Content pfp
Content
@
0 reply
21 recasts
21 reactions
phil pfp
phil
@phil
Two things can be true The Merkle team has been incredibly transparent about their reporting of user metrics, including hiding suspected bot accounts from the client experience An open protocol will have a high amount of spam and other machine driven activity that can inflate otherwise useful KPIs
3 replies
9 recasts
130 reactions
ildi pfp
ildi
@ghostlinkz.eth
Would you find it useful if Merkle were to create distinction between accounts and users? Accounts includes everyone and users would be something like all power badge holders.
1 reply
0 recast
2 reactions
phil pfp
phil
@phil
that will just create more spin what if i have two accounts: a main and an alt that each post on different topics is that two users? how would they know? this feels like bikeshedding instead of tackling the tough question of how to grow the protocol DAU
1 reply
0 recast
1 reaction
ildi pfp
ildi
@ghostlinkz.eth
I think it’s really important for any network to try to accurately report metrics. My profile says I have 36k followers, but in reality I bet less than 3k are actual DAU who do anything meaningful for the protocol. This is simply wrong and very misleading.
1 reply
0 recast
1 reaction
phil pfp
phil
@phil
if you think that the number should show something else, you're welcome to go use a client that does that if others agree, it will gain users this is like arguing over what % of SMS messages or emails are spam, which no one even tries to do instead, what actually matters is the experience for users & developers
1 reply
0 recast
2 reactions
ildi pfp
ildi
@ghostlinkz.eth
If we are properly optimizing for what developers require, then we should see growth in client ecosystem. Im looking at the raw protocol data and thinking “oh wow, this is a mess”. But maybe this mess is a better version than all the other options out there. Only time will tell I guess.
1 reply
0 recast
1 reaction
phil pfp
phil
@phil
Raw protocol data will always be a mess. It's up to clients to sift through this data for signal to present to users. That is the hard part and where we should be focusing our energies, not on bikeshedding around terms or trying to create legalese policies no one will read for a shared protocol.
1 reply
0 recast
2 reactions
ildi pfp
ildi
@ghostlinkz.eth
But Warpcast was doing a terrible job at hiding the spam as of weeks ago. The other clients are really far behind on that front. Neynar also ignored it and are now prob working on something. This is why users like me are frustrated. Cuz we can’t help but question the intentions of low quality DAU growth.
1 reply
0 recast
1 reaction
phil pfp
phil
@phil
Don't underestimate how much work goes into just keeping the client stable and running. Warpcast has done more to fight the spam problem in just a few weeks than most social media companies have done with years of tackling the same problem.
1 reply
0 recast
1 reaction
ildi pfp
ildi
@ghostlinkz.eth
I understand. I built a very bare bones client and it became completely useless after degen spam took over. Turns out my filtering wasn’t strict enough. The infrastructure we are all building on top of can also change so you need to stay on top of everything.
1 reply
0 recast
1 reaction
phil pfp
phil
@phil
Yeah, it's a high bar. I expect that there is an opportunity for corporations like Merkle and Neynar to offer APIs as a service for filtering, at the risk of centralizing a key piece of infrastructure. Otherwise, we have several decades of lessons in Bayesian and ML filtering to learn from in the SMTP ecosystem.
1 reply
0 recast
0 reaction
ildi pfp
ildi
@ghostlinkz.eth
I’m always looking at the client ecosystem for signals for how well everything is going and how excited devs are to build and maintain their projects. At the moment I wouldn’t say I’m impressed, but let’s see where we end up by EOY.
1 reply
0 recast
1 reaction
phil pfp
phil
@phil
Seems like a fair position to take.
0 reply
0 recast
1 reaction