"The new channels are terrible."

1. The old channels weren't working. So we opted to make a big, breaking change. Inevitably, some people were going to be upset. No risk, no reward.

2. We're actively improving channels. See the summary cast below for what we've already improved and what's on our radar. 

3. Our next big project is onboarding new users directly to a channel so they have a Day 1 channel to cast in and a community of people.

4. If your claim is "no one likes the new channels and they didn't take user feedback into account", see the second cast below for a slice of the network happy with the changes. Again, not claiming perfection, but let's start from a baseline that the changes worked for some users and did not work for others. And we're working to improve them for the people it didn't work for.

6. If you're still frustrated with channels, feel free to reply to this cast with what's not working for you.

https://warpcast.com/dwr.eth/0x9825d19e
https://warpcast.com/dwr.eth/0xcf622e07

1. I'm sorry your frustrated. Happy to help try to improve the situation.

2. What channels are you missing out on? Have you reached out to the channel owners for an invite?

***

> Improve the algo
Channel owners were unhappy with the algo and wanted more control. And most channel owners didn't want to spend all their time fighting spam.

Additional context:

Warplet

1. I'm sorry your frustrated. Happy to help try to improve the situation.

2. What channels are you missing out on? Have you reached out to the channel owners for an invite?

***

> Improve the algo
Channel owners were unhappy with the algo and wanted more control. And most channel owners didn't want to spend all their time fighting spam.

Additional context:
https://warpcast.com/dwr.eth/0x8bcd3f4b

I understand the reason who did it. 

But X and Threads are full of spam and low effort posts. Yet their algo manages to pick up great, interesting posts from all the trash. 

That’s not an easy ask: that’s why X and Threads are worth billions. 

But making channels exclusive to posts just discourages participation. So many great posts won’t be visible as ppl can’t go access those channels. 

I want to join ethereum and Farcaster channels.

Hear you, the challenge with making participation open, i.e. hashtags

1. Most channel owners want more control and don't want to give it up to an algo; if it's purely algorithmic, they don't feel like they are building a community

2. Content moderation is a ton of work; easy to weed out clear spam, but boring content is really hard.

3. Channels don't give any extra distribution because of this -- we can only boost content we know is interesting.

4. We're going iterate to give channel owners to make it easier to join channels, but we still think some friction is good for community building

***

X / Threads (i.e. Instagram) — multiple orders of magnitude more content, not decentralized (so can remove spammy accounts without anyone protesting and none of those apps have figured out how to scale communities like channels (Twitter has tried multiple times and failed). Twitter hashtags are super spammy even today.

Thanks for the feedback. Some growing pains because it was a big change, but feedback like this will help us to keep iterating and make it clearer.

Appreciate the effort to explain everything.

I'm bullish on what you are building and really want you to succeed.

But I don't want other creators to join Warpcast and then get frustrated because they cannot post into the channels they care about.

Hope we can find a way that it works for all.

1. Channel owners create incentive for people to build them as communities

2. We spent a year trying to make channels work as hashtags. They didn't work. People get frustrated about the algorithm. Doesn't decentralize well, either (channels should be consistent across clients, 100% open source algorithms are too easy to game).

3. They new model is closest a text-based Twitter Space -- some people on stage, a larger number of people in the audience.

4. You're underestimating how much spam will exist on a permissionless, decentralized protocol. Easy enough to catch 80% of it, but the last 20% can still be a large number.