[BRLTTY] Framebuffer terminal emulators

[Mario Lang]

Samuel Thibault <samuel.thibault at ens-lyon.org> writes:

> Mario Lang, le Mon 01 Dec 2008 11:12:02 +0100, a écrit :
>> Samuel Thibault <samuel.thibault at ens-lyon.org> writes:
>> > It was considered "ugly" by the debian-boot people to expose things
>> > via shared memory.
>> 
>> What did they propose as an alternative?
>
> Nothing.

Thats unacceptable in my opinion.  If we are being criticized for
the way we implement something, we should at least get a hint
how that person would find it acceptable.  Just blocking major accessibility
issues by throwing statements like this around isn't helpful at all.
The Linux kernel does currently basically the same with /dev/vcs.
I'd like to see real, valid reasons why exporting a SHM segment
is unacceptable as a solution here.

>> > Now, AT-SPI people would say "just implement the AT-SPI interface!"  I'm
>> > not sure we really want that.
>> 
>> I definitely think thats the wrong way to go, because of
>> the overhead involved.  This forces AT-SPI into text-mode
>> world.  Besides, its surely much harder to implement
>> without any apparent gain.  Besides, then, someone might
>> call it ugly because of its CORBA dependency.
>
> Well, AT-SPI is being ported to D-BUS,

Yes, but this effort will still take a few months.  Besides, that doesn't
really reduce the bloat, it just replaces CORBA with D-Bus.

> but I too think that it's a quite heavy dependency, particularly
> since we'd want to have the framebuffer terminal support in things
> like Linux installers...

Agreed.

-- 
CYa,
  ⡍⠁⠗⠊⠕ | Debian Developer <URL:http://debian.org/>
  .''`. | Get my public key via finger mlang/key at db.debian.org
 : :' : | 1024D/7FC1A0854909BCCDBE6C102DDFFC022A6B113E44
 `. `'
   `-      <URL:http://delysid.org/>  <URL:http://www.staff.tugraz.at/mlang/>