2009/8/5 John-Mark Bell <[email protected]>: > On Wed, 2009-08-05 at 14:24 +0200, Paweł Blokus wrote: >> Hi, >> >> Since the last progress report I have completed both the scroll widget >> and the select menu. Due to rounding errors the scroll sometimes has >> some off by one rendering issues but I am not sure whether much can be >> done about it. > > Where are these rounding errors?
The scrolls aren't aware of themselves and the box containing them, so their dimensions and positions are calculated seperately. For this reason there sometimes is a one pixel wide gap between the scrolls and a one pixel wide stripe of box content on the right of the vertical scroll. Actually, the latter also exists in the trunk with the difference that the stripe is on the left of the scroll. Maybe using round() at some places would help. Is performance on processors without a coprocessor the only reason why it's not used in the rest of the code? > >> At this point I am not sure what I should do next. I could start >> working on the keyboard navigation, but I don't know if I can reach a >> reasonable milestone as now I know my estimations of the required time >> were too optimistic while I was writing my proposal. Also, I think, >> once being reviewed, the treeview code would keep me busy in case some >> non-cosmetic-only changes are requested. But this would need the code >> to be reviewed before risc os working with it properly. Maybe there is >> some part of the code you already know you are not satisfied with? If >> so, don't wait for the main review with writing about it :) > > I've no good suggestions as to what you should do next :) Perhaps > Michael has some ideas. > > > J. > > >
