forums.silverfrost.com

[dfwlms]

Will an FTN95 program with ClearWin+ run properly on a Windows Tablet PC? For example, can one touch a displayed button on a tablet's display and produce the same response as if he had mouse-clicked the button?

And, can the orientation of the program's display be rotated so as to appear as a portrait instead of a landscape?

In general, does the FTN95-ClearWin+ program developer have to incorporate special measures for his program to run on a tablet PC?
_________________
David Williams
Retired Aerospace Engineer
Huntsville, Alabama

[PaulLaidler]

I don't know the answer to this question.
It will probably be a matter of waiting for someone to try this out and discover the related issues.
There is nothing in ClearWin+ relating to touch screens but maybe this is handled by the operating system.

[Robert]

I have a Windows 7 laptop with a touch screen. In general software 'just works' with touches looking like mouse events. The problem come with intricate user interfaces where the width of a finger can cover more than one control.

[LitusSaxonicum]

I think that we are all guessing before someone tries it in earnest, but a brief look at MSDN suggests to me that there is a whole new software interface for gestures in Windows 8.

Robert's observation that a touchscreen enabled Windows 7 device responds to touch like a mouse is useful, and indeed, you can overcome the lack of precision from a fingertip by either using a stylus or (up to a point) by having bigger icons, which probably means abandoning %bt buttons for rather larger toolbar buttons or clickable bitmaps or GIFs. It does seem to me that touchscreens in Windows 7 are handled via the mouse driver.

However, the new gestures are unlikely to be handled by a mouse driver, or at least, by existing mouse drivers. For example, the 'pinch' gesture or its reverse (expanding pinch?) gives two start coordinate pairs and two finish coordinate pairs, whereas (say) a box select has one start coordinate pair and one finish pair. The mouse driver always reports one coordinate pair (x,y) only.

When programming in FTN95/Clearwin, we cope with box select and line drawing by calling SET_GRAPHICS_SELECTION@ and then interpreting the next mouse operation in terms of movement and clicks as the start and finish coordinate pairs using GET_GRAPHICS_SELECTED_AREA@. The graphics selection mode is not always implemented, as it would otherwise get in the way of other mouse actions.

To implement responses to various gestures, presumably the processing by Clearwin+ would need to get smarter, decyphering all the gesture possibilities, reporting back the gesture type by using CLEARWIN_STRING@( CALLBACK_REASON) and then using another routine to get the coordinates if the gesture was 'PINCH_GESTURE' or 'REVERSE_PINCH_GESTURE' (say). (suggested variant of GET_GRAPHICS_SELECTED_AREA@ might be GET_PINCH_EXTENTS@).
A sideways or up-down flip would be like a click-n-drag operation without drawing anything, something I suggested to Paul might be a useful enhancement to SET_GRAPHICS_SELECTION@.

I would find it useful to have graphics selection permanently enabled, even without full gesture recognition, and to have an area selection reported with CLEARWIN_STRING@( CALLBACK_REASON).

As well as making the buttons and icons bigger, a proper touchscreen response is going to have to cope with a very low precision, and new user interface paradigms seem to me to be vital if the user experience is not to be awful.

Eddie