Full recording of the meeting can be found here.
Meeting Minutes below:
Patrick intro - welcome to monthly app miner meeting
- Patrick - want to give a little more of a vision of what the future of App Mining should look like. As it currently stand, app mining is actually pretty decentralized, we don’t vote, try to pull in best app reviewers. Over time the goal is to make app mining directionaly better month over month. In a future state app mining should become way more decentralized, and so the program will need to get better on a self evolving/protocol level. Purpose is to attract highest quality devs (you all on the call) and eventually end users. Super important that as app mining becomes more decentralized, it remains so fair that it is attractive.
Going to go through issues from this spreadsheet one by one: https://docs.google.com/spreadsheets/d/1AJR_qdXZBifgG5U7mnVMIGpWV1jRxYEdwRsREvXxnlY/edit#gid=0
-
Fraudulent voting on DE
-
We can tie critic back to an investor, but are as suspicious given the circumstances, same .id. We shouldn’t have the ability to censor votes.
- One question is if this was caused by the bug that happened on the same day?
- When DE found the bug, removed the votes that were accidentally granted, and patched the bug. Couldn’t have come from this as the votes would’ve been revoked and would not have been counted.
- Patrick: may make sense to have a third party auditor of votes or security audits
- One question is if this was caused by the bug that happened on the same day?
-
Gameability of Product Hunt votes .
-
Anyone can pay Product Hunt to re-feature their apps.
- We’re in touch with the PH team, and one potential solution is to have them agree not to do this for App Mining participants.
- PH team also believes they have an amazing system to minimize voting fraud. Haven’t seen any evidence to the contrary so far, but it’s something we should keep an eye on. If they’re serving the community well they should be kept, if not then they should be replaced. We need to be dispassionate with app reviewers.
- With the transparency side, all they have to work with is the number from their internal score, not a lot of clarity on why they do or don’t like specific things. Need more information.
- They haven’t been super awesome here from our perspective as well, and it’s something we’ve been prodding them to improve.
- PH reranked every app this month to make sure all apps were being ranked by the same people.
-
Mobile App disadvantage for PH , TryMyUI, and having app specific instructions for testing .
-
Stealthy is not the only mobile app: Afari and Timski are two other examples. Bugs with blockstack could have affected rankings. Instructions from TryMyUI directed testers to do things that you would never do with a chat app (upload a file).
- The question here is how much standardization vs. customization do we want here with TryMyUI?
- Should the burden on instruction here be on the App?
- They have the same disadvantage if they just come across their app in the app store - the creators need to provide an onboarding experience that educates the user.
- There’s really valuable user feedback here.
- How do you make your app understandable for someone who isn’t going to read instructions
-
Comes to an issue of should something only look good? Does that add value? Gets into the discussion around including Gaia. The values of the platform should be considered here too.
- The push back we get from investors and peers is that the apps are clunky and don’t work - so UX is one of the biggest issue and should be focused on.
- We completely agree with having a requirement for Gaia, the question is when. This is already in the works.
- What can we tweak with how TryMyUI is used to improve them short term as an app reviewer and then how can we bake that in long term.
- If a user gives an app a great score but the app isn’t usable, what’s the value in that score?
-
Why isn’t part of the ranking usage?
- As of yet no one has thought of a way to do that and measure churn in a way that doesn’t violate our ethos around tracking users.
- We want this and absolutely would love it, need to find a way
-
Right now, if I’m 100% going off incentives, shouldn’t I just use Auth but keep the rest centralized early on?
- Goes back to differentiated full fledged Blockstack apps vs. those that just use Auth. We all agree, and are moving in that direction.
- Our highest priority for engineering right now is platform reliability, so we are focused on this as well in terms of cutting out bugs/disincentives from using the full stack.
-
Community Voice as a replacement for DE Voting: Why are people who are actually in the community able to vote on apps right now?
- This goes back to the token offering - we took an extremely conservation/regulatory compliant approach. This is why only accredited investors currently have tokens and can vote. We are getting very close to the token being offered to everyone - see it more as a regulatory issue.
- Also practical questions solve for such as sybil attacks. Would you let chinese citizens vote in an american democracy? How do you draw a line on who can vote and how do you minimize the number and ability of bad actors?
- These folks do have skin in the game, and this also should expand as soon as possible.
- Should there be vote decay added? Should 40 people have this big of a say?
- One of the biggest issues with voting is voter fatigue.
- How do we get people to diligence apps and vote every month?
- We have some ideas here we’re working on, and may involve voter decay/voter incentivization.
- How do we get people to diligence apps and vote every month?
- One of the biggest issues with voting is voter fatigue.
- For right now though when there aren’t many voters, should one voters votes be able to skew the rankings that much?
- Great point, but appears to be more of a temporary pain. Most we can do at the moment to continue to do this safely is just to onboard as many token holders as possible and convince them this is worth taking part in.
-
Clarity on how bringing in external well funded apps helps the ecosystem
- The goal of App Mining is not to hit milestone 2, although it may certainly help drive towards that.
- The goal is to have apps created in the blockstack ecosystem, and then to get those apps to be increasing in terms of quality
- We’re not laser focused on Milestone 2 the way we were on Milestone 1. It is not something we consider material to the success of Blockstack - so we are not making any decisions purely based on trying to hit Milestone 2.
- The evidence shows bringing in external apps (SpringRole being the example) does bring them into the ecosystem and can make them take on our values. SpringRole has Gaia integrated as of today, it just took some time, planning, and dev work.
-
App reviewer that ranks based on performance metrics
-
covered above, but also what do you want to measure? Human attention? Whether someone’s logged in after 7 days? Understanding what we want to measure is a super important step here, but then also need to measure that in a way that doesn’t compromise our ethos/values/capabilities.
-
Match teams with a mentor - eventually we want to hold a growth meeting with App Teams on getting product market fit.
- Problem of how does this scale? If there are 1000 apps are there 1000 mentors?
- This is something we want to find a solution to, but in a way where as time goes on Blockstack PBC can be less and less invovled.
-
Apps that don’t work should be disqualified - if we disqualify apps it impedes the ability of the system to self govern and decentralize long term. We see this as a pain that needs to be addressed, but the route problem does. As time goes on we think having more apps involved and more app reviewers solves this problem.
- Let the game play out and make the game better rather than interfering with the game.
-
Originally apps were required to be open source. Many do not seem to be. Shouldnt this community expect that open source apps are the beneficiaries of app mining, not closed source apps?
- Interesting, do you think we should move to open source requirement?
- Goes into having some sort of decentralization score - open source would be a criteria there.
- We don’t see it possible/scalable to rank apps based on evaluating their code base. An open source requirement at best would be binary.
-
Rewarding too many apps? Interesting question, but does concentrating the rewards actually solve any problems? Interesting but not enough to stop paying all apps right now, especially looking forward to rewards increasing.
-
Categories - agreed, need to appear. Timing isn’t quite right, maybe as we get closer to 100 apps. But again, agree this should and will happen. Still thinking through how best to implement.
-
Question around continuing to allow TestFlight only mobile apps - Going forward we would like to move away from Test Flight only offered mobile apps. The app should be publicly available to everyone.