GrapheneOS: Talks shit about Fairphone and their 'ancient' hardware on socials, not considering them as an option at all.

Also GrapheneOS: Partners with Motorola/Lenovo despite more devices in their lineup having relatively mediocre chipsets than not.
@maddy Oh yeah, GrapheneOS, the OS that's fine officially shipping sandboxed proprietary Google Play Services but will not ship a screen reader because it's GPLv3!
@nytpu
👍 'Privacy'

👎 Accessibility
@nytpu @maddy wait, the fuck? they don't ship a screen reader? they're OK with being inaccessible???
@senil @nytpu I mean, they don't shut the fuck up about privacy and security, but I never see them spotlight accessibility.

@senil Yep, their social media account has told actual blind people to "just manually download a TTS engine" which is really useful advice when it is literally impossible for someone to interact with the device to get past the initial OOBE.

(They actually do ship TalkBack since that's part of AOSP, but they're missing a TTS engine. Which is like a Linux distro including a desktop environment but no graphics drivers and saying they "have graphics")

@maddy

@nytpu @senil And I assume this and Android generally being a dumpster fire is why entities flock to iOS, 'cause I keep hearing they've got you covered in terms of accessibility (and decently well) right out of the box.

@nytpu @maddy Ohhhh my godddd that's. Annoying as hell. "We ship the thing that would let you use a TTS system, but we don't actually provide one for you to get started with, if you need one you're on your own."

Surely there's like, some basic TTS engine that could be used for at least the initial OOBE and then post-setup assist the user in installing a more appropriate one if they REALLY don't want to ship a properly featured system.

@nytpu @maddy Truly why does this as a trend keep happening. And I imagine their take is "well just patch it in yourself if you want one from the start" since that attitude is ALSO disappointingly prevalent amongst OSS in general.

Fucking gross.

@senil @nytpu For real. It's incredibly infuriating to see accessibility as a low priority until they just happen to need it themselves.

@maddy @nytpu Even then, I imagine the shitty attitude of "it falls on me to implement this fix" comes up, which leads to no consistent answer for the problem and a ton of split efforts to adding a thing that shouldn't have been a concern from early on 

Ugh.

And folks wonder why people have few choices but to stick to proprietary software, and don't back projects collectively to enable folks who literally have no alternative to use FOSS-alternatives.

@senil I mean, unfortunately there's really not. Apparently GrapheneOS said "we can't use eSpeak-NG because they don't support launching early in the boot" and then eSpeak-NG promptly implemented that (and no other Android TTS engine supports to this day) within a few months so GrapheneOS had to reveal their real reason for not including it is "it uses the GPL".

eSpeak-NG does kinda suck but AFAICT it's the only FOSS TTS engine that works on Android and is actually usable by blind people rather than being a toy project. And it'd at least let people get the OS installed and navigate enough to install a proprietary one like Google's TTS engine if they wanted

@maddy

@nytpu @maddy Why... do they not want to ship anything with GPL? Like I'm sure they have their (presumably silly) reasons but "not shipping a decent enough TTS engine to get started with" does not feel like something they should be accepting.

And yeah, that's literally just my thought. A basic enough TTS engine that gets things off the ground to enable users to install a better, likely-proprietary engine post-OOBE. It'd just have to be good enough for that purpose.

@senil Since Android AOSP is Apache licensed and apparently "we want to be a drop-in replacement for AOSP". As if they couldn't just do what Debian does/did and have separate "minimal Apache" and "complete" versions or something?

Apparently LineageOS has the same issue for the same reasons after a search…

@maddy

@nytpu
But... Linux IS GPL, and not Apache licensed!
@senil @maddy
@senil @nytpu @maddy We ship plenty of GPLv2 software and have our own GPLv2 projects. We don't include GPLv3 software in GrapheneOS because it would add restrictions we don't want applying to the OS as a whole. We want GrapheneOS to be a drop-in replacement for the Android Open Source Project without having more restrictive licensing. Sketchy companies don't care one bit about the licensing and will fork it without complying with it while lots of ones we want to work with want to avoid GPLv3.
@senil @nytpu @maddy GrapheneOS is a privacy and security focused project. Including an app which handles sensitive user data as untrusted input from other apps is something which needs to be handled very seriously. It's a high priority for us to avoid doing harm to privacy and security which is very difficult when including a bunch of attack surface. We don't want to add a bunch of memory unsafe code. We also don't simply want to have something barely usable but rather a competitive TTS app.
@senil @nytpu @maddy Our approach to GrapheneOS is not hacking things together to get it working as quickly as possible. Our approach ahs always been putting in the effort to build high quality implementations where privacy, security, usability and other goals are all satisfied. It takes time to get things implemented to meet our standards. That's just how we do things in this project. It works out very well in the long term and is largely why people want GrapheneOS in the first place.
@senil @nytpu @maddy We have a fork of the open source TalkBack which is maintained by a blind member of our community and is included in GrapheneOS. We've developed our own text-to-speech implementation meeting our requirements for better usability, out-of-the-box functionality, security and acceptable licensing. Our text-to-speech app is currently available for early testing and will be included in GrapheneOS soon. We then need to add support to our Setup Wizard for integrating TalkBack.
@senil @nytpu @maddy We were in the process of investigating a bunch of open source TTS apps and saw none of them supported the required Direct Boot support for TalkBack so we filed it as a feature request with several. It wasn't done with the intention of bundling them but rather because we wanted the apps to be improved. We found 1 open source TTS app which almost met our requirements (SherpaTTS) but it's too slow to be comfortably used with TalkBack so we decided to make our own and did.
@senil @nytpu @maddy Our text-to-speech app project is at https://github.com/GrapheneOS/SpeechServices and we're actively working on polishing it up for inclusion in the OS as the default enabled TTS service. TTS services handle untrusted input from other apps on the OS combined with a lot of the data they handle being privacy sensitive. It's also crucial for it to perform very well especially for the TalkBack role. SherpaTTS provides high quality output but is much too slow, which was the reason to make this.
GitHub - GrapheneOS/SpeechServices: Fast, efficient, and high-quality text-to-speech for GrapheneOS using state-of-the-art models running completely on-device.

Fast, efficient, and high-quality text-to-speech for GrapheneOS using state-of-the-art models running completely on-device. - GrapheneOS/SpeechServices

GitHub

@senil @nytpu @maddy GrapheneOS does not come with sandboxed Google Play but rather people can choose to install Google apps as regular sandboxed apps on GrapheneOS.

We have no problem shipping GPLv3 licensed apps in our App Store. We use GPLv2 ourselves including as the license for our Vanadium browser app. We avoid GPLv3 because companies we want to work with avoid it.

eSpeak NG wouldn't have been included if it was GPLv2 or permissively licensed as it just doesn't meet our requirements.

@GrapheneOS

Using FUTO (based on openAI whisper) with local trained data without network access works fine for us...
Can be found @ fdroid

@senil @nytpu @maddy

@FamilyCyclist @senil @nytpu @maddy That's speed-to-text rather than text-to-speech. We're going to be providing both as fully open source implementations. FUTO keyboard is source available rather than open source and Whisper is a so-called open model rather than an open source model meaning the training data isn't available and it isn't possible to redo how they generated it. We want these to be open source and that's what we did for our existing text-to-speech app already.
@GrapheneOS @senil @nytpu @maddy I am using SherpaTTS myself for quite some time now, and besides the bad UI it works great for me. And for STT I use FUTO. But I'd be more than happy if you could ship a TTS & STT with the OS for better integration into the system.
@nytpu you realize gpl v3 is a restrictive license right?