Re: Blind-access patch for Linux 1.3.x

Kenneth Albanowski (kjahds@kjahds.com)
Fri, 8 Mar 1996 05:38:44 -0500 (EST)


On Fri, 8 Mar 1996, Theodore Ts'o wrote:

> Date: Thu, 7 Mar 1996 21:59:19 -0500 (EST)
> From: Kenneth Albanowski <kjahds@kjahds.com>
>
> I've thought through the various permutations, believe me: yes, your idea
> works much better then most, but it requires a custom (or at least recent)
> init and it doesn't give any help if the boot locks up before getting to
> init.
>
> You'd only need it for the initial installation, because the other part
> of my proposal was to make it possible to recompile a kernel that always
> did things in "blind mode" (at least according this particular braille
> device.)

Needing it at all is the problem. My goal is to be able to walk up to any
1.4.x kernel, and with a minor change make it usable with this type of
Braille reader. Yes, once you can install stuff, this isn't a problem. If you
can install stuff then little of this is needed as patching off hardscroll is
no trouble at all.

> As far as the installation software, we could just turn off hardscroll
> for the installation disk. (i.e., use a kernel in "blind mode" by
> default. Turning off hardscroll is slower, but there aren't any other
> bad effects.

Again, unless all kernels start out with hardscroll disabled, this
doesn't serve the purpose.

Perhaps I'm arguing too hard for this -- perhaps the "blind=1" parameter is
sufficent. I think that allowing the -2 or -4 video mode to select the
blind_flag is a useful tool, as it is easy to patch a kernel to vidmode -4 or
leave it set at -2 all the time.

> - Ted

--
Kenneth Albanowski (kjahds@kjahds.com, CIS: 70705,126)