Ethereum Cat Herders Community, 20 Jun 2022

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: URL/URI standard working group meeting #3

Description

Working document - https://hackmd.io/@poojaranjan/URLWorkingGroup

A

Welcome to url uri meeting number three. Thank you all for joining. um This is the third meeting in the direction of getting url url standards. Moving towards final, a quick recap for listeners.

A

In the last meeting we discussed proposal, 2400 url format for transaction receipts and proposal, 4804 that is 3 url to evm, call message, translation and, as next steps, authors um they're advised to make some changes like for proposal 2400.

A

The author mentioned that he would be revisiting the proposal and make a full request to bring the proposal to review.

A

He left some comment on the discord channel and I have added all of them in the working hack, md file. I'm going to share the link here in the chat for a quick reference.

A

And there was another proposal that is 48.04. uh The author was recommended to incorporate suggestions shared in the meeting and updating this meeting.

A

There was uh the third proposal- five one, three nine I mean that is not a proposal. Actually simon wilson mentioned that he would be working on creating an rpc endpoint standard, so he proposed an eip which is eip. 5139 remote procedure call providers list.

A

This is a new proposal and probably was discussed in the last meeting. So um I think we can go ahead and start with the proposal 4804 by the way, welcome greg.

B

B

C

Well, uh so let me share my screen, uh so I prepare a few slides um regarding the 4804. I think last time I think I introduced uh some uh actually erc, for example, and some uh questions and mckay uh gave some suggestions, so I feel, like a lot of people, may not have food context of what it's going to do so today I just prepared a few slides to tell what's going to do.

C

What are the some comments we had last time and how we fix fix them, and what are the remaining topics that we want to discuss and then also open for suggestions, comments on how we are able to uh adjust those uh concerns. So what is uh eip4804? Actually? It is uh http like it aims to be http like uh so that we are able to locate uh map3 here.

C

The web 3 is more precisely a evm compatible, blockchain, um so resources, so it's very similar to https, because we want to have give the user a similar style that they can easily using their existing understanding of http and to apply to the new url, but unlike http, which locate dns and also the corresponding web server resources and so for web3.

C

uh It will replace the dns with the ems, so here is an example of using dot ens and also is able to replace the port number with the chain id um so that it just behave just very similar to http. But you have different semantics of the the rest arguments in the path.

C

For example, the first one will be treated, so the arguments will be separated by slash backslash and then the first one will be treated as master id, for example, balance off and and then you can follow multiple arguments right now here. This one is real using uh ns as an address, so basically it's looking for uh zack's balance of a contract, and then it also will tell what the expected return types that will display on the browser.

C

um So if we already have a gateway extension that uh support this, so you can take a look at uh right now. I have. I have a firefox browser uh where I can uh put the ems homepage, dot eph, which is the home page of the ens.

C

uh Basically, when we expect respect the request, it will make a email rig, make us a list of requests which basically will call the corresponding contract that we saw from ens and then we'll call provide the corresponding code data that is passed from the corresponding method and also arguments. And then uh the control will tell us what the content will look like and then it will display on the basically the browser supported by. uh We have a extension here.

A

um Sorry to interrupt here, uh I I believe there was a hand, read grega. If you have any question.

D

Yeah, um what's the objective like I can understand the website logic like going to resolving a website, but what's the logic behind having a path name so like for methods and arguments like why append quarrying to this.

C

I mean how what what's the why we need a master id argument. Well,.

D

Actually, let me let me rephrase that sorry, I feel that was really confusing. Sorry um from our last, the last call, if I'm not mistaken, like one of the big things that I think myself and mika both were a bit confused, although I don't think I spoke much is um like how does this this much different than just like looking at the ens contract and resolving whatever the address uri is set within like that parameter? There.

E

Not not to forget to mention that these would need to be url, encoded or uri encoded, um which is not described so the period wouldn't even be allowed. In your current example, uh there was an issue with return types discussed last time, an issue with the usage of the web three scheme and a request to change the scheme.

B

Yeah correctly and.

E

The chain id here is done as port number uh following what http would do, however, every other url eip, so 681, 831, uh 2400, 1594, etc all use an at symbol, and then the chain id um in this case you're feeding user info through the url, which is a little weird, because the user info would probably come from the injected provider of the browser.

C

We do not rely on any like, for example, metamax or meta any, uh for example, uh wallet that we are still able to actually uh locate the webster resources um so how, for example, this user info. It doesn't need to pass to the wallet it can just go to the browser and brought up no, the gateway or the extension, so that so that it will basically translate this user into right.

E

But then we need a gateway or extension.

C

Yes, yes, that's right.

E

And that would have to point at some rpc id, which is then predefined by the gateway.

E

Right so you're.

A

E

With centralized servers, making the decisions for you and no way to upgrade to a situation where you could pick your own rpc endpoint, um so it.

C

Depends on the extensions, how do some extension implement, for example, forget for brave, like, for example, it has native integration ipfs right, um so then it can, for example, specify which gateway, uh which example, uh json offices, server, or maybe it can itself can be, uh become a like line so that you can just directly query uh the all the information from the of the companies and blockchain or any other blockchain, and so that it doesn't require uh fixed, strong dependency on the what kind of generalization rpc it can.

C

It can be used right so so this is just a purely way that, without any dependency, what kind of resources and what kind of the information the from I want to call the basic blockchain.

F

So a question about that that I have is um the tag, equals networking and order equals newest there. um Does that imply that that data gets sent to the contract, or is that some kind of like um browser side filtering that would have to go on there? Because if it's sent to the contract, then that means there's. um There would have to be some kind of like extra indexing that you wouldn't want to put in the contract.

F

But if it's on the browser side, then to make sense of the results that they come that get returned back, there has to be some special casing there. So I'm not really sure how that's supposed to be interpreted.

C

um I'm sorry the volume is pretty small like. I cannot hardly capture.

F

I don't maybe that that's part of the spec that I didn't see, but um that makes sense they would make more sense for that to be part of uh the fragment data to me.

F

If it's, if it's on the browser side.

C

Yeah from the browser side, I mean the the fragment data yeah that that's the naturally so so here I did not put the fragment part, but fragment, probably will also handle by the browser side, which is the same as http.

C

I'm not sure I explained the um this correctly yeah.

F

Well so like the return types there um shouldn't that be part of the fragment instead, because that's for the browser to interpret the data it gets back instead of being sent to the contract. Oh.

C

Yeah yeah yeah yeah. I got it. Okay, yes, uh for the return type. um The thing is that if we put it in the fragment and then if I open a browser to basically uh basically, we are discussing whether we should have a server-side, rendering or client-side right now, right, using fragment is kind of a client-side rendering, so it tells oh.

C

This is a new kind of an integer and if we put it in the question mark, that would be clock server, side, rendering which was suicide wheel and it can be an extension of native integration or gateway that can say hey.

C

I want to return an integer uh to the to back to the client, so here I my choice is to use uh basically the server side, rendering because like if I use fragment, then that means that I need to return a long html uh with uh the messenger, the the fragment master, to tell how I'm going to pass this to the front end, but with the server side that we just return an integer like in json format, and that will be much simpler to handle.

F

Okay, that makes sense to me.

C

C

um So then I can move forward actually. Regarding this query, query part: I also have a topic before I want to move this uh eip to the reveal stage. uh So for the last um meeting. um So basically, we are discussing a type, auto detection feature of this url link. uh It is a very, very important feature. The purpose is to simplify ul without introducing confusion.

C

uh So for each argument. Here uh we will basically uh we have a type detection that automatic detect, detect the type uh if it falls in some subspace of the agreements, for example, it is pure number, then we always treat it as you in 256 or if it's some fixed number of hex, then we will treat it as bytes or address all other types. Then we will treat it as dynamic size, bytes and then also uh we change. The fallback type detection right now is for any uh other types.

C

Then we will treat it as name so, for example, p for here. If here is this example, then we will treat it as a name so that we will try to resolve it while dot eth name resolver, so that and then we will translate to replace it with an address.

C

um So the update is the last time we prefer the preferred using the name over the stream for as a fallback type detection. So, for example, suppose I want to call a contract which will basically show an image with a string.

C

Then it will not work because originally we use the string as a full background right now with this change, then the user has to specify uh in include the type, but here is a string and with the extremely exclamation mark, uh which tells that hey, this is the following type is uh must be string, then it will call show image and parenthesis parameters, string method and with one dot jpeg that is encoded in utf-8 and then and that you, but for to, for example, for other examples like I can suppose.

C

I want to query the zucks.eth balance and it will just work by auto type detection and it will treat it as a name and it will also treat example, logic gpac, as of name but most likely. This jpg is not supported by either.

E

Gateway am I understanding correctly if I'm saying that, like you've added a bunch of ways that it should kind of try and auto detect whether it's a file on an image, an ens name uh or an id name or etc, etc, etc, and then 30 more things and then you're, adding a bunch of extra workarounds to specify that it is a string. So you can do string explanation mark. So you can say that it is a string and then like.

E

If I want to enter the string luke dot computer, then that would not resolve as an e s domain. That would resolve as a string. Even though that is an ens domain as well.

C

Yes, that's correct.

E

So there's there's a lot, a lot of ambiguity and room for error in your url system. Then uh why correct uh I mean follow this rule, because if somebody just defaults the input in there then without specifically specifying every single type and making a url's very long, um then it will probably cause some error at some point by something being interpreted as a un-256, a byte 32, an address, a byte or an e s address, or something different right.

C

So so, as long as there's no like explicit type, then it will apply this auto detection rule.

C

Right so so, for example, for with this explicit type, so there's a exclamation mark which basically will will not be in in any of this uh basically uh rule in this uh in this four in the abcde. So as long as I detect a exclamation mark, then I will treat it as explicit uh type and then, if, for example, this this can be also wrong right, it can be abc, then it will just return. Oh, this is uh the user request is wrong.

C

So there will be in this express type and also auto type detection. They will order their this subspace of their arguments will be, should be completely unoverlapped.

A

I wonder if there is a um fem link yeah, I think it should be there in the proposal right.

C

uh I mean for the for this discussion. That's.

A

C

Yeah uh so there's a eip and actually I have an extensive discussion with. uh Actually there are two eips, so I should probably put another eip here, but not another pr here. Actually.

C

Sure I should search for close.

C

um So there's a one, and then I have a lengthy discussion with mckay uh regarding what uh what what's the default behavior of name, auto detection, and why we would like to use the uh name over string because, like uh for example, is uh making sense like before it feels like people wants to use the name service like address more than a string of names, um so that so that we we decided to use the name service as the first class and then string just need to be explicit string type in the argument.

C

Yeah, there's also a discussion regarding other types of just just some wording of the of the object of the of the rules. Yeah.

A

That's right, that's nice and I think maybe you can go ahead with your presentation and uh we will share it with community. So if people have more questions, they can reach out to you on the discussion to link and add their thoughts.

C

Yes, um so yeah, I also removed the specific name, server sorting, and so uh following the mckay's comments, so today I want to uh discuss a topic. Actually it's also closely to the question just raise is where to put the specified return type here, at least a couple of the possible way to put the return type. It can be just following the the name, the contract name or the, unless name with the return time.

C

This is, I feel like a bit awkward, because it's quite different with http style and another is I put this one address after solidity, uh just after the master ready it's more close to solidity style because, like I have a method and then return something, but probably the http users may still have no idea what what what is and I'm more close to http style, which I gave example it's just using the query part and then and then let the server will decide uh what kind of return types is going to be returned, but because all this can be basically random by server, so it just.

C

We just need to implement the corresponding server or native integration implementation. We just want to make sure that, what's the best type that follow that meet our final goal and so yeah, I'm open for uh any other suggestions.

D

um Something that I just thought of, and I'm actually really confused about uh for, let's use option c, the one that you highlight because it's easiest to read um is foo dot eth like a token address, or is that a user address.

C

It's a token address, oh more, like something you can see: uh usdt dot equals.

B

Okay: okay, okay,.

D

Oh, I see and then bar.ether would be the user or the address that you're looking at rather yes, I see um so this is where I get a little confused. Like is the objective in this case like if you're looking up foo dot e here, then that's just going to return. You like a number in the browser kind of like a json blob. The way like you can return like api data.

C

Is that the idea? uh Yes, it's just uh it's kind of like uh using this way. I can query any uh pick. Some exam contract I can replace with address and then be able to return any json file without using the example mask or any tunes just like yeah.

B

Yeah yeah, okay,.

B

Okay, interesting.

D

And then I guess it's up to your extension to determine what your gateway is going to be.

C

I can gateway, and hopefully I hope that the extension or maybe native integration can make it as a kind of like line of I know, like portal network, is also doing some very cool things. I'm not sure he's able to integrate that so that, in this case, when you used to type this, it doesn't need any example gateway or inferior. It's just teddy.

D

Yeah yeah, you can just use like load stars like klein or something like that. Just get.

G

D

G

Yeah yeah yeah.

F

I think it might be interesting to take examples in the wild of like doing uni swap swap or interacting with various dapps um and seeing like what it would look like the various kinds of calls that the dapps make to the contracts and seeing what it would look like using this format in, like whatever level of disambiguation.

F

You need with return types or argument. Types whatnot, um because, like something like unit swap, has a lot of parameters. So that could become unwieldy. And I don't know if that's problematic.

D

I think the problem is the post data manipulation. That's really the issue here.

B

You have to do so much to do so much parsing after the fact just to get things like standardized.

D

Yeah that'd be really annoying the only. I say this because, like we just wrote, 3.js in a way that you can like standardize at like global level that you want, like all units to return to like big numbs or whatever so and so forth.

D

It's just like doing that with like uniswap, for instance or even die, is just like a pain in the ass.

C

Heard in my language, so you mean to kind of standardize uh what the region types it is.

D

Well, it's just like kind of like what trey was saying it like I'd, be curious to see what like a complex like call, not even a complex call, but like a more common call than the space would look like, because a lot of the time like functions aren't even returning you anything. You actually have to look at like the structs internally. You know and like map things together to get.

C

Like a reasonable.

D

C

um So I think couple of things that I haven't covered did not cover too much details in this eip guide, fixed array uh struct, um all this things that right now we just say, hey just follow the api uh interface, but definitely we are able to make it more clear. uh What I expected to be returned. um Definitely, I think, is a very great suggestion. uh We are able to fix some ems or, for example, unit swap right now.

C

I think we have a uni swap which uh that is on link b, uh so so resource, but right now it is still. uh Basically the the the web page itself is still calling uh money mask to uh query the corresponding price on the information from blockchain, and definitely we are able to uh replace that and so that we are able to, uh for example, for any user. Even they don't have any wallets.

C

No money masks they're, still able to browse what the blockchain looks like and if they want to somehow really want to exchange some token, then starting you have to install wallet, and these are these all the candidates that you can use.

D

Yeah, but I think we're solving like two different problems here. Right, like you just described the provider injection problem, which is what metamask removed originally back in the day like a year ago um and you're kind of talking about.

E

I need to collect my thoughts. What.

D

E

To do, if I understand correctly, is wipe out the entirety of the injected provider and make it work purely through links, which would mean he could use like a text based browser or something else. But.

A

E

From my perspective, there's a lot of overhead a lot of ambiguity, a lot of non-compatibility with the current standards so like web 38. Dapps would need to adopt this and like lose out on a bunch of functionality, lose out on a bunch of responsiveness and like it's going downhill.

E

um So I don't think this is at the moment, in its current state, um a feasible uh mercy.

C

um So for one example that I'm trying to use is, for example, uh create a, for example, charlie mentioned, creating a blog on the blockchain uh in his recent non-financial article of usage of blockchain, and with this, then we are able to fix and share my instant tweets share my articles or share my blog uh with dell, basically, which all asked for on the blog blockchain, but without asking a user saying you have to install is the metamask so that you can watch all my, uh for example, articles uh just type.

C

For example vitalik uh you just everybody can type uh what the oh it's broken.

B

Oh, it's psycho dos the other earlier today because he put up like a new blog post.

C

B

His site went down for a bit today because of a blog post. Oh.

F

Yeah, it's all sit down for me.

B

A

Yeah, the 40 liter backpack.

B

Yeah, it literally caused him to go down. It's ridiculous.

C

Oh, no, I can type in puppy yeah. So um so when he shared the uh blah blah, then we don't need to ask the user to say: hey, you have to download a wallet um some some kind of, like maybe re, read only applications that we are able to easily to upload more back to users.

C

That's one target that we're thinking about.

A

Question here, so um I'm just trying to follow up with the greg's earlier comment that here the proposal is trying to come up with the solution to two problems. uh Is there a way like we can maybe uh separate it into a smaller eips like.

D

Smaller well like well like yes, let me explain that. I think I have a better understanding of what I was trying to say so, there's kind of like two problem statements that have been made and they're kind of. Not, I don't see them as related, but I see them as potentially one is codependent on the other. One being you know, as a user I'll use user stories as a user.

D

If I don't have a wallet, I want to be able to query data efficiently, that's kind of like one problem statement right and then the other one is kind of like as a developer. I want to provide my users with like a very seamless way to quarry data right and.

E

D

Are kind of like two separate problems that you seem to be trying to hit with the same solution if that makes sense, all right.

D

And it's like yeah, I want to solve the like client realistically, but, like you know, you don't really need a wall. The lifeline could live even in like the dap um but yeah, that's kind of where I'm at right. Now.

G

A

So I mean like: uh is it possible to have it like into smaller eips like solving one for user and one for the developers.

D

Well, I I think, like the bigger question is like what is like what is the core purpose of standardizing like an http like url pattern like what is like the biggest benefit you get right.

D

I think that's like the number one question to ask: is it for access to dapps or is it for quarrying data or is it for both? You know what I mean like, and I think really just narrowing it down to that. One example is probably the best here.

D

D

Yeah, it's like, if you append the question mark right, so the query to a url.

D

If you don't have it, does it resolve a website and if you do have it does it resolve data like I'm, trying to like figure out like that middle ground of like so.

C

So basically, right now the particle it is that if there's no return types so.

E

C

Can specify and we'll it will treat it as a website, but look. Obviously this is not a website, so it will. uh I think the browser still display it, because it's a json format, some browser will auto detect. But if there's an explicit, oh, let me see yeah.

C

Let me double check. Oh yeah yeah.

D

So I see where I see what you're saying but, like my question is like the ens protocol. Technically already does this right because you can just do like whatever find the uri within um keys at pronounced. Like I don't know, john doe dot eat. You know what I mean and like that can resolve the website.

D

So that's what I'm kind of trying to say is like if this is purely a quarry format. Why not just follow suit with the ens name flow.

C

D

C

D

I think the big problem here is like this might be a better eip for the cips, which is like the cross chain standards working group with, like pedro from wall connect and all those folk versus like an eip.

D

Oh you see, I'm trying.

C

Cip mm-hmm or what's this see, content.

D

uh It's the cips, I don't know like the type in like cip cross chain, uh like league league and like pedro, work on it yeah right there cips, that's the first github link.

C

D

C

D

Now these are the ceramic improvement proposals. Lol.

D

ah Let me let me find it because, like this is touching on like okay, this is like a personal opinion, but, like I think, a lot of um eip editors, like standby, too, is kind of like we want to try and make the eips as like, non-network agnostic as possible, not necessarily networking but, like.

D

I think it's for ethereum mainnet at the end of the day, not necessarily for like the wider evm ecosystem right, um although we have to keep that in mind for security reasons like I I worry this is like too generic. If that makes sense,.

C

I mean for two generations.

A

I think I think they are also available in the wallet discord key. If you are there on that channel.

C

uh What it is called.

A

That's right, um I may be able to send you the link later on, but if I'm getting greg's point correct, uh uh the interpretation is this proposal is not ethereum specific, but it is solving uh um like problems to more evm related chain, and the idea of a standardization related to eip is more ethereum. um Centric proposals is that there.

D

We go yeah, sorry they're, not cips, they're caips, apologies.

D

Like a lot of like basically a lot like I'm, so I do a lot of work in the bridging space um and like these are like standards for that stuff. Because it's like helpful to for like multi-chain focus- uh and I feel like this- is more of like a chain agnostic way for like doing url standards right and.

C

D

Get like wallets to implement them accordingly, in browsers versus like and.

C

Definitely apply to any evm compatible blockchain right.

D

Yeah or just any chain in general, really.

C

D

That's just my two cents. Personally, you know it's like because you can easily replace with dot mirror with dot soul with dot whatever right. So it's like, why not just make this like a.

C

Caip but they will have different master id argument encoding so that that's something that I want to touch at the moment right. So these.

D

Are like those, but those should always be different, no matter what no.

D

Those are arbitrary quarry strings are arbitrary. It doesn't matter how you encode them.

C

Yeah you just wait, but right now here is more specifically how we call the evm message.

E

I think I agree with you greg that this is probably more of a cross-chain situation.

D

Yeah yeah, it's just like, I think, like standardizing, I don't know I'm just having a tough time like with the with the the use of like why I would.

E

Let alone the claiming of the web 3 schema.

D

I care less about the web. 3 schema, that's like irrelevant to me. I can claim I can claim http plus https for fun. It's just about whether people actually like adopt it.

E

Yeah, of course,.

D

D

Like I'm literally just like trying to understand how this would be implemented and how it would be consumed right, like you need browser adoption for it to be consumed or wallets need to natively integrate it. I.

F

Also think it's just it's very, very generic and like how you would interpret what data you would get back from. It really depends on the use case, so it's difficult for someone to just give you a url and you just be able to properly display it, and you can.

E

F

Like hints on the side to be able to figure that out, but that puts a lot of effort on to wallet developers and browser developers.

D

Well, it's kind of like the graph right. It's I don't know if anyone uses it but like half the time the data you get back is absolute gibberish and you need to end up parsing it for like an hour until you can parse it into something like consumable. You know what I mean it's like.

D

I think the intention is right, but the question for me is like: do we actually need this as a form of quarrying data, or do you just need a playground to quarry david? You know what I mean it's like in web 2, you don't really. You never really see like a.

D

I don't know yeah like.

F

D

Like this is for like a website to parse and like decompil decompile, but it's like a known schema for the website when you start doing generic schemas, it's kind of weird you just get like a db dump against an api.

F

And, like you, don't you don't have like a content, encoding header when you're making like a call to.

E

A contract that you.

F

Would have in http request, so that means that the urls get really ugly and not human readable.

D

Yeah yeah, it's just like contracts, aren't written nicely. You know, as I was saying before, it's like the the usefulness of the data you get back from a contract is usually against the next chained call.

D

If that makes sense to anybody, it's like, usually you have to chain like six calls together to get like useful data.

C

um That something that would actually we would tr for sure is like kind of like uh you can use the basically evm contract, as kind of like, uh for example, as a comeback uh backhand that is able to store, for example, contract uh webpage, json, javascript or even, for example, uh blogs or even tweets, uh as as, for example, uh charlie also mentioned here. uh So if we are able to do it on chain, then we can solve a lot of problems, especially like ipfs static with vm ebn.

C

We can compose all the users comments, articles put it together and be able to dynamically create them. But the front point is that we, the ethereum lack of a presentation layer that we are able to directly access the content without using the wallet so that uh when I share a picture of each other's post, then sorry you cannot read the post. You have to uh install minimize before you can continue, um there's something that I've been put application that I I don't think you're thinking about.

E

Can we take a step back? Did you say storing your blog on the blockchain? Yes, in, like http format on ethereum mainnet.

C

That can be easily mended and that's something. Some project has already done fixed most cycle brokers. They just upload all the svg um to the ethereum net, but definitely we can find cheaper solutions with l2 or something else to support that.

E

Like are we even ipfs so like? Basically, what we already have? I'm sorry isn't all of this, like already implemented by most like browser-injected, wallets, aka, metamask, brave wallet, et cetera, uh so you mean the brand.

C

You can literally.

E

Just enter any ethereum, url and it'll already do schema resolution and load the whole website of it, so your entire gateway would not be necessary. In that sense, I mean um calling customer functions from contracts through get requests is kind of an odd one, because it really only restricts you to being able to execute certain read-only functions because remember this is not connected to any provider, which means we can't like execute stuff sign stuff. Whatever.

C

Right not not necessarily because, like you can also so right now, the standard is more like the just located resources just to read it like kind of like get http cad, but definitely we are also able to uh post, but suppose, like the client, the browser has some way to just sign signature. Even without my mouse, then you can compose this transaction and be able to directly broadcast to p2p and that's still possible, but but I think we just want to step.

C

Do things step by step, how we are able to solve pure read-only operations, how we are able to, for example, set up a broadcast.

C

C

We just create a contract because they have no idea to um to render it on her blockchain. So it's still even they said they have all the svg on the blockchain, but they're not able to rename it so just run any type of token id. Then all the content here right now is composed by svg uh kind of like html, but all purely on property filament and unfortunately they don't have the way to fix someone using this web3 standard to specify in the token url. So they still have to use the.

C

uh Let me see so they have the token url just and right now you still have to rely on rtfs as a copy, even though they said all the metadata has been uploaded on blockchain because like if they said no people will never worry about by losing the data, because all things is on top of easier. So I feel like this is something that now right now is missing. Then then we, if we are, we can fix them and replace this fpfs with directory and then point to the theorem method that gives them this contract.

C

I just create a counter and then repeatedly call the random random master and then is able to retrieve it directly on top of ethereum without any using some other third party service, ffs or maybe even wallet.

C

Yeah, so this is pure on chain render, I just change the 20, and then you just, I think it. It internally has kind of iterative generating method that I'll put the svg panel, and I just put them together in the contract, and now I'm able to display it just using the standard and so random blocker is the master, and this is just you in 256.

C

Questions of this kind, like application in terms of having more focus on web, rendering using rendering using uh big kvn as the um as the backend and then and then we use the data and then so. The evm can compose this data and then be able to dynamically random directory uh without relying on any uh volume installation.

A

All right, um so, if I understand correct, would you be making a new pr with updates that you have done so far, uh I'm wondering because they were looking into uh adding one more pull request.

C

uh Any animals.

A

uh Adding a request to update this proposal.

C

uh Update of 4804.

A

Yeah, my question to you is like, as a next step, um are you proposing to maybe make uh changes, discussed and update the proposal.

C

um uh Change change calling out today's discussion.

A

C

um So um but what was the conclusion of today's discussion.

A

Yeah, that's a good question so greg I don't know if you would like to maybe summarize like, should we go ahead and continue this as an eip, or should he be reaching out to caip group and maybe moving in the direction of making it a caip? What do you think.

D

I honestly don't know.

D

Maybe perhaps like the best course of action is to determine to see, if, like there is more demand for this um reaching out to caip, so like pedro from all connected league will definitely help there.

D

I think taking it down that road is probably the best first course of action.

D

Do you have like a lot of support for this um in the wider community or because maybe, if there's other people that can shed light on like how useful this is for them then like it would be really helpful for me to understand, but I'm just having like a really really hard time.

D

um Seeing like the the use case here, um that's not like a burdensome for developers.

C

And users um yeah um so so right now we so this is uh this- is this standard is just composed in february, so it's kind of in early stage, and I mean we are also looking for searching for more applications.

C

One one thing is like cyber brokers, because we notice this the claim to be the first project to put all the uh their metadata on the blockchain and about how they achieve that, and they found them using the presentation part so that even they have all the data they are not able to use them on top of eastern, they have still using a copy of ipfs.

C

um So I think this is something when there are more and more kind of like the applications. I think somebody will be using svg to render the data or, for example, even ens. They are proposing the pfp for their earnes names uh and then, if they are able, once you say, they have algorithm they're according to your name, they're able to randomly generate a profit, uh your the profits uh picture.

C

The point is that where, where to retrieve that, there's no way to be able to retrieve directory on block on property theorem, even the corresponding rendering logic, the generation logic is already implementing by external smart contract as supported by ens. And this is something that, having said hey now, you can use standard, and then you can display the basically the order profile picture of that that you randomly generated just directly on top of each other blockchain. By calling the contract that you created.

C

Another example that I'm finding just a few days regarding the uh he complained about like maintaining his blog on ipfs all the issues. And then then, if you dump on chang, then um we are able to do much better and but if we don't uncheck how we are able to render it and how we are able to uh another question how we are able to store it right.

C

That's another big problem, but at least there are a lot of, I feel like the more and more and thoughts regarding how we are able to extend uh the non-financial applications of evm uh definitely can be part of the financial application, but sv svg generation compose composing, different svg or articles of the web generation. This is something that we feel like that they're more and more thing that evm can can do great for uh especially enable this kind of what we call dynamic, dynamic, decentralized website.

A

Earlier I remember you shared that there is a ud swap implementation on some of the test nets. uh Just wondering is it being involved in one of the devs from the swap or unislab team, or are you in touch with that, or is it by the proposal development team.

C

uh So this one uh it is the the web content is stored on the rinkeby uh in this web. But actually the website can work out for any network, for example instagram or any any network. um So because this website is just purely a copy of ethereum front-end and it still connects to the wallet.

C

But the website itself is is retrieved from this standard and returned to the browser and it's display just as uniswap.org.

C

Is but the good thing is that using this one, but definitely uh it can, for example, using ens pure ens, plus the plus the website generated. Then you can manage the website on top of blockchain using dao something else. Then it will be much better for sensors resistant.

A

Right now, my question was like: uh are you working in association with any image laptop? That means just trying to get the information on like? Is there any interest of uni swap for this proposal to support.

C

Yeah, uh actually, we don't have direct contact of unison. We are interested if anybody can introduce and tell them how you are able to do something differently, uh rather than using ipfs yeah.

A

All right, that's not a problem um as uh greg was suggesting that it is important to maybe reach out to projects to gather uh support for this eip, and you already shared some of the use cases that you think would be useful for so maybe as a next step for the next meeting. If we can get some more, you know feedbacks from some of the projects that may be implementing this proposal. Inclusion may help increase the chances of getting it moved towards final, a little bit more um yeah.

D

Yeah that that's sort of what I'm suggesting is like, maybe um before we discuss this any further, if we can have like some support from other teams, that would be leveraging this that I think that'd be a really good start yeah, um because it'll be up to them to help like shape it. um I think at this point, like we've kind of set our opinions, and it's kind of you know up to the community to adopt it um for us to like push it forward.

C

Yeah um so one direction, I'm talking about this uh called the eth website subgroup. I I already joined this discord. There are a bunch of people they're, also discussing uh other way, besides ipfs hosting, because they develop a lot of issues and they're looking for some new solutions, and I think we will have a uh eth subgroup meeting in a few weeks, one or two weeks, and then they asked me to give the presentation and demo on how to use this standard and how?

C

uh Because right now there are also a couple of uh ens, I would say, gateway like example, uh the pth link of nemo. So I'm trying to convince them to also adopt these standards to integrate into the gateway. So there's not only our gateway, but multiple gateway can also support this standard. So there's a couple of steps that we are making forward to a wider adoption and and here how people are using them and how what kind of feedbacks do's or bads, and then we are able to incorporate the eips.

C

D

Yeah, I think that's the best approach because, like I feel like one prom that we have plagued the ips repo with is like a lot of eips that, just like don't do anything and stay very stagnant yeah and I think it'd be really nice. If we could get this included once we have- um and I forget the status code but like once, we actually have like some actual adoption and usage of it um and have some people experimenting. It that'd be great.

C

Yeah yeah yeah, that's something I think I think I'm really uh looking forward to it and, at the same time, definitely I would like to see existing projects rv compound and if, if you have connections and tell them hey, there's a different way to host your website in more decentralized way. Now you can using your dial token, to manage how it looks like um then. um I think that that would be also a very good uh exploring direction.

A

All right that sounds like a a good point uh to maybe break off for today, and we have some next steps for the future meeting.

A

um I know there were other proposals as well proposal 2400 to be discussed today, but I don't see the author on the call and, moreover, we are on the uh like an hour mark, so um people can maybe refer to the comment that the author left on the discord. I have added it to the work in progress document for url uri.

A

One question for this group: when do we want to have our next meeting the next meeting? If we keep it on uh like alternate week, it's going to be on july 4th, but I know july 4th is a public holiday in in the us. So do we want to move it to july 11th or.

A

A

So yeah, I'm not sure, like many people from us will be able to join on july 4th. So is it okay? We can keep it for july. 11.

F

I'm fine with that. I know it's me yeah as.

D

Long as we have an agenda.

F

A

Course, yeah. uh What I'm gonna do like going forward is like create a tentative agenda and uh share it, along with the meeting invite and on the public channel of ethereum catholic, so people to follow- and I hope that will be useful for like uh having it on a regular basis agenda and the summary and conclusion that we are discussing in this.

A

Well, thank you, everyone for joining us and I hope to see you in the next meeting. Thank you. I will share the recording on the ethereum cat, headers youtube and also on the squad channel. Okay,.

G

B

So much me too.

E