[tcmart14]

I'd be really skeptical that there exists companies who would love to be the portal for a smaller cut then Apple. At least longer term. I can see someone, just to get their platform off the ground, trying undercut for a little while. But 30% is sort of like an industry standard. Google takes 30%. There are alternative app stores for Google, but really how prevalent are they and are they really worth mentioning. Nintendo takes 30%. Steam I believe takes 30% [1]. So even in markets where alternative app stores/portals can exist, 30% is still pretty much 'industry standard.' Gog is 30%. Microsoft store is 30%. Playstation is 30%.

[1] https://www.ign.com/articles/2019/10/07/report-steams-30-cut...

[chandler5555]

why wouldn’t microsoft make a portal and do 0%. they get all of the market share and analytics and good will, plus the value of being the main “other” store. they can break even on hosting fees with some ads

[tcmart14]

Because breaking even isn't stocks going up and to the right. If shareholders aren't seeing good returns on building a portal, it isn't gonna be a good time. That is true for most of these companies. Good will and analytics alone isn't enough for investors.

[flemhans]

I'd do it for 1-2%

[tcmart14]

You would. But would the venture capital firms you would need to bank roll the infrastructure and upstart costs be fine with this. I imagine most of them, might be fine short term with low percentage (to eat market share), but eventually they'll ask, why can't we take 30% like everyone else?

I do think this is why EpicGames only takes like 12%. It isn't really good will, or it is surface level goodwill. Its market share. They'll eventually crank it up to 30 like all their main competitors.

[flemhans]

Of course 1-2% will pay for the infrastructure. It's not such a huge deal. F-droid runs on donations.

Let's hope for more competition in the future, at least in EU.

[tcmart14]

On the scale f-droid runs at, with a fraction of a fraction of a fraction of a fraction of a percent of android users, yes. As much as I would love it, I don't think f-droid could scale in such a way to really be competitive against the google playstore, eat a serious portion of Google's market share in android app stores. I hate it, but we see it time and time again. Open project requires infrastructure, which relies on donations. When its a small set of users its fine. When the user balloon up, the donations don't and the project has to shut down. Unfortunately, its only a hard core subset of users who donate.

[dlubarov]

What kind of costs are you thinking of? Payment processing could be significant, maybe 3% or a bit more if the average amount is small. But isn't that about it?

Like bandwidth becomes essentially free at a certain scale, but even if we're paying s3 retail prices, that's under $0.001 to transfer the average (~40mb or so) iOS app. Compute should be less than that.

There would also be fixed costs to develop the service etc, but Apple earns something like $80 billion / year in App Store revenue, so if a competitor could capture a nontrivial portion of that, any reasonable development costs should be pretty negligible by comparison.

[tcmart14]

Assuming you want to compete with Apple at scale or Google with roughly similar quality of service. You would probably need to stand up multiple servers around the world. Deal with the regulatory bits that each country would require to operate in. Probably also development to develop your own APIs. For instance, your gonna distribute the apps, but the apps still need to rely on Google's payment API for in-app purchases? So you need developers, if its FOSS and you get enough developers, may not be a cost sync. Authentication?

Could you do it cheaper than Apple and Google. Yes, in theory. But probably need quiet a bit of upfront capital to get it going, which might require investors who are eventually gonna wonder why they arn't collecting that sweet 30% while their competitors are.

I do wonder how that would work regulation wise. Is the store PCI compliant? How do you manage PCI compliance in a FOSS project? I do a little work on payments processing and what happens when VISA makes up a new rule on the fly that they force the payment gateways, issues a fine and the gateway companies (like World Pay) want to pass that fine on to the developers using the payment gateway? This almsot happened at my shop, we almost got hit with a $25K fine because Visa made up a new rule, tried to fine WorldPay and WorldPay tried to pass that fine onto us.

Edit: If anyone has some answers or experience to the last paragraph, please let me know! I've only dealt with payment processing on the proprietary software side and doing integrations for brick and mortar stores. So working with physical cards and payment terminals. I would be very interested to learn how this works or would work in a FOSS project.

[flemhans]

I totally agree, but I do think a few percent would cover it, still. Mostly would go towards payment processing and payout bureaucracy rather than the IP traffic and servers.