Angus> BufferView_pimpl.C: Angus> I like the look of your changes here. The points that Angus> follow aren't a criticism of what you've done. They're Angus> thoughts on where all this should go in the end.
Abdel> I agree with all what you are saying below but I'd like Abdel> to point out something first: I am taking an evolutionary Abdel> approach. Multiple views will come after. If I derive too Abdel> much from HEAD all my work will be lost so my goal for Abdel> now is to put all this in a working shape (without Abdel> additional functionality) so that I can merge my branch Abdel> into trunk. Then, we can start another short-lived branch Abdel> for the multi-view case. Sure. But it's importatnt to have know where you're going when you set out on a journey or you'll get lost... Abdel> I am moving the code step by step. The first step was to Abdel> move the WorkArea creation out from the Bufferview. Then Abdel> I agree with you that we have to transfer LyXView Abdel> functionality to BaseFrontend. The Qt4 frontend will Abdel> then manages the WorkAreas and the Views. It will be interesting to see your design when you come to think about it properly. The devil is sometimes in the detail. Abdel> There is a lot of Cut&Paste indeed but making Abdel> everything fit together was quite a hard job. Of course. You'll find it easier to sell people on your ideas if you can given them the big picture and tell them which bit you're biting off. FWIW, André proceeded in a step by step manner in his clean up of mathed. The clean up took more than a year but he sold us on his architectural overview fairly early and thereafter was allowed to be fairly autonomous. He committed his changes without having to jump through the hoops that (I suspect) you feel you have to jump through at the moment. He could do so because Lars was comfortable with where he was heading. Have a look at the README in mathed. Put together something like that and sell that to the rest of us. Angus
