Re: [orca-list] audio infrastructure for assistive software?

[Didier Spaier <didier slint fr>]

Hello,

Actually Qemu is easily cross-platform.

At least I could install it in Windows just getting and running this installer:
https://qemu.weilnetz.de/w64/qemu-w64-setup-20190815.exe

As I don't have Windows on bare metal I installed and ran it... in Windows running in a Qemu VM <smile>

I don't have macOS installed, not even in a VM, but looking at this page:
https://www.qemu.org/download/#macos
installing it should be fairly easy.

Generic link for downloads: https://www.qemu.org/download/

Best,

Didier

Le 25/01/20 à 01:23, Nolan Darilek a écrit :
> Adding on a bit to what Didier says:
>
> Who is your target audience here? Have you actually identified a group that would benefit from this?
>
> Speaking for myself, I'm done with VirtualBox. I used to run it exclusively. Then I set up a Windows VM 
> under Virt-Manager and haven't looked back. VirtualBox's audio support was so terrible and laggy. Qemu was 
> comparatively blazing. I've even been able to do light audio editing, including near-real-time MIDI 
> playback, by using USB devices hosted by the Windows guest, which thanks to passthrough was as easy as 
> plugging in the device once the VM was booted. I think things may have gotten even better with the release 
> of Qemu 4.0 and some unofficial pulseaudio patchset Reddit was having kittens over for months until it 
> finally got upstreamed.
>
> I know you can't use Qemu and be easily cross-platform, which is my point. Why create a Linux VM on a 
> virtualization technology with such bad audio? What group are you targeting who will download and get use 
> out of such a thing? I'm genuinely not being sarcastic--I'm trying to understand what problem this solves 
> or what group it helps. And why would I want a bunch of software I won't need, when I can get the software 
> I do need as easily as typing "dnf/flatpak/snap/brew install ..."?
>
> On 1/24/20 6:10 PM, Didier Spaier via orca-list wrote:
> > Hello Rich,
> >
> > I couldn't find where to download Perkify on your website. is a
> > pre-version already available?
> >
> > Now, let me be blunt in the hope to help.
> >
> > I think you need to really use the a11y software and integrate them so
> > that Perkify can have a real added value. To provide an actually
> > accessible system, a good integration of the shipped software is a must.
> > In other words, just shipping a lot of software a11y related is not
> > enough: they should be configured to play well together. As an example,
> > even if Debian ships all that's needed wrt accessibility, as far as know
> > it is still not configured by default in such a way that you can keep
> > speech switching from a graphical environment to a console.
> >
> > The right question about ALSA and PulseAudio is rather "how do I
> > configure them so that they play well together" rather than "which one
> > should be preferred."
> >
> > Also, the process to get Perkify working looks to complicated to me as
> > it involves Perkify itself (indeed) as well as Vagrant and and a
> > pecific version of VirtualBox (which is not the best choice wrt a11y in
> > my opinion as the sound support is not in par with that of Qemu).
> >
> > If I understand well Perkify is an Image for VirtuslBox, and is huge
> > (10G?) Here is what I suggest, technically speaking:
> > 1. Provide an image of the installed system for Qemu, compressed.
> > 2. Install in it a minimum system: just what is needed to be accessible
> > with Braille and speak and include the utilities and instructions needed
> > to download and install all other packages. Indeed downloading the
> > packages will take some time, but I think that's more acceptable than
> > downloading a very big image for people with a slow and possibly
> > unreliable Internet connection.
> > 3. Provide instructions to install Qemu and a small shell script to set
> > up a VM with Perkify in it.
> >
> > I have done something like that for Slint recently: the compressed image
> > weighs only 302M. This stuff is in:
> > http://slackware.uk/slint/x86_64/slint-14.2.1/minislint/
> > cf.: the README there.
> >
> > HTH and good luck,
> >
> > Didier
> >
> > Le 24/01/20 à 22:35, Rich Morin a écrit :
> > > I've been making significant progress on getting audio output to work from Perkify
> > > (my attempt at a blind-friendly VM, based on Ubuntu, Vagrant, and Virtualbox) to
> > > macOS.  The plan is to have it work with Linux and Windows hosts, as well.  The
> > > current setup only uses ALSA and PulseAudio, but I'm planning to add JACK to the
> > > mix <g>, by means of the pulseaudio-module-jack package.
> > >
> > > I've installed a number of audio-related accessibility packages, including BRLTTY,
> > > Fenrir, Orca, etc.  I'd also like to support Emacspeak, the vOICE, and others.  So,
> > > I'd like to know if there are any common themes in how these sorts of packages use
> > > audio output.  For example, do they generally tend to use ALSA and PulseAudio?
> > >
> > > Feel free to contact me off-list or on the Perkify mailing list (perkify.groups.io)
> > > if your response seems to be veering too far off-topic for this group.  I'm also
> > > interested in finding folks who would like to help by testing and/or configuring
> > > these sorts of packages.
> > >
> > > -r
> > >
> > > P.S.  For more information on Perkify, visit its Introduction page:
> > >        http://pa.cfcl.com/item?key=Areas/Content/Overviews/Perkify_Intro/main.toml