Notices by Paul Haddad :tapbots_logo: (paul@tapbots.social), page 5

So the AMA ended up being 20k almost exclusively negative comments and 14 non answers that just pissed people off more.

Dave Chappelle has some salient advice for next time:

I don't think a 2 day Reddit blackout is going to do anything. Guessing they'll just wait it out and do nothing.

If it was set up as "until things change” it'd be a lot more interesting. Mostly because I think Reddit would panic at some point and do something really stupid, which would be fun to watch.

I have two questions.

1. How much did Apple pay for this hashflag?
2. Why did Apple pay for this hashflag?

https://www.macrumors.com/2023/05/30/wwdc-2023-hashflag/

Neat.

I'm sorry, if I'm buying a $4k machine the last thing I want to see is a tip % option. I'm all for them to make more money and have better conditions but no way this crap.
https://www.bloomberg.com/news/articles/2023-05-03/apple-s-unionized-store-workers-seek-tips-and-higher-holiday-pay

Oh my!

Man at those speeds its almost tempting to just self-host a bunch of stuff and downgrade our hosted server. Not going to, but it sure is tempting.

This doesn't really answer the question *, sure clients will have various monetization models (though WTH is transactions). The issue is what is the server/Bluesky PBLLC business model?

Are they going to start putting ads in the feed? Are they going to charge users or clients for access? Stay free and make it up in volume? I have no clue, not sure anyone does either.

* Image copied from https://mastodon.social/@atomicpoet/110280403283967447 since I don't know how to get a web link to that and there's no quotes here…

Just for the record, I wish Bluesky good luck. Being founded by Jack doesn't bother me. Being a Mastodon "competitor" doesn't bother me either. Seems like there's some good people working on it and if it causes more folks to stop using Twitter 2.0 that's a win in my book.

That being said I’m not going to be looking at it anytime soon, it's too early from both the API and specially # of users standpoint.

New hobby. Googling "As an AI language model" site:amazon.com.

Twitter finally announced their “Basic" API access. Completely useless as expected.

Things of note:
They totally disabled access to the V1.1 API (probably because they had no one that could implement these limits on it).
The V2 Tweet Cap went from free 2M Tweets per month, to 10K Tweets per month for $100.
The rate limits on a bunch of calls have been slashed significantly.
The sign up link for the Enterprise API, you know the thing that costs $42k/month, is just a Google Form doc…

When creating a new account we show a list of ~40 suggested follows (should be the same as explore -> for you on the web client). I'm debating just auto-selecting them (possibly with a “skip" button).

I don't want to be accused of forcing people to follow other users, but the first launch experience when not following anyone is pretty terrible.

My guess is that most will tap continue without even reading anything. So it's an auto-follow but for a good purpose...

Thoughts?

Since a few have asked. The list comes from the individual instance (mastodon.social in this case). The admin can make changes to the list, a user can opt themselves out of it. It's not separated by any kind of “category”. Outside of the two company accounts we just show what the server returns.

My gut says that anything short of having a diversely populated timeline on launch is going to cause most users to run the app once, delete it and then go complain on Twitter that Mastodon is too hard.

And wired is reporting the same numbers too.
//ht @shiruken

https://www.wired.com/story/twitter-data-api-prices-out-nearly-everyone/

If I'm reading between the lines correctly, they are merging the Enterprise + Public APIs into a single product. The old Public API calls will all get put under a Tweet cap (50M Tweets for this small plan). The Tweet cap counts each Tweet returned by the API call.

50M seems like a big number, but as a point of reference before Tweetbot 6 came out, I implemented a single capped API call and sent it out to ~100 beta testers. They blew through the 500K cap we had in about 2 days.

I've seen this $42k/month number for the lowest level of real Twitter API access mentioned a number of times. Image could be made up but seems credible.

https://twitter.com/BlackForestBoi/status/1633983548708253696

More leaked prices for the other enterprise plans.

We never had any kind of analytics to get a better # but my gut says a Twitter client would probably download 100k-250k Tweets/month per account per device. So if we get about 200 people willing to pay $400-500/month we could come back!

https://twitter.com/stokel/status/1634262169721221120

I think this onboarding update ended up coming out pretty nicely. Let's you create a new account without a lot of fuss.

I also popped in auto-complete as part of login, mostly because I got so tired of typing stuff 100 times while testing onboarding. Though still surprised there isn't a built in UIKit component for this.

"Better to talk to people than communicate via tweet." says the guy who only “communicates” via Tweets.