using DispatchResult everywhere

2010-01-19 Thread Pavel Sanda
Jean-Marc Lasgouttes wrote: > What seems to be a good solution is to define a new > LyXFunc::dispatch(FuncRequest, DispatchResult) that just returns what > needs to be done, and add change the normal LyXFunc::dispatch to be a > wrapper around it that does the screen updates. This is what I started

[PATCH/RFC] using DispatchResult everywhere

2010-01-19 Thread Jean-Marc Lasgouttes
I tried recently to get rid og all-insets-toggle in favour of inset-forall. This works, but I noticed at this occasion that the performance of inset-forall is horrible, in particular because the screen is redrawn for each inset. This is definitely wrong. What seems to be a good solution is to def