Comment by minimalist | Hacker News Reader

Comment by minimalist | Hacker News Reader

This is good! Hopefully Purism can build on their recognition to funnel resources toward FOSS-mobile applications because this will help us all.

This reminds me of something else. What would be the best model for a FOSS "app store" with mechanisms for compensation? Lowering the barrier for transactions to FOSS projects really helps their survival and sustenance by individuals who cannot contribute code.

Here's a very half-baked thought: One can imagine users supporting community servers (matrix, activitypub/mastodon, etc.) and applications through a cryptocurrency wallet that is loaded in the same way that people load their g--gle/a--le wallets. One can also imagine applications that can tie into this wallet to compensate publishers based on some model (a transaction API like g--gle/a--le or something like what brave is doing with the BAT). My question is what are the failure modes of such a model?

For example, this still wouldn't stop the perverse incentives of:

- Content publishers going for the types of bait that generate the most clicks. Clicks=compensation regardless of it's from ads or some sort of attention token.

- Application developers casing addiction mechanisms that go after whales. The same loot-box mechanisms can be implemented with a FOSS game that uses a crypto wallet API.

I'd be happy to use an app store that tracks active screen-time per app and disburses a monthly "subscription" proportionally to the developers of all the apps I use.

I don't have any concern about the type of currency used to pay.

A quick check of my screentime over the past few days shows only two apps which wouldn't benefit from this: DAVx and Syncthing. Both of these are essentially daemons I configured once, and ideally never see the app interface of ever again until I get a new phone. Otherwise, apps I benefit from are represented in the screentime of the past couple days.

I've started developing something like that, and I started by noticing like you that basing on screentime isn't fair to apps that are purely in background, so I tried to compensate for that.

Then I found that some apps, even though they were useful, didn't actually do much under the hood. Like say Frost for Facebook which is mostly a webview with few extensions. So I wanted to add a section for things that are mostly a "viewer". An additional example of such a thing is a media player: if you look 10h a day at videos, it's thanks to the content, not to the viewer.

And then, there are behaviours you maybe don't want to trigger, like getting people addicted. 2048 using the screentime metric would have gained a lot of money (and it does deserve a lot!), but IMO people who got addicted to it should give them less.

And what about apps that have paying variant on Google Play? Which deserves more money, the app that is simply free on fdroid and Play Store, or the app that is free on fdroid, but paying (or free reduced) on Play Store.

So, I spent few days trying to make something fair, and at the moment I believe that the most fair I can do is to give equally among all apps that I've used in the last 7 days.

Yeah, there's plenty of edge case handling you might want to do.

My first pass would be to do it based on screentime.

My second pass would be to give the screentime ratios to the user, allowing them to tweak the ratios themselves.

> My second pass would be to give the screentime ratios to the user, allowing them to tweak the ratios themselves.

Absolutely. Look at what Humble Bundle does, where you can tweak the proportion of your payment that goes to various things. That's still quite novel.

Give equally among all apps used in a time period sounds like the best metric.

In fact, I would argue that some of the most useful apps might be ones that we use less - because they accomplished the required goal.

"we use less - because they accomplished the required goal."

Exactly.

This may result in applications trying to keep you hooked in, addicted, or having unnecessarily long processes.

Subscriptions make sense for dynamic content, and service. Fantastical and 1Password are examples of subscription-based platforms which were once buy to play (term from games industry).

However if you self host the data (you probably can and should) or sell your soul to the devil (Google etc) you paid for hosting. So a subscription for the software doesn't make sense from customer PoV. You essentially don't pay for service, compared to Disney+, World of Warcraft, or Netflix.

I wouldn't want any monitoring of which applications I use. In any case, such an incentive would make developers of apps that should have quick, fast actions do things to make the apps slower. For example, if it takes a user 1s to use the app, they'll add a 0.5 sec animation to open it and 1.5x their money and so on.

And those seconds will be dominated by the tens of minutes I spend in other apps per day. And Android and iOS already monitor which apps you use. The calculation of ratios could easily be done locally.

> And Android and iOS already monitor which apps you use.

I know that. I don't want it as I stated above.

> And those seconds will be dominated by the tens of minutes I spend in other apps per day.

That was just an easy to understand example. The real money will be in maxing out total screen time. Such examples exist as Facebook, Netflix, or advertising driven platforms that hyper-optimize for watch-time or "engagement" to our detriment.

> I'd be happy to use an app store that tracks active screen-time per app and disburses a monthly "subscription" proportionally to the developers of all the apps I use.

Screentime is a really, really bad proxy for usefulness though. Weather apps, email clients, maps (without turn-by-turn directions), the alternative app I use to control my Hue lights, ... will starve, while the Kindle app and social media will get showered in money. That alternative camera app, hardly any screentime at all. My to do app shows me lots of notifications but gets very little screentime. Netflix however...

also ... "tracks" - that will be gamed and privacy abused.

I wonder how other "all-you-can-eat" stores handle this.

apple music, spotify, netflix, audible, kindle unlimited, etc.

Isn't apt-get, pacman, emerge, etc. more or less equivalent to an app store ?

I don't think there's necessarily a need for projects to be funded or paid for - giving back to the mobile ecosystem the same way we give back to the desktop ecosystem seems like a pretty fair way to go.

Not really, because they have no discovery mechanism. (And no, a search command that kinda sort matches on description strings doesn’t count)

There are a few gui apps that leverage tags for browsing dpkg/apt?

In what way? I don't see any fundamental difference between discovering GNU/Linux packages and Android ones (in the "stores" I've used -- f-droid, /e/ apps, and Aurora).

Why it doesn't count? It works perfect for me. How else do you search for apps in app store? New, popular?

Algorithms that match intended words and synonyms over verbatim pattern matching. Seriously, apt search sucks because if you are searching for something that does not include the word you use, it falls over completely. Just like every single time I try and brew install a package I sort of remember the name of, it will just not give me anything.

This could be an area of improvement. E.g. I don't think adding metadata to packages would be particularly complicated and that could help (e.g. using tags) with discovery.

It's interesting to read the definition of a linux distribution on wikipedia:

"A Linux distribution ... is an operating system made from a software collection that is based upon the Linux kernel and, often, a package management system."

When I read that, I realized just how fundamental the package manager is and it makes sense.

> the same way we give back to the desktop ecosystem

How do we give back to the desktop ecosystem?

Writing small software scripts, posting them to the public, making small improvements to the distro of your choice, etc.

This is good, but certainly not enough for a thriving ecosystem.

I'd be interested in a Purism app store that charges a small fee per app "purchase". The apps would be Libre so nothing stopping people packaging/compiling etc themselves (or getting from an alternative source) but to me the convenience of having that done already and curated/integrated would be worth a couple of bucks per app and even if it was optional, just having the default be paying a small fee might give enough of a nudge in the right direction.

I'll pretty much pay for anything which takes the packaging out of the equation. There's value in not having to build and distribute my own stuff (and those mechanisms need financing).

On top of that, I'll gladly pay for hosted versions of self-hosting systems. Nextcloud, Mailserver with DNS, gitea, drone/jenkins, piwik, etc.

Sounds counterproductive, to pay someone to host something that is meant to be hosted by yourself, but most often, I just need the option to move to selfhosted when time arrives.

Sure, I can host my own piwik, or gitea, or nexcloud (I do) and even (painfully) manage my mailserver and DNS with mailinabox. Out of principle. But the TCO of that is mountainhigh, if I'm honest. Tucking someone €9.99 a month to deal with one of these for me, would save me a lot, in the grand scheme.

Edit: what I'm trying to say: I'm convinced there is a businesscase there. There's money to be made in this area of FLOSS, which can trickle down into the upstream projects or be a businesscase as funding for the projects themselves even.

Then you might consider buying Librem One services by Purism: https://librem.one.

> Hopefully Purism can build on their recognition to funnel resources toward FOSS-mobile applications because this will help us all.

What about FLOSS-mobile applications?[0]

- FOSS is "Free + Open-Source"

- FLOSS is "FOSS + Libre"

[0] https://opensource.stackexchange.com/questions/262/what-is-t...

I've always thought the F in FOSS was for "Free as in Freedom" not "Free as in Gratis" but yes, FLOSS was the intention here, thank you for the reminder.

Always thought the same. By that definition I care much more about LOSS than about FOSS or FLOSS but I don't see that used ever...

This is just an annoying result of OSS projects squatting on FOSS, re-muddling the waters about what "Free" was supposed to mean (free as in freedom/"libre"). The original dichotomy was just OSS / FOSS. I guess the resulting division is precise (opensource you pay for / opensource and gratis / opensource and free as in freedom), I just find "floss" somewhat gross and just say Free Software.

The "L" part was an effort to include all the speakers of Romance languages who say "software libre" or "software livre". In my experience living in South America, very few people say "código abierto" or "fuente abierto" at the Linux conferences except if they are companies presenting business plans. For most, they don't see the need for the division between "free software" and "open source", so FLOSS is a more inclusive term to cover more of the world.

My personal belief is that we need the discourse of both the OSI and FSF because they appeal to different segments, and we need both the business community and people care about user rights to be in the movement if we want FOSS to succeed.

This is what Coil does for the web: https://coil.com/

https://qbix.com/token

Example of what can happen

Reply on news.ycombinator.com