Aral Balkan @aral@mastodon.ar.al

@ben @inmysocks @ekaitz_zarraga “Dominant” in P2P is a failure mode.

Blog: Boogie Board: a beautiful, modern, portable take on the blackboard

https://ar.al/2018/12/16/boogie-board-a-beautiful-modern-portable-take-on-the-blackboard/

(Not to mention inexpensive at $35-$45. I got mine delivered from the US to Ireland for ~€50.)

@deni @ekaitz_zarraga Any business model that can achieve exponential growth so that they can have their exit (can be sold). None of the business models capable of fulfilling that core requirement are compatible with our welfare as individuals or with the continued existence of our habitat or of the commons.

@fabricedesre You can store it anywhere, DAT is agnostic about that. So in a single file, in a key-value store/database (see hyperdb), multiple files (see hyperdrive), DNA*…

* TBA ;)

@Danielle

@strypey @rick_777 Ah, gotcha. Lack of threading in the interface alters the semantics :)

@ekaitz_zarraga @inmysocks Yeah, and the code I was just testing tonight for hyperswarm is anything but smoke. It works like a charm :)

@ekaitz_zarraga @inmysocks I’m fine with stealing good ideas from bad enterprises for a good cause :)

@fabricedesre You don’t store the content in the append-only log. The history that is stored there cannot be deleted without deleting the history of the object (timeline, document, etc.) but you can use granular objects and delete the whole shebang. Whatever is your root “identity” log will have history that cannot be deleted unless you delete that identity.

@inmysocks @ekaitz_zarraga This is all you need to know about IPFS.

@fabricedesre That’s what’s great about DAT (& what separates it, from, eg., SSB): you _can_ delete content. And even if you implement a CRDT on top, as I need to do, you can still delete content if you use a WOOT family of CRDTs (eg., LSEQ). Also, it does sparse replication. I don’t know if anything better at the moment for what I want to build and fuck knows I’ve been looking for years :)

@ekaitz_zarraga The difference is that IPFS is backed by fucking venture capitalists and can go die in a fire (I mean that in the nicest possible way) whereas DAT is backed by a not-for-profit and created for all the right reasons and without a fucking blockchain in sight :)

Oh my fucking goodness, hyperswarm + multifeed DAT works like a charm under real-world circumstances.

I can’t wait to try out kappa-core next.

More excited than I’ve been in a long while. The DAT folks deserve a fucking medal.

@brennen Thanks for sharing :)

Blog: Kappa Architecture Workshop

https://ar.al/2018/12/15/kappa-architecture-workshop/

@bob Hehe, I was just playin’ with you :)

Blog: Remember directory from last session in Tilix with zsh

https://ar.al/2018/12/15/remember-directory-from-last-session-in-tilix-with-zsh/

@bob But what does that have to do with your hair, Bob? I’m so confused :P

Following along with the excellent Kappa Architecture Workshop (https://kappa-db.github.io/workshop/build/01.html)

Multiple writer P2P replicated chat in under 100 lines of code, did you say? Sure: https://github.com/aral/kappa-architecture-workshop-work-files/blob/master/multi-chat.js

(Very simple example.)

Check out the excellent workshop if you’re interested in creating P2P apps.