@rabble Most days, the return on writing OSS software is an inbox full of hate mail. Not very appealing to me anymore, but maybe this guy will do it?
If the failure mode of OSS is lack of incentives, an easy start might be making ppl doing the work feel appreciated rather than berated
@joshisledbetter @glichfield @liron @ensvision @LensProtocol If you build an app, and it uses “distributed” data, it doesn’t really matter if any client can access it. What matters is what people are using, and if your app starts manipulating/censoring content, the only remedy is to get *everyone* to switch to a diff client. Same as web2.
@joshisledbetter @glichfield @liron @ensvision @zapper_fi “If Twitter does something shady, everyone will just switch to a new API.” The world doesn’t work like that. OpenSea banned my NFT in January, so why hasn’t everyone universally coordinated a switch to a new API yet?
@joshisledbetter @glichfield @liron @ensvision Switching cost is not about moving data, it’s about moving community. It’s not difficult to make a Twitter clone; you can even move the data. The hard part is that you have to get everyone to switch. If web3 solved this I would be able to see my OS banned NFT in my own wallet!
@joshisledbetter @glichfield @liron @ensvision Unless they create an API that mediates everyone’s view of NFTs and then get everyone to switch to that from OS’s; doesn’t matter. Until then, if OS bans your NFT it disappears from everywhere, including your wallet. Fixing it requires getting everyone to switch. Same as web2
@joshisledbetter @glichfield @liron The problem is rarely data portability (migrating from rackspace to aws), but community portability (moving from FB to anything else).
I can build an OpenSea alternative easily, but it’s meaningless unless I convince everyone to use that API instead of OpenSea’s. Same as web2.
@TierSigma @Qghoste @burnt_disk @kumatomoon The app store is full of apps that are rebranded Signal running their own servers. There are literally hundreds of them bc it’s the ~easiest way to launch a chat app.
I’ve already linked you to everything you need. There’s nothing stopping you. Maybe you just want to complain?
@TierSigma @Qghoste @burnt_disk @kumatomoon The server is untrusted. The whole point is that it doesn’t matter what code is there, because it doesn’t have access to any data or significant metadata. It’s also OSS, so if you want to run your own, you can! And then you can federate it however you’d like. Go for it!
@TierSigma @Qghoste @burnt_disk @kumatomoon OTF is an “internet freedom” grant making org. Signal got grants from them, Shuttleworth, Ford foundation, and many others so we could build OSS and non-profit
Honestly if you think that’s a lot of $ for a project like this, you probably haven’t tried building software at scale
@kumatomoon @stocksultan It's not. Signal never has access to ~any user data. There's no need for "server side encryption" (whatever that is), because what touches the server is already encrypted.
@matthew_d_green Ah, yeah I've seen this before. There should probably be something in the group empty state view for the creator immediately following a group creation involving invites which more explicitly clarifies that invited members won't see messages from before they accept.
@matthew_d_green C doesn't see any messages from before they accept (it would be weird if people sent messages to the group not knowing who was receiving them). That text should probably clarify that they'll never see msgs from before they join - I see the ambiguity in there around "until" rn.
@matthew_d_green 2. The Signal UX strives to function as it appears. If you send a msg and C isn't yet in the group, C never sees the msg. When C joins, it will first display "C accepted the invitation," and then subsequent msgs will go to them.
@matthew_d_green 1. There is not yet anything about C to display to B. Alice had to invite instead of add because she had not been approved as a connection by C, so Alice doesn't have C's profile key or profile information. So there's nothing to display to B about C, b/c nobody has it yet.
@chadloder @_NosirrahSec_ I’m not the Signal CEO anymore. I’m not sure why you think loading recaptcha into an isolated webview is sketchy, but if you build and maintain forever an alternative captcha service with equal efficacy at the same price point, perhaps Signal would consider using it instead.
@thisisnotawill web3 ppl keep telling me it “doesn’t matter” OS banned my NFT. In my experience, it very much matters, and is functionally identical to being banned from a web2 platform. None of the forces are moving away from that, all are moving towards exacerbating it. Good luck out there!
@thisisnotawill The thing that changed from web1 to web2 was not data portability, it was the introduction of a community/audience. How “portable” that is has always been the issue. Nothing I’ve seen in web3 so far helps, and in my brief experience it already exactly reproduced web2 conditions
@thisisnotawill OS is not the internet, but every client/server on the internet uses their API. If they ban your NFT, it disappears from ~all wallets, galleries, Twitter, games, etc
It doesn’t matter if other APIs exist unless everyone switches, but they will always be less complete than OS.
@thisisnotawill You can list the NFT, but does anyone want to buy it if it doesn’t work anywhere on the internet? If I don’t like Twitter, I can post my tweets on my own website instead. I can even move the data. But it doesn’t matter because the switching cost is in the community not the data.
@thisisnotawill I get that this is the mantra, but the “cost” of switching in web2 has never been about the data - it’s about the community/audience. I can switch to looksrare, but my NFT still doesn’t appear in my wallet, on Twitter, etc. To switch, *everyone* has to switch. Same as web2!
@stack1a And yet after you clear the cache, the content is still visible! Your web browser also has a cache to improve page load time, that doesn't mean a web browser is anything other than a view onto a server.