Does someone synchronization peers between anonymous users using Gun?
I use fingerprntjs2 to generate a key for Gun
let fingerprint
Fingerprint2.get({}, function(components) {
const values = components.map(function(component) {
return component.value
})
fingerprint = Fingerprint2.x64hash128(values.join(''), 31)
})const peer = new Peer({
host: 'localhost',
port: 9000,
path: '/',
debug: 3
})
if (peer and fingerprint) {
gun
.get('peers')
.set({
fingerprint,
peerId: peer.id
})
.bye()
.put(null)
} gun
.get('peers')
.map()
.on(function(data, id) {
// this.put(null)
// console.log('data', data)
if (data) {
if (data.fingerprint && data.peerId) {
if (data.peerId !== peer.id) {
// console.log('data', data)
connect = peer.connect(data.peerId)
setTimeout(function() {
connect.send(`Hello world from ! ${peer.id}`)
}, 1000)
}
}
}
})bye().put(null) does not work.
@rm-rf-etc w00h000! Phew, half a year out! Commitment! :D
@mimiza w00h000! :D This is a good idea, but it becomes a "standards" thing because we need pretty much every machine to run the code to enforce the rules, and more code = more bloat, etc., it sounds like you can get away with doing what you want just via a user space, no?
@finwo nice answer!!!
@trosel oooooooo! I like that idea.
@narnikgamarnikus yes, do you need the data to merge tho (?)
I didn't quite understand your 2nd example, is this just checking the hash on data? Or you just want online-only with
.bye() extension? I think you submitted a PR... will be looking into it. :)@canadaduane I couldn't find any example of GQL :P did you see any?
@go1dfish YAAAY it worked :D :D yeah
.grant( .trust( specifically.@matthewhall.ca_gitlab WELCOME TO THE COMMUNITY!!!!!!!!!!!!!!!!!!!!!!! Multiple relay servers is probably smartest, depending on your needs. The easiest route would be just to use different data graphs,
.get('foo') versus .get('bar') and save everything under those.Never heard of yuno.
reselect module. I think this would help with certain situations for sure. Reducers in buckets can only work on the data they have at hand, so by creating a merged view from multiple buckets, it would allow us to compute states we otherwise couldn't. But it doesn't solve the situation where we want a state change in one bucket to trigger a state change in another bucket. I'm leaning towards loop detection for that. If it's good enough for React, it should be good enough for Weir. I'm just worried a little that someone might deploy a change to prod that triggers a loop which they weren't able to catch in QA, and then they end up with jacked up user data that's hard or impossible to fix.
We build software for "Location based VR" - basically anytime you're using VR not for personal use. VR arcades, businsses using VR for training, schools with VR labs, etc.
Right now we use a very traditional client to server approach where we have some Windows desktop software (.NET Core), VR apps (Unity + VueJS - its a bit unique) and some web applications (VueJS) all using a GraphQL API (PHP, Go, Node microservices)
Looking to move that section of our software to a more decentralized approach for a few reasons - Biggest being speed and reliability. Want to be able to better handle clients with crappy internet, having communication on a LAN between our web software and the desktop software without having to go to our servers, etc.
GUN so far looks like it checks all of our boxes pretty well. Biggest things we're trying to figure out still are
1) How to make our current APIs read/write to GUN
2) How to include it on our desktop software effectively
3) How to host our GUN databases server side for multiple tenants
Hoping to ship a very rough MVP just to test behind the scenes in conjunction with the current approach in the next few months.
Not hiring anything specific to this project just yet but we are hiring a .Net/C# desktop developer :)
