Date: Sun, 30 Jul 2000 09:42:21 +0300 (IDT) From: Eli Zaretskii X-Sender: eliz AT is To: Ian Miller cc: djgpp-workers AT delorie DOT com Subject: Re: MS mouse with wheel In-Reply-To: <003201bff99d$7c25d8a0$2b8da6c3@shelob> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Reply-To: djgpp-workers AT delorie DOT com Errors-To: nobody AT delorie DOT com X-Mailing-List: djgpp-workers AT delorie DOT com X-Unsubscribes-To: listserv AT delorie DOT com Precedence: bulk On Sat, 29 Jul 2000, Ian Miller wrote: > > On a Windows 9X system with an MS mouse that has a wheel (between the two > > buttons), I see a strange phenomenon. When a DJGPP program (Emacs in > > this case) initializes the mouse driver, the returned info says that the > > mouse has 3 buttons. > > I do not know how to confirm or deny that observation. Like I did: run Emacs under GDB, put a breakpoint in the dos_ttraw function, where Emacs initializes the mouse driver and figures out how many buttons does it have, and see what does function 21h of Int 33h returns in the BX register. Or just write a short program that calls function 21h of Int 33h, and see what it yields. The idea that this might be the problem struck me when "C-h l" showed that Emacs thought I clicked mouse-3 instead of mouse-2 for the right button. > > The problem with Emacs is that it thinks that the right > > button is mouse-3, and mouse-2 is absent. > > And you are, for example, unable to navigate a directory structure > using dired using just the mouse. Me too. Yuck! I can easily repair that, by providing a function to forcibly set Emacs' notion of the number of mouse buttons from Lisp; users could then avoid the problem by calling that function from their .emacs init file. I posted my message to make sure it wasn't something specific to the machine where I saw the problem. Thanks for the feedback. I think I have my confirmation now, and can proceed with the solution.