Things we don't need for ethical technology to reach the masses:
- multiple distros
- mutliple desktops
- multiple toolkits
- "the unix philosophy"
- server-side decorations
@brainblasted Beneficial to whom?
The problem with multiple platforms is that they're competing for mindshare, users, and developers. It dilutes already scarce development resources, and means there isnt's a single brand that the general public can see as the ethical alternative to the status quo.
@brainblasted Well, it depends on the goal.
If your goal is to make have fun making software, it's perfectly fine. But it's not helpful if your goal is to get rid of surveillance capitalism.
@tbernard @brainblasted at the same time, multiple vertically-integrated, high quality platforms helps make sure everyone will find what they need and won't ever go back to the current non-ethical solutions.
I'll take multiple vertically-integrated platforms (GNOME OS, KDE OS, Elementary, …) any day over multiple generic distros all doing almost the same thing in slightly incompatible ways, each breaking the upstream UX differently.
This is much larger than platforms vs. distros though, it's about overall goals and motivations.
I'm not saying there can't be multiple platforms, just that it's not helping us reach the goal of giving the masses access to ethical technology.
@tbernard @mathieu @brainblasted I'll agree that a single large "lighthouse" project/platform/banner can help draw a lot more attention from users (which is the truly scarce resource), certainly in terms of advancing the idea of ethical systems but in terms of "platforms" (as in Twitter vs. Mastodon, Whatsapp vs. Signal) not having multiple independent offerings would be outright dangerous. There can be no benevolent dictator.
@tbernard @brainblasted Wait, that toot wasn't sarkastic?
Yes, I think we absolutely need multiple platforms, brands and implementation. If there was a single platform, there'd be a single point of failure, as well as the biggest ever incentive gain control over that one platform, and thanks to the network effect, it'd be almost impossible to make something better. Diversity is good, not just for humans and ideas. Any marketplace, monetary or not, needs diversity.
@tbernard @brainblasted This is not really true.. Multiple platforms come to life because it offer something the developer didn't get. Like programming language for example. A python developer will not want to learn perl to join an existing project.
You create something because you have a vision... You may not like the previous existing project for any reason. This means that developer will either create a alternative or nothing at all.
So no... quality does not come with numbers.
Yeah, this sort of party-rule, one size fits all, technovanguard stuff skeeves me out.
Ethical tech doesn't have to be Unix-inspired, as such (although good luck escaping its influence entirely), but leaving so little room for multiplicity and diversity of approach ... <shudder>
@tbernard kill those themes! 😩
@tbernard since we are very clearly going to get neither of ethical technology for the masses nor a satisfying technical dispensation, entirely regardless of whether i'm willing to capitulate on the second, i think i will continue wanting both in equal measure.
@tbernard whatchu have against the unix philosophy?
@tbernard oh that ol' thing, not the unix philosophy i have in mind
@snwh hence the air quotes ;)
@tbernard honest question: how do you see Pure OS on the Librem laptops?
It seems to me like it's much more of a distro than a vertically-integrated platform (like GNOME OS would be)
@mathieu I'd be very happy to see PureOS (and every other distro in the conventional sense) go away in favour of vertically integrated platforms.
PureOS is just rolling-release Debian with no proprietary software. In a hypothetical future where everything revolves around platforms instead of packaging formats, it would just be GNOME OS without proprietary software.
Yeah, we might have different values here when considering ethics or 'ethical systems'
@tbernard @amenthes I disagree on the Unix philosophy part. A separate tool for each job, which does its job well and can be easily integrated is great way for dealing with complexity. Sure it does not matter much to the end user, but somebody will have to maintain your fully vertically integrated utopia. An that's much harder for one large monolith with a lot of tightly coupled interdependent parts.
@sebastian You're missing the point. The two philosophies play out in completely different discussions.
You can have completely unethical monolithic things and completely unethical unix-philosophy tools. You can have ethical monoliths and ethical unix-philosophy tools.
They are not related to each other at all. "Unix Way" is just an implementation detail.
@sebastian would you say, grep and less support each other? They happen to work with each other. They are modular. I think that's a low bar for interoperability.
@sebastian sorry, last tweet was meant to say "interoperability is a low bar for ethical software". It's a desireable trait, regardless if it's ethical software or not.
@amenthes But the original toots I replied to implies that we don't nee X to build ethical software. And we agree that the solution to ethical software is an optimum somewhere out there in the design space created by many different independent axes. Saying 'we don't need X' to achieve ethical software implies that we do not care where we end up on the X axis of our design space. That's a little short sighted to me.
@amenthes E.g. I think ethical software should not lead to vendor lock-in or network effects, which is hard to achieve if we focus all efforts on a single implementation of 'one great platform', because there will be no migration path away from it in lieu of viable alternatives to it.
Similarly having single large 'brand' associated with that platform and hence with ethical software will likely burn the 'ethical' brand as soons as the platform fails.
@sebastian I don't see "Unix way" changing any of that?
@amenthes That was more related to the multiple distros/mutliple desktops/multiple toolkits parts of the original toot.
But coming back around to your less/grep example: I can trivially replace either one of them and the replacements could even coexist as part of the same installation.
So you have at least a basic modularity to it and you are locked more to interfaces between components and less to concrete implementation of them.
@amenthes After all part of the original argument is to centralize all development so that endresult is feature complete and polished will be widely adopted. And that's in my oppoinion a very short sighted course of action.
Server run by the main developers of the project It is not focused on any particular niche interest - everyone is welcome as long as you follow our code of conduct!