@abucci@anthony.buc.ci & @prologic@twtxt.net: thank you for resolving the issue with refresh property!

@win0err@kolesnikov.se I guess it worked after @abucci@anthony.buc.ci nuked his pod’s cache? 🤔

it does raise an interesting question of whether I should try to re-implement the exponential back off algorithm again?

I tried once, but I couldn’t get it to work just right

@abucci@anthony.buc.ci Haha 🤣 You jerk 😜 (kidding) – Would you have time to go over the implementation/algorithm I once tried but was never happy with? 🤔 🙏 Maybe we can bring it back and improve/fix it.

@abucci@anthony.buc.ci Also sorry about the confusing shit™ last night (it was late for me) 🤦‍♂️ I should have jut said:

1. Go to Setting
   
2. Poderater Setting
   
3. Reset Cache
   

👌

@abucci@anthony.buc.ci Here you go!

https://git.mills.io/yarnsocial/yarn/pulls/1108

For some context, I was trying to work out a way to calculate the moving average of a feed and how often it updates and converge on this. So feeds that update frequently are always fetched at the fetch interval configured by the pod operator (default 5m) and feeds that update slowly are fetched less frequently.

For cross-pod fetches, this has no effect as pods now should be using WebSub by default (depending on the version) and at some point I might also add WebSub support to the feeds.twtxt.net service 👌

@abucci@anthony.buc.ci Yhe diff? 🤔 Is it? 😆