P. A. on 18/10/2009 20:46
Are there particular "UNIX" environments that you have in mind? It would be good if you can be more specific about the supported environments.

I'm struggling to work out how such a utility would be useful in isolation. Perhaps you can elaborate?

If I were to bid, the technology would support Windows, Mac and Linux based on the intel archiecture.

Paul

H. Y. on 18/10/2009 23:37
Hi Paul,

Windows, Mac and Linux will do fine for now. I realise people will be curious to know the context for this, but I would rather not elaborate at present.

A slightly fuller spec of the requirement is:-

The rectangle is no larger than a credit card, but the user must be able to change its dimensions, up to this maximum. The entire rectangle should be visible, although it should be possible to align any edge with an edge of the screen, if that is what the user wants.

The accuracy required is at screen pixel level.

The output parameters should include the pixel positions of the top left and bottom right corners of the rectangle when the user clicks on “Save”.

This output needs to be stored in the local PC such that other applications can access it and display the rectangle in the user-fixed position. We need to be able to do this, even if the user has changed the screen settings since the calibration was completed.

A suitably fast method might allow the user to drag the entire initial rectangle to the approximate favoured position, then pick up its corners and drag them independently to the near-precise required position (2 corners should suffice). There may need to be some form of fine tuning • nudging one or two pixels at a time - to achieve the required accuracy of the final calibration.

Our current implementation is in Java. It works, but it is clumsy. A novice user needs two minutes or more to perform the calibration. It does, however, include a “nudging” process, which might be useful in the final implementation.

Please feel free to ask any other questions.

Howard