Newsgroups: comp.windows.ms
Path: utzoo!utgpu!news-server.csri.toronto.edu!helios.physics.utoronto.ca!alchemy.chem.utoronto.ca!mroussel
From: mroussel@alchemy.chem.utoronto.ca (Marc Roussel)
Subject: Re: Mouse in a windowed DOS program
Message-ID: <1991Apr4.171426.28650@alchemy.chem.utoronto.ca>
Keywords: Mouse window DOS
Organization: Department of Chemistry, University of Toronto
References: <1991Mar31.185751.15696@uvm.edu> <1991Apr3.012535.17977@mrspoc.Transact.COM> <1434@macuni.mqcc.mq.oz>
Date: Thu, 4 Apr 1991 17:14:26 GMT

In article <1434@macuni.mqcc.mq.oz> dcrowley@sunb.mqcc.mq.oz.au (David Crowley) writes:
>    If the dos-app is windowed then you will *NOT* be able
>    to use a mouse with the app even if it supports it. This is because
>    windows does not release control of the mouse when you move the pointer
>    into the dos-app window. And anyhow if it did, how would windows know
>    how to get hold of the mouse again? (detect when the mouse got to the
>    edge of the dos-app? = not easy).

     X-Windows does this.  Besides, why is it so hard?  A GUI has to know
where a window's edge is so it can do clipping.  It also has to recognize what
sort of object the pointer is sitting on so it can take appropriate
action on clicks.  I honestly don't see why detecting what kind of
window the pointer is sitting on and taking appropriate action is so
difficult.

				Marc R. Roussel
                                mroussel@alchemy.chem.utoronto.ca