Hello, this is somehow a follow-up of: http://linux-speakup.org/pipermail/speakup/2019-April/061808.html I have setup a procedure allowing to handle the same way hard synthesizers, be the speakup drivers built-in the kernel or modularized. I did that to allow users of hard synths to keep the same kernel parameter when switching from the former case to the latter. So, e.g. a user can in both cases append speakup.synth=apollo to the boot command line. In that aim I have included in /etc/rc.d/rc.S (the main startup script run in Slackware and derivatives), soon in the startup sequence, this code snippet: # We want to handle synthesizers the same way regardless whether the # speakup drivers are built-in the kernel or shipped as modules. So, # parse the command line and if speakup.synth=<model> is found, try to # load the corresponding module speakup_driver=$(grep 'speakup.synth=' /proc/cmdline| \ sed 's;.* speakup.synth=\([[:alpha:]]*\).*;\1;') if [ ! "$speakup_driver" = "" ]; then modprobe speakup_$speakup_driver start=1 2>/dev/null fi Of course modprobe will fail if the driver is built in but this won't hurt. However, this leads to a question: as now speakup handles devices connected through USB, should I handle this case with an udev rule (well, eudev here)? If yes, please someone give an example of such a rule if possible. As an aside, shouldn't spkguide.txt be updated? As an example it stills states: Note: Speakup does * NOT * support usb connections! Before someone ask: I am willing to help updating it, but am unable to list all the changes that occurred. Best regards, Didier _______________________________________________ Speakup mailing list Speakup@xxxxxxxxxxxxxxxxx http://linux-speakup.org/cgi-bin/mailman/listinfo/speakup
