Can we please make a viable (federated!) amazon alternative? I have an idea!
-
What about the online food ordering market. I reckon that might be an easier first step than consumer products. Here in the Netherlands JustEatTakeaway has a market share of around 90% and requires restaurants to give them a 14% provision. Restaurants don't have much of a choice, if they're not on there they miss out on a huge part of the market, it's like they don't exist. Why don't restaurants unite and develop a FOSS protocol that let's them federate, so the consumer has a central place to browse the food delivery market, but simultaneously makes the providers independant because they can run their own instance if they please. Have these types of ideas been pitched to branche organizations? Restaurants have a clear interest to develop this to free themselves from the platforms with a monopolistic venture-capital-driven strategy.
-
I fully agree that this would be a valid application. The reason any company doesnt adopt such strategies is the cost of pioneering it. Most companies who spearhead such an idea want it to pay off -> proprietary. Also most people are specialized in their industry. Developing an app is not native to food industry for example.
-
@muntedcrocodile Exactly why was all the effort done to separate logic and presentation and visual design if we are now back at embedding the logic in the presentation ? IDK I need to get used to that idea. Might be just old and tired...
-
Since this was another round of no additional input, I'll repeat myself too:
People have already suggested that. But thanks for participating.
-
F [email protected] shared this topic
-
I wont correct you since I'm not the authority. I think your point is valid.
In my idea, the shop you visit - lets call them computer store - will sell you a range of computers, some of their own assembly, with normal margin, like its done today. What changes is that they partner with another shop (or many) that sell adjacent products. That could be a desk for the computer, software or other products. Those products are manually federated, ie the partners have been vetted by computer store. If you buy the computer, the seller makes their typical margin. If you buy the desk, no matter if additionally or exclusively, they will only manage the order process and payment. The rest will be done over classical dropshipping. Meaning the original desk seller will handle everything after the sale has taken place. Same as amzon does with many of their products, same as aliexpress and ebay but better than ebay because the computer store owner keeps control of the vendors they partner with. They receive a small fee only which would not be enough on its own but they arguably dont have any work besides processing the order.
-
Since this was another round of no additional input, Iโll repeat myself too:
I don't think so. I also can imagine you moved on to ethical business and suggesting ideas because you had personality conflicts where people actually do something.
-
Very much decentralized. Just with the caveat that payment decentrlization needs its own project. Successful foss software projects typically have a narrow scope and concentrate on them. Feel free to do the payment part in an adjacent project so that we dont have yo rely on stripe and paypal.
-
I like htmx cos ur entire application state is in one place.
-
This is incredibly valuable advice! Thank you so much!
My current stance on federation is of course opt in and requires the main seller to trust the downstream vendors.
The main point is that this already happens for a large portion of thing you can buy. I sell computers and adjacent services, classical system integration if you will. Of course I have to buy the systems from vendors and resell them to my customers.
Many system integrators have shops where some of them rely on custom integration of vendor apis. Take minecraft server sites for example that have an automated integration with a hosting company's api (eg hetzner). you as a customer just order a server, their automation makes the order processing with hetzner and provisions the server for you.
Now make this over a non custom but standardized api, eg activity pub.
I might still be overlooking stuff but from a technical standpoint this should be doable. The legal aspect is interesting, although I think this could be done similar to already existing resellers.
Feel free to point out flaws obvious to you. I appreciate your feedback massively.
-
I'm very happy you reveal your actual intent by personally attacking me instead of taking the hint. Good bye.
-
You don't seem to understand the retail operations of Amazon. They provide logistics and marketing services to retailers, they also directly compete against those retailers because those retailers can't do better at logistics and marketing without using Amazon's services.
-
You're missing that these points have already been adressed in a lot of other comments and have been stated way more constructively.
Of course having a whole logistics setup in place will be far superior to only doing dropshipping. But this is a whole different (additional) project. It absolutely has it is place. What I'm dismissing is the claim that the idea is dependent on somehow cloning the arguably much more expensive and complex parts of amazons business.
Again, i do agree that amazon has a huge machinery in place. But I also wish to discuss things without being treated dismissively myself.
-
Thank you for this friendly and encouraging offer. This goes a long way sowing trust. I feel a lot more positive about looking into it now. Have a nice weekend.
-
no sketchy pricing based on bullwhip procurement.
Walmart's procurement has been abusive to their suppliers (who often go out of business because of their relationship with Walmart) for decades. I think you may need to reassess your perception of their procurement strategy.
-
Sorry to bother, good luck.
-
Amazon doesnโt handle shipping for a lot of the things they sell.
This is false. Very few products sold via Amazon are shipped independently from Amazon's logistics services.
-
how to process payments? how to ship goods? how to handle refunds? how to handle contestations?
The problems are solvable, but the solutions taken together are couple times as complex as Amazon itself. This translates to cost. Which is naturally the reason Amazon came to existence earlier than that solution.
I think that layers of storage\messages and actual logic should be firmly separated, an instance going down when someone wants a refund for an operation that involved it seems not good enough. If the operation is a cryptographic contract with an escrow, and "instances" are just servers providing message storage probably privileged for some users (might be members of a community, might pay for that storage, that's lower layer anyway), this is less of a problem. But that's not a federation.
By the way, however I dislike OP's attitude, if you suggest this idea like a federated ads and reviews platform, it becomes useful.
-
a) Youโve never run a business
They might have run a small business or been present in a bigger one in management position, doing their own job well enough to avoid painful understanding they don't get it as a whole. Arrogance is not always cured by experience, actually I doubt it's ever cured in humans and we all have it.
b) youโre more interested in fantasizing than a realistic conversation.
That much was clear from the very beginning, I tend to have such ideas too, but I have BAD and thus mania periods.
-
I think it makes sense. It would allow a decentralized unified search across all stores. With Lemmy I can search posts as long as the instance is federated. With this I could find products.
-
In this example instances are stores, stores are users in instances? How stores are protected to be defederated by competitors (we are talking about money and making a living here).
What it adds to just a simple centralized service that any store can join. If you don't want it to be another amazon, make that service a coop. or some kinds of non-profit that it's paid by the stores that want to become part of that.
I think here we are in the classic conundrum of "a solution in search of a problem".
Fediverse and ActivityPub is cool, but it's a social media thing. And decentralization is cool when needed, for instance social media. But it doesn't have to make sense for every use case.
For what's being proposed there's zero actual need for decentralization or ActivityPub.
