From owner-freebsd-current Fri Oct 11 01:19:28 1996 Return-Path: owner-current Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id BAA27143 for current-outgoing; Fri, 11 Oct 1996 01:19:28 -0700 (PDT) Received: from covina.lightside.com (covina.lightside.com [207.67.176.1]) by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id BAA27112; Fri, 11 Oct 1996 01:19:23 -0700 (PDT) Received: from localhost (jehamby@localhost) by covina.lightside.com (8.8.0/8.8.0) with SMTP id BAA10056; Fri, 11 Oct 1996 01:19:17 -0700 (PDT) Date: Fri, 11 Oct 1996 01:19:12 -0700 (PDT) From: Jake Hamby To: sos@freebsd.org cc: current@freebsd.org Subject: Bug with moused and XFree86 Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk I was very interested to discover the recent addition of moused to FreeBSD-current. GPM was always a nifty program to have for Linux, and it's great to see that FreeBSD now provides the same functionality (with a nifty arrow cursor, no less!). However, I've noticed two strange bugs affecting XFree86 (I've consistently reproduced them with 3.1.2-S as well as the 3.1.2-G beta) when moused is running. With twm as the window manager, when I click on the title-bar, the "four-corner arrow" stays active AFTER I release the mouse button, UNTIL I move the mouse (it _should_ change back to the regular arrow immediately after releasing the button). When fvwm95 is the window manager (which defaults to click-to-focus, not focus-follows-mouse), a different bug occurs. When I click to activate a new window, any characters I type are ignored, until I move the mouse (as little as one pixel), then all the characters I've typed display. I'm running FreeBSD-current with a Microsoft-compatible serial mouse on COM1. "vidcontrol -m off" was in effect during these tests, but the moused was running (started by /etc/rc.i386). I believe these are two symptoms of the same problem. My uneducated guess (and don't even _think_ I have time to look at the source code right now :-) is that moused is filtering XFree86's access to the mouse device and somehow certain events (in this case mouse-down) block it until another mouse movement is received. I don't ever remember having troubles like this with Linux and GPM, so I'm guessing that moused, not XFree86, can be patched to fix these problems. Thanks! -- Jake