2009/8/17 Art Hunkins <[email protected]>: > I'm dealing with another issue as well - one that is equally important: 4:3 > va. 16:9 displays. All my text has to be on a single screen; but 4:3 screens > will display more *lines* than 16:9. What I'm trying to do is for my script > to query for screen size (Walter showed me how - thanks, Walter!), and > depending on the answer, compress some of the text so that it takes fewer > lines. > > As a result, my script will need to determine the following: > a) XO or not - result: set appropriate font size > b) 4:3 or 16:9 - result: set appropriate font size (there *appears* to be a > difference here, unfortunately - I've yet to figure out this disparity), and > adjust number of lines and/or font size for selected text.
This is a bad approach. If filling the screen (or an area of it) is important to you, then you should render some text using the font and measure it's size, then adjust accordingly. Measure is one example of an activity that does this. Otherwise, the only thing that is safe to assume is that 1 point means 1/72 of an inch. This alone is meaningless though, you should also consider the size of the screen to see if that would feel proportionally OK, and you also have to consider how far you expect the user to be sitting from the screen. Size 10pt on the XO looks uncomfortably large, but measure it and you'll see that it's exactly the same size as size 10 on your regular monitor, where it looks fine. The difference is that the XO screen is proportionally much smaller and you generally sit closer to it. The approach I'm advocating here is that nobody is encouraged to mess with DPI (and Sugar does not touch this). Instead, Sugar ships a default font and size setting (e.g. Sans 10) but allows it to be easily overridable by deployments. Additionally the numeric part of this can be changed through 2 new buttons in the control panel: "Make fonts smaller" and "Make fonts bigger." But this has not been implemented yet. Daniel _______________________________________________ Sugar-devel mailing list [email protected] http://lists.sugarlabs.org/listinfo/sugar-devel
