Speakup in Linux Open Source Summit

[Janina Sajka]

I can confirm Gregory's point that the internal ISA card was noticeably
more performant as compared to a serial port connected device.


I would add two additional facts from the historical perspective ...

1.)	Especially with the internal ISA, but also with a serial device,
one could debug a not quite working installation. The system might not
boot, but the blind user could review on screen messages until the
problem became evident.

2.)	This was often necessary because one had to compile one's own
kernel in the early days of Speakup, starting with a download of kernel
and speakup source, followed by the proper patch command, followed by
the make config. There were no prepackaged kernels with Speakup until
the early 2000's.

Best,

Janina

Gregory Nowak writes:
> On Fri, Mar 29, 2019 at 04:04:58PM +0100, Samuel Thibault wrote:
> > Hello,
> > 
> > Okash Khawaja, le ven. 29 mars 2019 11:04:35 +0000, a ecrit:
> > > 1. Title: I am thinking "Linux Speakup Architecture and Status" but it
> > > sounds mundane.
> > 
> > It could be
> > 
> > "Linux Speakup makes Linux talk to users: past and future"
> > 
> > > The major portion of the talk will be about architecture of speakup's
> > > kernel code, but see the abstract for clearer idea.
> > 
> > 
> > > Abstract:
> > > ---------
> > > 
> > > Linux Speakup started out as a kernel based screen reader. It has now
> > > grown to encompass a number of related projects. According to its
> > > website[1]: "The speakup project is basically a bunch of blind people
> > > who like messing around with Linux and writing cool and, hopefully
> > > useful, software."
> > 
> > It is not really so accurate nowadays. The Linux speakup community is
> > around the Speakup reader mostly.
> > 
> > 
> > > This talk will cover three aspects:
> > > 
> > > 1. History
> > > 2. Architecture
> > > 3. Current status and future
> > > 
> > > Speakup has been around for a while now. The first section will cover
> > > roots of the project and will take a look at its community.
> > 
> > It would be worth starting with mentioning that speakup was initially
> > written by a blind person (which is awesome!)
> > 
> > Then explain how it was at the time: people would either plug an
> > external device or even put an ISA card in their computer, which gives
> > them an additional serial port. That explains the initial way that
> > Speakup accessed serial ports by hand.
> 
> The ISA cards also provided an internal speech synthesizer. This was
> versatile, since I recall the internal cards being more responsive
> than the external serial synths were on older systems. This also meant
> an internal card like the DoubleTalk PC could be used in situations
> where an additional serial port would cause an IRQ conflict, since the
> DoubleTalk PC only uses I/O addresses for communication. This could
> also tie into the to-do section in speakup. 
> 
> > 
> > Then it'd be useful to explain that that approach became harder and
> > harder to maintain, and didn't allow USB, PCI serial ports etc. and then
> > you took the time to settle a proper tty discipline :)
> > 
> > The access to the console content and snooping the keyboard was also
> > initially not very neat, but over time I got to make Linux include
> > proper interfaces for them, so that Speakup now does things the proper
> > way.
> > 
> > > Benefits to Ecosystem:
> > > ----------------------
> > > 
> > > From technical standpoint, speakup is a very interesting project with a
> > > an active and helpful community. Maintainers and contributors are
> 
> I would just say "with an active community."
> 
> > > extremely patient and easy going. All this makes speakup a great way to
> > > learn about kernel development. Speakup has been around for quite a
> > > while which means, for the more adventurous developers, there are many
> > > opportunities to innovate.
> 
> Paired with a hardware synthesizer, I believe speakup is still the
> only GNU/Linux screen reader available from boot up to
> shutdown. Without that, I would personally not have learned how to compile the
> linux kernel, since I wouldn't have had access to the kernel panics to
> find out why the system didn't boot to a login prompt.
> 
> Greg
> 
> 
> > 
> > I'd say you can add that it's useful for kernel developers to know how
> > accessibility typically needs to plug itself into the kernel, to get
> > screen reader shortcuts working, to peek the console content, and emit
> > data over a serial port.
> > 
> > Samuel
> > _______________________________________________
> > Speakup mailing list
> > Speakup at linux-speakup.org
> > http://linux-speakup.org/cgi-bin/mailman/listinfo/speakup
> > 
> 
> -- 
> web site: http://www.gregn.net
> gpg public key: http://www.gregn.net/pubkey.asc
> skype: gregn1
> (authorization required, add me to your contacts list first)
> If we haven't been in touch before, e-mail me before adding me to your contacts.
> 
> --
> Free domains: http://www.eu.org/ or mail dns-manager at EU.org
> _______________________________________________
> Speakup mailing list
> Speakup at linux-speakup.org
> http://linux-speakup.org/cgi-bin/mailman/listinfo/speakup

-- 

Janina Sajka

Linux Foundation Fellow
Executive Chair, Accessibility Workgroup:	http://a11y.org

The World Wide Web Consortium (W3C), Web Accessibility Initiative (WAI)
Chair, Accessible Platform Architectures	http://www.w3.org/wai/apa