On 03/24/2012 11:53 PM, lajos kamocsay wrote: > Hi Dick- > > > Thanks for pushing me for a better solution. I spent some more time > going through the drawpanel/frame code and found a much nicer way to > do the panning with the same functions that autopan and many of the > zoom functions use. > > The key is this pretty cool function: > EDA_DRAW_FRAME::AdjustScrollBars. It does all the limit checking and > optimizes the scroll bar size, so I didn't have to duplicate all that > logic. > > The attached patch (diff from revno 3477) adds middle button panning > to eeschema, cvpcb module viewer, pcbnew and gerbview. > > > Thanks- > -lajos
-lajos, I see you found a solution to the limitless dragging option, but not the bounded cookie tray option that I described. Using your mode does not recognize the boundary of the cookie tray, and reduces the scrollbar handle size proportional to the travel outside the cookie tray. I was hoping you would support at least the mode which recognizes the cookie tray boundary, so we could compare. The reducing scrollbar handle size is a way to address the discontinuity, but also means the playground becomes needlessly large, seemingly irreversibly for that session. Do zoom levels factor into the usability of the mode I described, twice now? Why did you not code it that way [too]? Are you making a statement, or simply ran out of time? Dick > > > On Fri, Mar 23, 2012 at 9:50 PM, lajos kamocsay <[email protected]> > wrote: >> OK, makes sense. I should have time to work on it this weekend, I'll >> post my progress. >> >> Thanks- >> -lajos >> >> >> >> On Fri, Mar 23, 2012 at 11:32 AM, Dick Hollenbeck <[email protected]> wrote: >>> On 03/23/2012 09:58 AM, lajos kamocsay wrote: >>>> Hi Dick- >>>> >>>> I wish I checked my email more often... last night I already started >>>> on extending the scroll region similar to what auto pan does. Your >>>> suggestion of locking the cursor on limits would be a faster solution. >>>> >>>> Do you want me to continue with extending the region, or would you >>>> rather me just lock the cursor as you described? >>>> >>>> By the way I wasn't implementing anything with ctrl/shift, I was just >>>> referring to behavior of the mouse wheel scrolling function, that's >>>> already in kicad (ctrl+wheel scrolls one axis, shift+wheel the other). >>>> >>>> Thanks- >>>> -lajos >>> -lajos >>> >>> My only concern about extending beyond the scrollbar limits is you impart a >>> discontinuity >>> when the user switches back to the scroll bar with the mouse, under >>> circumstances of >>> having travelled beyond the scrollbar limits. Then moves back into the >>> normal range. >>> This concern is what directed my original thinking. >>> >>> Maybe you have a solution for this. If you have something you really like, >>> or want to >>> offer TWO solutions, controlled with an >>> >>> #ifdef >>> >>> your patch can also simply do it both ways. We can manually supply the >>> #define for >>> testing, don't think its worth the time to put that #define in the >>> CMakeLists.txt >>> configuration yet, or ever. I don't think this raises to the level "we >>> need to support >>> both ways long term". We will probably come to a conclusion on which is >>> best and stick >>> with one. >>> >>> If you still have a lot of work on the harder one you mention, then maybe >>> skip it. >>> >>> Dick >>> >>> >>> >>>> >>>> >>>> On Fri, Mar 23, 2012 at 2:03 AM, Dick Hollenbeck <[email protected]> wrote: >>>>> -lajos >>>>> >>>>> You might not drink milk, here is a better metaphor. >>>>> >>>>> Put a rectangular cookie tray on the table. Place a rectangular cutting >>>>> board in side the cookie tray. Tell the cookie tray to never rotate. >>>>> >>>>> Jam a thumbtack anywhere into the cutting board. Move it around the >>>>> interior of the cookie tray, it never rotates, but can hit one or both >>>>> walls. Your thumbtack never moves, it is your handle. You release the >>>>> middle mouse button, the thumbtack comes out, you can move it and >>>>> re-insert it anywhere on the cutting board and try again. >>>>> >>>>> :) >>>>> >>>>> I know you eat cookies. >>>>> >>>>> >>>>> >>>>> >>>>> _______________________________________________ >>>>> Mailing list: https://launchpad.net/~kicad-developers >>>>> Post to : [email protected] >>>>> Unsubscribe : https://launchpad.net/~kicad-developers >>>>> More help : https://help.launchpad.net/ListHelp _______________________________________________ Mailing list: https://launchpad.net/~kicad-developers Post to : [email protected] Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
