[BRLTTY] Default braille driver and device.

[Jason White]

On Wed, Oct 12, 2005 at 12:28:03AM -0400, Dave Mielke wrote:
 
> What about changing the default braille driver to "auto", i.e. that
> autodetection be performed? If so, what about if one or more drivers are built
> in (configured with --with-braille-driver=)? Should the autodetection be then
> limited to those drivers which are built in, or should full autodetection still
> be performed?
On the first question, what proportion of braille displays can be detected
correctly? Are there any actual reports of displays or other devices entering
undesirable states as a result of auto-detection? If there are, maybe it would
be better to allow, as a compile option, auto-detection to be suppressed for
certain devices. Restricting it to drivers built into the binary would be one
way of doing this.
> 
> What about changing the default braille device to "usb:"? I like this because
> it'd mean that BRLTTY would no longer ever do unsolicited I/O on an unexpected
> device. What if the user specifies a braille driver which only supports serial
> devices? Should some special accommodation be made for this case, or, to err on
> the side of safety from an I/O perspective, should the default device still be
> "usb:" so that the user must specify the serial device which is to be used?
Again, I think that depends on our perceptions of how big the problem is of
writing to unsuspecting devices. Do we have enough evidence with which to make
a real assessment of this? To the extent that problems have been caused in
practice, there is a stronger argument for having USB as the default and
requiring the user to specify a serial port if necessary.

Note: I'm away at a conference for the rest of the week and won't have access
to e-mail until Saturday.