[LIP-5] Opening Up the Garden #29
Conversation
Requested Lip-5 proposal to progressively open Lens Protocol.
|
Not sure as to how to comment on Github or if it is possible, but I would like to thank @ZKJew for taking the initiative and writing such an important and timely proposal. I have been on Lens since August of 2022. It was a very odd experience pioneering into Lens and Web3 Social Media but it has improved drastically every day. I remember using hey xyz in the browser of Zerion & 1inch Wallet. Now there is Buttrfly, Orb and Phaver, each within the comfort of the largest App Stores. The maturation of the Lens Protocol and its client apps is very real and there is no reason for it not to be permissionless. We now have Momoka protecting & scaling us from Layer 3. From here on out, Lens only grows by scaling at the technical level and inviting more users on and figuring out the limitations of what has been built. I think soon is too late. I think now is the time. |
ZKJew
changed the title
|
Some things that I think could be discussed are whether the socially based invites should have qualifications to them as Lens comes out of beta. An example of an introductory cohort would be 50,000 WL that have to pay for a handle, then also 50,000 free profile and handles to Farcaster users. On top of that, have 5k-10k profiles reserved for users that have met certain qualifications like posted 30 days on Farcaster, Deso, ext. Another thing to discuss could be the evaluation to delay or accelerate the pace of users added based on how polygon and Momoka handle the increased usage; however, I think this is best to be left to the team and that about 105k per month is a fairly aggressive initial growth and stress test for the protocol. Lastly, there may be some gripe about why I suggested for invites to have free handles, while WL do not. I think for one it will allow those who want to get off the WL most get a profile (spammers will get weeded out or left with just profiles), while invitees are people the protocol wants to bootstrap before decentralization, so reducing barriers seems necessary to get the best community members. |
|
I have been using and creating on Lens almost since its inception and can say that there is nothing like it anywhere on the internet (not even the recent hype of Farcaster). It's a technical marvel that is enabled by social graphs, one of its kind. Use -> Break -> Iterate; the sooner the better. |
|
Thanks so much for opening this LIP; I fully support opening up the protocol to everyone. As you said, it’s time for Lens to open the protocol to the community. Everyone is asking for it, and I think we should do it. Note we must think about the spam which can happen when we open up, so it is not as easy as just saying, "ok, we do it". We must first put a few things in place to ensure the experience is still good for most applications. I greatly favour this LIP, and I think we will move this forward. Thanks for raising it with us. We have heard this demand for a while now, and we need to give the community what they want. Excited to see what people think of this all ears |
|
It makes sense to open Lens more. Though, the needs of the protocol and of the apps have to be considered. Too many new users will cause problems, I guess. The "quality" of new users is important. I don't think Farcaster users should be addressed explicitly, though. Most think FC is the top deal and explicitly addressing those users will result in bad takes about vampire attacks, etc. Some ideas for onboarding creators were good, e.g. Refraction. The problem was that those creators couldn't post to their followers on Twitter "hey, come over and join me on Lens". |
|
Thanks @ZKJew. Lens being gated is the most frustrating thing not only for me but also for most of the app developers I've talked to. Opening Lens to be permissionless is table stakes for building a successful network, at least at this stage after being around for ~ 2 years and removes the main barrier for devs and apps to build their userbases, which is the foundation of building momentum in other areas. I'm really happy to see this discussion and have folks like @joshstevens19 chiming in with his support and excitement to move this forward! |
|
Also in favor of this personally 👍 |
|
I'm very excited that this is being brought to the community and fully support the initiative to bring more users onto the Lens social graph! My only concern with the proposal as written is that I feel that it doesn't go far enough. There is no such thing as "semi-permissionless". The protocol is either permissionless for any address to create a profile and access the network or it's not. There are two primary arguments against a permissionless model for the Lens social graph:
I believe the same mechanism can be used to deal with these concerns as the underlying Polygon POS network, gas cost. The Polygon POS network is permissionless and it operates continuously, producing blocks with a finite gas limit on a constant basis. The Polygon network can regulate the rate of on-boarding / spam, and Momoka scaling infrastructure is always available as an alternative method for interactions which don't require on-chain transactions. I propose to make the Lens Protocol fully permissionless with the caveat that any address that joins the network through permissionless onboarding should pay the gas for profile creation, and is only eligible for sponsored transactions using Momoka infrastructure. These addresses could eventually be given access to sponsored on-chain transactions through the Lens API when they reach a certain threshold of graph connection and engagement. Since this is a substantially different scope than the original design, I would like to gauge whether the community agrees with this approach and if so, create an alternative LIP to enable fully permissionless innovation. |
|
I think I have basically one (and a half) concern(s) regarding lens going permissionless and that's I guess some barrier of entry is required to avoid spam at scale. Rn the barrier of entry is getting an address allowlisted or 13 USD (current price of a lens handle on opensea). Both doesn't sound crazy to me but it looks like that it's a (the ?) reason for devs not to build on lens. Which is a major issue if true, imo. Imo lens becoming permissionless should come with an "onboarding program" helping new users to get an overview of the ecosystem and first steps to have fun and find frens on lens. the upside :-) I hope lens becoming permissionless helps devs to find and onboard niches easier, they can build unique UX for. Solving irl problems and generating irl value. |
I fully agree with the take that Lens needs to be opened in a way where apps, front-ends, and developers can handle all the increased demand. While, I personally do disagree on the method of acquiring the users. I think we saw with the regular invites that said "Hey, come join me on lens," are too tedious and lackluster of an approach (I don't think there is quarriable data on this though), so its just the sample I see. That being said, I feel it is the best idea to invite Farcaster users first. Here's the rational, it tests the idea that Lens will be able to fulfill its purpose as a portable social graph. Meaning, Farcaster users will be able to link and port their social graph into Lens as their followings and followers join as well, so they can have access to the Lens ecosystem without leaving their community behind. Additionally, Farcaster users will be able to give excellent feedback to the community and team due to their similar background. To address the social implications, I do not consider it to be a vampire attack because in the end I do not think Farcaster users should have to "switch sides" and only post on Lens. I think Farcaster users should be able to post directly to Farcaster and Lens at the same time, while maintaining their social graph. |
I really agree, I think that something needs to coincide to reduce churn. I think I have an idea it's a little silly, but I will get my ideas down and post it to see if we can get feedback. I completely agree that there is a chicken before the egg problem on Lens hands. Developers can't make scalable apps without users and users will leave if there aren't quality apps. So, there needs to be a way to disincentive churn as developers build products for users. I do think an onboarding program would be a really good idea. I think however, if gamified we could get more out of it. I think I'll make a separate LIP for this call it LIP [5.1] "Lens Quests" where users can complete quests curated by the community, developers (which should help with developing apps), and the core team (to stress test momoka, ext). During these quests, users would gain "experience points" which level up their Lenny unlocking attributes or customizations. |
While I understand the perspective, I do disagree. Mainly because I disagree with the categorization of the current plan. It's not "semi-permissionless" imo it's a path to become permissionless. I fully agree that the ability to sponsor transactions can be done well with retroactive inclusion in a permissionless way, but I think you're underestimating the large incentive for spamming Lens. According to Stani of the total 134k profiles about 40k were likely sybils on day one. While momoka and polygon might be able to withstand a large amount of stress, the apps in the ecosystem might not be able to handle an extremely large increase in demand. Ergo, until there is no longer an incentive for spam on the network and developers are just bootstrapping their apps, Lens should have a careful phased approach. Now, maybe a more permissionless onboarding WL could be employed like "link a year old twitter account", or another broad stroke mechanism of inviting a broader audience would be a good middle ground leading up to the point of permissionless. But, I would say due to the large likelihood of spam and the need for developers to have quality user activity and feedback the progressively phased approach is best. (Would love to hear others opinions on this). |
|
To close out my opinion on the main topic of debate, if the core team is confident in a full opening right out the gate as well as the app developers are confident that the apps can handle the additional stress, I would be 100% for it. I just know that once the gates open, they won't be closed and that there is a large sybil/spam presence looming, so I would want to make sure Lens doesn't get overrun by a large amount of unwanted spam early on. |
|
It seems that there is a consensus around the fact that Lens Protocol should be permissionless, which is great and I fully support it. Given Lens is community-driven reaching this agreement helps now to move to the next topic which is the costs. In the current Polygon network indeed there are gas costs as mentioned by @defispartan whivh mitigates network level spam and of course regarding bot activity (which we can see on other social networks as well) costs could mitigate excess profile minting. Not sure exactly what would be the appetite but I would believe 5-10 MATIC could be a good starter when users come from the main claim flow and then give apps credits to mint for free with a weekly allowance. This could be interesting incentive to drive users to the apps and also give the apps control over their own on-boarding strategy based on the results and spam mitigation measures apps could receive more/less credits depending of the user quality. Once there are enough apps supporting with their own on-boarding flows the main on-boarding flow could be taken down to progress decentralization. Would love to hear thoughts on the above. |
That's a clever way to permisionless. Gives more responsibility to apps to manage spam (which is what any social app on the internet should be doing anyways). Also we can define a clear and transparent strategy for credit allowances. If only one app has the perfect onboarding, and most of the credit allowance goes to them, then it's completely fair and merit-based. Which is a great incentive for apps to manage their own onboarding. Also sometimes you can have apps that limit their onboarding start to a very select number and quality of users that they deem relevant for their use-case. Which this kind of program allows. More flexibility and more incentive for building strong onboarding systems. All the while we keep the vanilla claim flow working with payments. Even 10-20 MATIC range is a good one tbh. It's a lifetime access, so it's quite fair to have it gated a bit more with a mild to high price. And prices can fixed and changed according to monitoring metrics at some point. |
|
@ZKJew thank you for opening up this LIP, it's about time. Quick side note about this:
I believe any kind of sophistication around entering any social app is a superfluous and is an additional friction. That's why I'm personally more inclined to direct access approaches. @EthWarrior 's proposal of app-relative onboarding is a decent way of releiving the onboarding UX burden from the protocol, and would mitigate spam on the long term. All the while on paper the social protocol is just open. You either pay, or the app you want in decides how you get in. And then it's to each their own. That being said, appreciate you raising all of these ideas. This conversation is great, and hope it leads to great progress fast! |
|
I wrote a commit maybe 5 months ago to open lens to public was closed but hope this closes with the progress I was expecting, also expecting profile's not to be free so the network doesn't gets spammed also a consensus vote mechanism to allow every interested wallets to be reviewed by community would be nice in this case |
|
On the topic of spam prevention I agree with this approach:
Ultimately imo this is the type of thing where Lens would really shine, different apps will take different approaches and users will have the choice of gravitating towards what they like best. Some may have the option of filtering out activity from users that don't meet certain requirements. We can observe all those trends and get a sense for what works well and doesn't across the ecosystem over time, but aside from logistical questions around gas (what @defispartan outlined makes sense imo with users paying gas for account creation) I don't see that many downsides since I imagine front-ends would adapt to user needs here. I do wonder though if some kind of referral/invite system can be preserved and utilized as a signal of the reputation of an account once Lens is open to all. Detecting high quality vs spam accounts will need solutions, but even without a referral system there is a lot of rich data from the current user base to figure out which accounts are low engagement spam. Curious to hear more about this comment by @joshstevens19:
^and whether the team has a good sense of what those things will look like more concretely or if it's an area of open discussion at the Lens protocol level (vs what individual apps will come up). And also curious to hear about people's experiences with spam prevention today and whether we could do anything more today for spam prevention, for example low quality comments like "Great." being spammed are already hidden on some front-ends. I could see some front-ends for example giving users some type of slider to adjust the spam detection sensitivity (maybe even by comments content vs. profile reputation). Where a really sensitive setting of this could even be based on having a high probability you will like the content/comment regardless of who wrote it; all this to say I think there's some room to be creative with these solutions. Lens is also well suited for solutions that prevent spam since some front-ends already offer the choice between different algorithms. I'm slightly unclear on the alternative/current plan prior to this proposal, but to me seems as good a time as any to open things up to all
|
|
My takes to support onboarding:
|
would you want to fork the request and reset the specifications based on the community consensus? |
|
Great points from everyone, trying to address a few here from the perspective of Phaver as we've been building exactly these things for the last 2 years alongside Lens and actually a lot of these I think should be purely app level, not protocol level, improvements. Permissionless claiming: We believe gating can never lead to a successful social ecosystem as nobody is really qualified enough to fairly decide who to let in and who not, while there need to be mechanisms to curb/remove the bad actors. In Phaver's case this means we actively ban bot accounts (40k+ in last 3 months alone) and use the Levels to give more relevance to users who voluntarily prove their genuinity either by holding certain level of assets or completing KYC or similar free items. For Lens I agree with Stani that keeping a fee on minting is smart at this point to prevent mass-botting while we have also already successfully onboarded 10k+ Phaver users to Lens and have another 10k of very active validated users waiting for their well-deserved free profiles from the app credits. Longer term we of course need to find a balance to let everyone in without any friction to have any chance of mainstream adoption and challenging incumbents, but in the meanwhile, whatever we do, let's not put the friction on transaction level and individuals paying gas per transaction as that totally destroys the UX. User onboarding: User churn: Spam and filtering: |
|
I am in favor of this. |
|
Interesting discussion and nice to see the support and all the takes. @hanahem " Also we can define a clear and transparent strategy for credit allowances." On this I agree that there could be a follow up LIP for defining strategy post permissionless. Ideally its should be community-driven and the design removes the need form the Lens core team the discretion. @RickyEsclapon agree on the gas aspect, its the only small concern and web3 community is still small enough that that gas might be handled and if not, we will break Polygon and figure out the next solution for scaling. Regarding the invites, one idea is to use the public reputation score that is computed into the Big Query that could be a basis. I would say eventually the apps would create their own algos and could lean onto the public reputation score. @joonatanlintala right on point, Phaver's Cred is good example where there is app specific algo that could be used to on-board. Agree as well that there could be small fee in the beginning and as these algos become proven it could be reduced or even removed at some point. But good to start from somewhere. Regarding the spam and filtering as long as we educate the community to report posts and profiles that already goes a long way. For the initial costs 5-20 MATIC is the range now. Would love to hear more thoughts where the number should land? Mind that this would be from the temporary main claim flow (apps would have credits to on-board without a fee with only gas on the mint). |
|
I think this proposal comes just in time. People are seeking for other social apps and Farcaster ist gaining more and more traction. So stay competitive and providing access the garden needs to open. Josh already mentioned the potential of spam which needs to be attacked, but I guess this will develop over time and get better and better. It will be a tough ride, but a good one in the end. |
|
I fully support for Lens to fully open to public. I think this is the right time isn't it? As the attention to decentralised and permissionless social app is currently on the rise right now. But, only if the team are ready and confident to open the flood gate. Probably the main concern is the spam protection but seems everyone already addressed this. |
page not found error |
|
I agree that the lens should be released to the public now, and we should open the garden. I've been using Lens for over 2 years, and it has evolved into the finest product with extensive improvements. Now, it appears to be a market-ready product. The main concern is how to address the spam issue and minimize it. One solution could be implementing a whitelist criteria where users need to stake a certain amount on the platform, acting as a fee to enter the market. This approach could help prevent spam in the garden and ensure a more controlled and quality environment for the existing and new users. |
|
This sounds like the next logical step -- opening up the Lens. I am looking forward to reading more about how this unfolds. |
|
I think we're just getting started! 🌿 |
|
With the development of the project, it seems that users have always been drawn by airdrops. Now it has disappeared, users have lost a lot, and even many apps based on lens have been consumed. My suggestion is to take advantage of this wave of good market, seize the time to send airdrops, and continue to maintain the second and third airdrops in the later period to develop the community. |
There was a problem hiding this comment.
With the development of the project, it seems that users have always been drawn by airdrops. Now it has disappeared, users have lost a lot, and even many apps based on lens have been consumed. My suggestion is to take advantage of this wave of good market, seize the time to send airdrops, and continue to maintain the second and third airdrops in the later period to develop the community.
|
Impressive insights on lens permissions and onboarding strategies. The emphasis on permissionless claiming and tackling challenges like fake signups shows a commitment to a robust and open ecosystem. It's intriguing how Phaver handles user authenticity through the Cred score and addresses potential bot accounts. I'm also looking forward to Orb V2. I love how seamless it is. |
|
Thanks so much for every single person who has commented on here; your thoughts really resonate, and it's super encouraging to see everyone rallying around the protocol. It feels like we're all in this together. Diving into the heart of the matter, the big challenge we will face opening up is the spam issue. You don’t want it to become too easy for folks to mess things up for everyone, especially with the low gas fees. I'm on board with the idea that there should be a cost to join Lens, something like 5-10 MATIC seems fair. And with the payment option which exist today to use crypto, fiat, or even a card to pay, it makes jumping in pretty smooth for everyone. But here's the twist - I personally think it should be the apps, not the protocol itself, that take care of bringing users on board. We've got some amazing apps out there that can handle this perfectly. This got me thinking about using credits to kickstart the onboarding process for major apps, making it essentially free for them at first. We could then check in and see how it's going, ensuring the process keeps spammers at bay. If an app starts to slip or gets exploited, we could pull the plug on their credits. Apps could also set their own fees for onboarding, which could open up new revenue streams for them. The cool thing about smart contracts is they offer so many flexible ways to manage access, whether through NFTs, funds, or other creative methods. On another note, we've been leveling up our machine learning game. Now, each profile gets a score from 0-100%, helping us spotlight the high-quality ones and keep the low-quality or spammy profiles out of the limelight. While we can't completely eliminate bots, making them less visible helps preserve a quality experience for everyone which we be adding ordering of these % throughout the lens API (of course apps can opt-out). I am hoping to also see apps build their own algorithms with the public big query dataset to eliminate the bots as well. Lens also offers a gasless experience, which is key for usage on the protocol. The beauty of the blockchain allows users to cover their own costs if needed, but our API aims to sponsor genuine profiles while sidelining the bots, again hopefully helping bots. I'm really excited about this hybrid approach. It not only helps us onboard users effectively but also leverages our ML advancements to improve profile visibility. This could be a win-win, empowering apps to stand on their own and contribute to the ecosystem. |
I am 100% with this |
|
200% with it,let's go man!!!🤗 |
|
Thank you for initiating this discussion on the LIP; I wholeheartedly endorse the idea of opening up the protocol to the wider community. It's clear that there's a strong demand for it, and I believe it's time for Lens to embrace this openness. However, it's important to consider the potential spam issues that may arise once the protocol is opened up. We need to implement certain measures to ensure that the overall user experience remains positive for most applications. I'm fully in support of this LIP and am eager to see it progress. This is a request we've been hearing for some time now, and it's imperative that we listen to the community and meet their needs. I'm looking forward to hearing everyone's thoughts on this matter. |
"This got me thinking about using credits to kickstart the onboarding process for major apps, making it essentially free for them at first." I'm not a fan of this privileged-app approach as it feels like we're going in the opposite direction from permissionless. I'm also skeptical on how that would help the spam situation when nearly all of the spam came from users that onboarded through Hey. Most of the major apps on Lens, which recently raised millions, don't really need the help. It's the smaller independent developers that would benefit from some type of credits. If I'm launching a bootstrapped competitor to Phaver or Orb I'm at an extreme disadvantage, having to pay 5 - 10 MATIC for each user I want to onboard, when it's free for them (despite venture backing). |
|
Great insights on spam prevention at different levels, onchain, offchain, and profile. But anyway this is the time for opening Lens, and the proposed measures like onchain gatekeeping and offchain algorithms align with the community's vision for a permissionless protocol. Concerns about spam, profile reputation, stress testing, and differentiation are valid. A wellthoughtout onboarding flow, differentiation, and targeted marketing can be the key. Also, incentivizing developers and creators seems necessary for Lens's success. Exciting times are ahead for the Lens community! Cheers. with Love and Respect. |
Building new client here as well and agree that creative, differentiated and innovative apps should be not put at a disadvantage especially when bootstrapping or launching. Lens true power lies in a vast amount of apps and new creative experiences. Spam is enemy #1 The thesis has always been however that Apps would compete to create better algorythms and improve the protocol. The problem/risk being that users that have a bad experience on one app have a tendency to think the entire protocol is bad. Weeding out spam "should" be apps responsability provided they are given the tools to do so at the API level? For example: – If Apps had their own Lens domain name on profiles creation (ex: UserA.hey or UserA.phaver) then App could have the ability to block profiles their content from a spammy client or undesired content? If we can't prevent spam at the profile creation (the real problem being spammy content) then maybe App can be given more options or tools at the API level to weed it out? |
|
Bluesky opening up this week
https://bsky.app/profile/retr0.id/post/3kkswdei6hb2x https://bsky.app/profile/bossett.bsky.social/post/3kkswrmuaed2c |
|
Hey all there is going to be a community call on Lens about LIP through Buttrfly Spaces this Friday so anyone can join for final thoughts https://buttrfly.app/space/xlp-pgtj-dyo |
|
Looking forward to the spaces this Friday on Buttrfly 👍 |
|
🔥 |
|
i'm so ready for the garden to open up |
|
Hi everyone, I'm glad to see this proposal coming to Lens, I think that eventually the protocol should be permissionless. However, like most of the people who have spoken in this PR, I'm afraid that if it isn't managed very carefully, openness will seriously damage the protocol because of the numerous spammers. This solution would limit the risk of a wave of bots claiming handles unduly. Also, I don't particularly agree with this idea, which seems rather risky to me: if an app doesn't manage to protect well against bots, the whole protocol will be impacted. This is also true in the case mentioned, where apps are put "in competition" with a certain allocation of accounts. It would only take one hack at app level to impact the whole protocol, and even with all the goodwill of a team, it's sometimes not possible to guard against all the risks. In any case, I'm really looking forward to seeing Lens open up, with only the bots to deal with at first. |
|
I'd love to explore using Lens protocol for online education credentialing and building an "education persona" for each identity. I've been working on blockchain + education for the last 6 years at BitDegree.org and the identity was always a problem. Lens solves this. Let me know with whom we can explore this. |
|
Just open up lens for public use.. |
|
Once we open up, there will be room for deep level protocol and app-level assessments. For a start, there should be a fee to mint a profile. Apps can be given some free invites to onboard but it should be strictly monitored. Each app will eventually come up with it's own stricter measures of combating bots and farmers using different metrics just like the way Orb and Phaver is tackling this. I'm happy we are about to open the garden. |
|
I beleive lens should open free to use for everyone. The dapps must his own rules to combats bots and for spammers the team set by stani is enough. Lets go |
|
I think it's time to open up the garden and let things expand. One thing I do believe is crucial is to mitigate spam and bots that try to flood the graph, as it's a given in this space. Charging for each new account created is something I don't necessarily agree with, though. I think there are other avenues to work around fighting off bots, whether that be captcha blockers during the onboarding process or some other verification method. Many people will look to join Lens, and if they get hit with a fee (even minimal), most will turn away. Relaying what @joshstevens19 said above, "I personally think it should be the apps, not the protocol itself, that take care of bringing users on board. We've got some amazing apps out there that can handle this perfectly. This got me thinking about using credits to kickstart the onboarding process for major apps, making it essentially free for them at first. We could then check in and see how it's going, ensuring the process keeps spammers at bay. If an app starts to slip or gets exploited, we could pull the plug on their credits." This avenue would make a lot of sense, in my opinion. In any case, I will have to brainstorm a bit more, but I think opening up the garden is a must at this point. Just need to navigate how to do so in a manner that prevents bots and is a net positive for the Lens ecosystem. |
|
I think it is long overdue to open the protocol. if cost are an issue, other protocols have shown users willingness to pay |
|
well done all! LGTM |
There was a problem hiding this comment.
can you please restore the LIPs/lip-template.md file and we can then merge @ZKJew
Requested Lip-5 proposal to progressively open Lens Protocol.
When opening a pull request to submit a new LIP, please use the suggested template: https://github.com/ethereum/LIPs/blob/main/lip-template.md