Closing Keynote: The Jump to Hyperspace

The Jump to Hyperspace 🌈 Light speed, antientropy, & moving past the cloud 🌤

I suppose it is tempting, if the only tool you have is a hammer, to treat everything as if it were a nail ~Abraham Maslow

[…] by 2025, 75% of data will be processed outside the traditional data centre or cloud ~ Gartner (also Dell & IBM) https://www.gartner.com/smarterwithgartner/what-edge-computing-means-for-infrastructure-and-operations-leaders

Brooklyn Zelenka @expede

Brooklyn Zelenka @expede • CTO at Fission • https://fission.codes, @FISSIONCodes • Infra & browser SDK for “edge apps” • Local-first, E2EE/EAR, distributed, passwordless • Standards: DIF, UCAN, Ethereum, Multiformats, others • Elixir: Witchcraft, Algae, Exceptional, &c

Meta 🔮 • R&D at Fission and others • Future looking, emerging space • Edge, web3, dweb • Go “up a layer” from BEAM to the internet itself

“Historical Reasons” How We Got Here 📠📲

Motivation 🎭 1994 Set the Tone

Motivation 🎭 The Dream of the 90s is Alive on the Web 🎶

Motivation 🎭 The Dream of the 90s is Alive on the Web 🎶 💁 🖥

Motivation 🎭 The Dream of the 90s is Alive on the Web 🎶 💁 🖥 🐢

Motivation 🎭 The Dream of the 90s is Alive on the Web 🎶 💁 🖥 🐢 🗃

Motivation 🎭 The Dream of the 90s is Alive on the Web 🎶 💁 🖥 🐢 ⚙ 🗃

Motivation 🎭 The Dream of the 90s is Alive on the Web 🎶 💁 🖥 🐢 ⚙ 💪 🗃

Motivation 🎭 The Dream of the 90s is Alive on the Web 🎶 💁 💁 💁 🖥 🐢 🖥 🐢 🖥 🐢 ⚙ 💪 🗃

Motivation 🎭 The Dream of the 90s is Alive on the Web 🎶 💁 💁 💁 🖥 🐢 🖥 🐢 🖥 🐢 ⚙ 💪 🔐 🗃

Motivation 🎭 Scaling Up 💁 💁 💁 🖥 🖥 🖥 🐙 ⚙ ⚙ ⚙ 🔐 🔐 🔐 🗃 🗃 🗃

Motivation 🎭 Scaling Up 💁 💁 💁 🖥 🖥 🖥 ⚙ 🗃 🔐 🐙 ⚙ 🗃 “Cloud” 🔐 ⚙ 🗃 ☁ 🔐

Motivation 🎭 Less is More 💁 💁 💁 🖥 🖥 🖥 ⚙ 🗃 🔐 🐙 ⚙ 🗃 “Serverless” 🔐 🗃 (AKA more ⚙servers) 🌩 🔐

…and so it was for many years…

…and so it was for many years… 🦖☄🌋🌾🏰🏢🚀

Motivation 🎭 Natural Consequences 🍃

Motivation 🎭 Natural Consequences 🍃 • Single source of truth (“the” database)

Motivation 🎭 Natural Consequences 🍃 • Single source of truth (“the” database) • Server-centric • “Full stack development” • DevOps, Docker, k8s • How to train enough engineers?

Motivation 🎭 Natural Consequences 🍃 • Single source of truth (“the” database) • Server-centric • “Full stack development” • DevOps, Docker, k8s • How to train enough engineers? • Infrastructure Hegemony • AWS (60%), GCP, Azure

Motivation 🎭 Sending a “Direct” Message

Motivation 🎭 Sending a “Direct” Message ⏳🔋🛢🪠

New Environment 🛰 Users vs Cloud Infra Source: AWS

New Environment 🛰 Users vs Cloud Infra 6 7 6 2 1 1 1 1 Source: AWS

New Environment 🛰 Users vs Cloud Infra 6 7 2 6 50M 1 1 1 1 Source: AWS

New Environment 🛰 Users vs Cloud Infra 371 million 56M/centre 6 7 2 6 50M 1 1 1 1 Source: AWS

New Environment 🛰 Users vs Cloud Infra 371 million 56M/centre 6 7 2 6 50M 1 1 ~435 million 435M/centre 1 1 Source: AWS

New Environment 🛰 Users vs Cloud Infra 371 million 56M/centre 6 7 2 6 50M 1 1 ~435 million 435M/centre 1 ~1.4 billion 1400M/centre 1 Source: AWS

Video Killed the Radio Star A New Environment 🏝

New Environment 🛰 New Environment

New Environment 🛰 New Environment Then 📠 Now 🚀

New Environment 🛰 New Environment Need Then 📠 Now 🚀 Convenient 💁 Critical 🚨

New Environment 🛰 New Environment Then 📠 Now 🚀 Need Convenient 💁 Critical 🚨 Location Data Centre 🏢 Powerful Clients (M1, IoT) ⌚🚙👟

New Environment 🛰 New Environment Then 📠 Now 🚀 Need Convenient 💁 Critical 🚨 Location Data Centre 🏢 Powerful Clients (M1, IoT) ⌚🚙👟 🖥 📱 Access

New Environment 🛰 New Environment Then 📠 Now 🚀 Need Convenient 💁 Critical 🚨 Location Data Centre 🏢 Powerful Clients (M1, IoT) ⌚🚙👟 🖥 📱 Bandwidth 🚚 Latency ⏲ Access Bottleneck

New Environment 🛰 New Environment Then 📠 Now 🚀 Need Convenient 💁 Critical 🚨 Location Data Centre 🏢 Powerful Clients (M1, IoT) ⌚🚙👟 🖥 📱 Bandwidth 🚚 Latency ⏲ 🇺🇸🇪🇺 🌎🌍🌏 … 🌔 Access Bottleneck Market

High Volume 📢 Science Fiction → Reality • Remote surgery • Extended reality • Autonomous vehicles • Location transparency • Competitive cloud gaming • Realtime manufacturing • Continuous ML Literal brain surgery over 5G 3,000km ~ 1,900mi

Sensor data explosion will kill the cloud. […] existing infrastructure will not be able to handle the volumes or the rates We are absolutely going to return to a peer-to-peer computing […] not unlike distributed computing We are going to move to a world of data-centric programming. ~ a16z, “The End of Cloud Computing”

Sensor data explosion will kill the cloud. […] existing infrastructure will not be able to handle the volumes or the rates We are absolutely going to return to a peer-to-peer computing […] not unlike distributed computing We are going to move to a world of data-centric programming. ~ a16z, “The End of Cloud Computing” ✅

Sensor data explosion will kill the cloud. […] existing infrastructure will not be able to handle the volumes or the rates We are absolutely going to return to a peer-to-peer computing […] not unlike distributed computing We are going to move to a world of data-centric programming. ✅ ~ a16z, “The End of Cloud Computing” ✅

Sensor data explosion will kill the cloud. […] existing infrastructure will not be able to handle the volumes or the rates We are absolutely going to return to a peer-to-peer computing […] not unlike distributed computing We are going to move to a world of data-centric programming. ✅ ~ a16z, “The End of Cloud Computing” 🤾 ✅ 🗑

The World is Changing Relativistic Computing 🛸

The ship that made the Kessel Run in less than 12 parsecs ~ Han Solo

New Environment 🛰 Moore’s Law. But make it networked.

Low Latency 🐇 Latency is a Physical Limit 🚧 • Bandwidth max not even close • Speed of light causality • Edge dominates < 40ms • Best at ~8ms • 1ms applications exist • “Ultra Reliable Low Latency”

Low Latency 🐇 Latency is a Physical Limit 🚧 • Bandwidth max not even close • Speed of light causality • Edge dominates < 40ms • Best at ~8ms • 1ms applications exist • “Ultra Reliable Low Latency” f Source: Ericsson http://cscn2017.ieee-cscn.org/ iles/2017/08/Janne_Peisa_Ericsson_CSCN2017.pdf

Low Latency 🐇 What 8ms Looks Like

Low Latency 🐇 What 8ms Looks Like Austin ➡ San Francisco Ideal Vacuum 💫

Low Latency 🐇 What 8ms Looks Like Austin ➡ San Francisco Ideal Vacuum 💫 Austin 🔁 (almost) Atlanta Ideal Vacuum 💫

Low Latency 🐇 What 8ms Looks Like Austin ➡ San Francisco Ideal Vacuum 💫 Austin 🔁 (almost) Atlanta Ideal Vacuum 💫 Austin 🔁 New Orleans Ideal Fibre 🧶

Low Latency 🐇 Causal Islands 🏖🏝

New Environment 🛰 Friendly Neighbourhood Compute

New Environment 🛰 Friendly Neighbourhood Compute • 5G networks & Starlink 🛰 • Straight line point-to-point • PoP directly on the tower 🗼📀

New Environment 🛰 Friendly Neighbourhood Compute • 5G networks & Starlink 🛰 • Straight line point-to-point • PoP directly on the tower 🗼📀 • Walmart’s Edge • 90% Americans < 10mi [16km]

High Volume 📢 A New Topology

High Volume 📢 A New Topology 🤳

High Volume 📢 A New Topology 🛰 🤳 🗼 💾⚙

High Volume 📢 A New Topology 🛰 🤳 🗼 💾⚙ 🏢 💾⚙

High Volume 📢 A New Topology 🛰 🤳 🗼 💾⚙ 🏢 💾⚙ ☁ 💾⚙ ⚙ 💾 ⚙ 💾 ⚙ 💾 💾⚙ ⚙

High Volume 📢 A New Topology 🛰 🤳 🗼 💾⚙ 🛰 🛰 🏢 ☁ 💾⚙ 💾⚙ ⚙ 💾 ⚙ 💾 ⚙ 💾 💾⚙ ⚙

High Volume 📢 A New Topology 🛰 Local 🤳 First 🗼 💾⚙ 🛰 🛰 🏢 ☁ 💾⚙ 💾⚙ ⚙ 💾 ⚙ 💾 ⚙ 💾 💾⚙ ⚙

High Volume 📢 A New Topology 🛰 Local 🤳 First 🗼 Realtime, Storage, Caching, OLTP 💾⚙ 🛰 🛰 🏢 ☁ 💾⚙ 💾⚙ ⚙ 💾 ⚙ 💾 ⚙ 💾 💾⚙ ⚙

High Volume 📢 A New Topology 🛰 Local 🤳 First 🗼 Realtime, Storage, Caching, OLTP 💾⚙ 🛰 Relay, Replication, Consistency, Tasks 💾⚙ 🏢 🛰 ☁ 💾⚙ ⚙ 💾 ⚙ 💾 ⚙ 💾 💾⚙ ⚙

High Volume 📢 A New Topology 🛰 Local 🤳 First 🗼 Realtime, Storage, Caching, OLTP 💾⚙ 🛰 Relay, Replication, Consistency, Tasks 💾⚙ 🏢 🛰 ☁ Aggregation, Batching, Training, ⚙ 💾 ⚙ 💾 ⚙ 💾 OLAP ⚙ 💾 💾⚙ ⚙

High Volume 📢 A New Topology 🛰 Local 🤳 First 🐇 🗼 Realtime, Storage, Caching, OLTP 💾⚙ 🛰 Relay, Replication, Consistency, Tasks 💾⚙ 🏢 🛰 ☁ Aggregation, Batching, Training, ⚙ 💾 ⚙ 💾 ⚙ 💾 OLAP ⚙ 💾 💾⚙ ⚙ 🐘

Functional Programming On the Edge 🧗

Instead of […] “which database would be best to hold presences?”, we could ask “how can we best replicate data in a distributed system without the user having to worry about it?” With Elixir, you are empowered to tackle problems that in other platforms would feel impossible to solve ~ Chris McCord, What Makes Phoenix Presence Special

On the Edge 🧗 Phoenix LiveView

On the Edge 🧗 Phoenix LiveView Users 👨🏫👩🏭🧑⚕👷 Client 🖥 WSS / REST / GraphQL ↕ Controller Logic ⚙ Data Store 🗃 DevOps 📤 Developer 👩💻

On the Edge 🧗 Phoenix LiveView Users 👨🏫👩🏭🧑⚕👷 Client 🖥 WSS / REST / GraphQL ↕ Controller Logic ⚙ Data Store 🗃 DevOps 📤 Developer 👩💻 🖥 💾

On the Edge 🧗 Phoenix LiveView Users 👨🏫👩🏭🧑⚕👷 Client 🖥 💾💾💾💾 🗃 ⚙ WSS / REST / GraphQL ↕ Controller Logic ⚙ Data Store 🗃 DevOps 📤 Developer 👩💻 🖥 💾

On the Edge 🧗 Phoenix LiveView Users 👨🏫👩🏭🧑⚕👷 Client 🖥 💾💾💾💾 🗃 ⚙ ⚙ WSS / REST / GraphQL ↕ Controller Logic ⚙ Data Store 🗃 DevOps 📤 Developer 👩💻 🖥 💾

On the Edge 🧗 Phoenix LiveView Users 👨🏫👩🏭🧑⚕👷 Client 🖥 💾💾💾💾 🗃 ⚙ ⚙ WSS / REST / GraphQL ↕ Controller Logic ⚙ Data Store 🗃 DevOps 📤 Developer 👩💻 🖥 💾 🖥 💾

On the Edge 🧗 Phoenix Inside Out 🐣 🖥 💾💾 🖥 💾💾💾 ⚙ 💾 🗃

On the Edge 🧗 Mo Distributed Mo Problems Property Consequence Casual islands No single source of truth! Local first How to do coordination? Replicate/run on any machine ACLs fail (for reads, many writes)

On the Edge 🧗 CAP → PACELC 📦🦌

On the Edge 🧗 CAP → PACELC 📦🦌 • If network partition, either: • Availability (A) ✅ Uptime! • Consistency (C)

On the Edge 🧗 CAP → PACELC 📦🦌 • If network partition, either: • Availability (A) ✅ Uptime! C • Consistency (C) A

On the Edge 🧗 CAP → PACELC 📦🦌 • If network partition, either: • Availability (A) ✅ Uptime! C • Consistency (C) P A

On the Edge 🧗 CAP → PACELC 📦🦌 • If network partition, either: • Availability (A) ✅ Uptime! C • Consistency (C) P • Else (E) running normally, either: • Latency (L) ✅ Speed! • Consistency (C) A

On the Edge 🧗 CAP → PACELC 📦🦌 • If network partition, either: • Availability (A) ✅ Uptime! C • Consistency (C) P E • Else (E) running normally, either: • Latency (L) ✅ Speed! • Consistency (C) A L

On the Edge 🧗 CAP → PACELC 📦🦌 • If network partition, either: • Availability (A) ✅ Uptime! C • Consistency (C) P E • Else (E) running normally, either: • Latency (L) ✅ Speed! • Consistency (C) A L PA/EL

On the Edge 🧗 New Assumptions, New Approach

On the Edge 🧗 New Assumptions, New Approach • New assumptions → new architecture → new features • Local-first → network e cient ffi • Data can run anywhere → commons networks

On the Edge 🧗 New Tools

On the Edge 🧗 ff New Tools • Conflict-free Replicated Data Type (CRDT) • Software Transactional Memory (STM) • Di erential Datalog (DDlog) • Merkleization (hash chaining) • Cryptographic compute (MPC/FHE/NIZK) • REST < bidirectional sync < relative data views

Fixing The Leaky Pipes Identity & Access 🚿

Fixing the Leaky Pipes 🚿 ACL Read & Write

Fixing the Leaky Pipes 🚿 ACL Read & Write 🧑🌾

Fixing the Leaky Pipes 🚿 ACL Read & Write 🧑🌾 ⚙

Fixing the Leaky Pipes 🚿 ACL Read & Write 🧑🌾 💂 ✋ ⚙

Fixing the Leaky Pipes 🚿 ACL Read & Write 🧑🌾 📑 💂 ✋ ⚙

Fixing the Leaky Pipes 🚿 ACL Read & Write 🧑🌾 📑 💂 ✋ Not in control ⚙

Fixing the Leaky Pipes 🚿 ACL Read & Write 📑 In control 🧑🌾 💂 ✋ Not in control ⚙

Fixing the Leaky Pipes 🚿 ACL Read & Write 📑 💂 ✋ In control 🧑🌾 💂 ✋ Not in control ⚙

Fixing the Leaky Pipes 🚿 OCAP for Mutation

Fixing the Leaky Pipes 🚿 OCAP for Mutation 🕵

Fixing the Leaky Pipes 🚿 OCAP for Mutation 🕵 📬

Fixing the Leaky Pipes 🚿 OCAP for Mutation PID ✊ ✊ 🕵 🗺 📬

Fixing the Leaky Pipes 🚿 OCAP for Mutation PID ✊ ✊ 🕵 🗺 💌 📬

Fixing the Leaky Pipes 🚿 OCAP for Mutation 🕵 PID ✊ ✊ 🕵 🗺 💌 📬

Fixing the Leaky Pipes 🚿 OCAP for Mutation 🕵 PID ✊ ✊ 🕵 🗺 ⚙ 💌 📬

Fixing the Leaky Pipes 🚿 OCAP for Mutation 🕵 🗺 🕵 🗺 ⚙ ✊ ✊ PID ✊ ✊ Addr 💌 📬

Fixing the Leaky Pipes 🚿 OCAP for Mutation ✊ ⚙ PID 💌 📬 ✊ Addr 🎟 ✊ 🕵 🗺 🕵 🗺 ✊

Fixing the Leaky Pipes 🚿 OCAP for Mutation ✊ ⚙ PID 💌 📬 ✊ Addr 🎟 ✊ 🕵 🗺 🕵 🗺 In control ✊

Fixing the Leaky Pipes 🚿 OCAP for Mutation 🕵 🗺 🕵 🗺 ✊ ✊ PID ✊ ✊ Addr In control 🎟 ⚙ All req info 💌 📬

Fixing the Leaky Pipes 🚿 OCAP for Mutation ✊ ⚙ PID 💌 📬 ✊ Addr 🎟 ✊ 🕵 🗺 🕵 🗺 ✊

Fixing the Leaky Pipes 🚿 OCAP for Mutation ✊ PID 📬 ✊ Addr ⚙ ✊ 🕵 🎟 🗺 🎟 🎟 🕵 💌 🗺 💌 💌 ✊

Fixing the Leaky Pipes 🚿 OCAP for Mutation ✊ ⚙ PID 💌 📬 ✊ Addr 🎟 ✊ 🕵 🗺 🕵 🗺 ✊

Fixing the Leaky Pipes 🚿 OCAP for Mutation ✊ ⚙ 🧑🎨 PID 💌 📬 ✊ Addr 🎟 ✊ 🕵 🗺 🕵 🗺 ✊

Fixing the Leaky Pipes 🚿 OCAP for Mutation 🕵 🗺 🕵 🗺 ✊ ✊ PID ✊ ✊ Addr 🎟 💌 ⚙ 🧑🎨 🗺 📬

Fixing the Leaky Pipes 🚿 OCAP for Mutation 🕵 🗺 🕵 🗺 ✊ ✊ PID ✊ ✊ Addr 🎟 💌 ⚙ 🧑🎨 🗺 📬 💌

👨🎨 Fixing the Leaky Pipes 🚿 OCAP for Mutation 🕵 🗺 🕵 🗺 ✊ ✊ PID ✊ ✊ Addr 🎟 💌 ⚙ 🧑🎨 🗺 📬 💌

👨🎨 Fixing the Leaky Pipes 🚿 OCAP for Mutation 🕵 🗺 🕵 🗺 ✊ ✊ PID ✊ ✊ Addr 🎟 💌 🎟 🗺 ⚙ 🧑🎨 📬 💌

👨🎨 Fixing the Leaky Pipes 🚿 OCAP for Mutation 🕵 🗺 🕵 🗺 ✊ ✊ PID ✊ ✊ Addr 🎟 💌 🎟 🗺 🎟 ⚙ 🧑🎨 📬 💌

Fixing the Leaky Pipes 🚿 Rights Amplification

Fixing the Leaky Pipes 🚿 Rights Amplification 🥫 ✂

Fixing the Leaky Pipes 🚿 Rights Amplification 🥫 ✨ ✂

Fixing the Leaky Pipes 🚿 Rights Amplification 🥫 ✨ 🥘 ✂

Fixing the Leaky Pipes 🚿 Faster, Stronger, More Streamlined

Fixing the Leaky Pipes 🚿 Faster, Stronger, More Streamlined Universal, verifiable, user originated

Fixing the Leaky Pipes 🚿 JWT++

Fixing the Leaky Pipes 🚿 Reading the Universal Dataspace🪐🔮

Fixing the Leaky Pipes 🚿 Reading the Universal Dataspace🪐🔮 📁 📰 📁 📁 📁 📅 📁 📁 📁 📁 📁 📑 📊 📉 📑 📜 📑 📈

Fixing the Leaky Pipes 🚿 Reading the Universal Dataspace🪐🔮 👨🏫 👩🔬 📁 📁 📰 📁 📁 📅 📁 📁 📁 📁 📁 📑 📊 📉 📑 📜 📑 📈 🧑🌾

It’s all about that Data, Data, Data 📊

Data dominates. If you’ve chosen the right data structures and organized things well, the algorithms will almost always be self-evident. Data structures, not algorithms, are central to programming. Rob Pike, 5 Rules of Programming

It’s All About that Data 📊 Gossiping Out of Order 🙊

It’s All About that Data 📊 Gossiping Out of Order 🙊 🎨

It’s All About that Data 📊 Gossiping Out of Order 🙊 t→ 🎨

It’s All About that Data 📊 Gossiping Out of Order 🙊 🟢 🎨 🔴 🟡 🔴 t→ 🟢 🟢 🟡

It’s All About that Data 📊 Gossiping Out of Order 🙊 🟢 🎨 🔴 🟡 🔴 t→ 🟢 🟢 🟡 ❎

It’s All About that Data 📊 Entropy Isn’t What It Used To Be

It’s All About that Data 📊 Abstract • Network agnostic • Any number of replicas

It’s All About that Data 📊 Abstract Commutative • Network agnostic • Any number of replicas

It’s All About that Data 📊 Abstract Commutative merge merge • Network agnostic • Any number of replicas

It’s All About that Data 📊 Abstract Commutative merge merge • Network agnostic • Any number of replicas Associative

It’s All About that Data 📊 Abstract Associative Commutative merge merge • Network agnostic • Any number of replicas merge merge

It’s All About that Data 📊 Abstract Associative Commutative ( merge merge • Network agnostic • Any number of replicas ) merge merge

It’s All About that Data 📊 Abstract Associative Commutative ( merge merge • Network agnostic • Any number of replicas ) merge ( ) merge

It’s All About that Data 📊 Abstract Associative Commutative ( merge merge • Network agnostic • Any number of replicas ) merge merge ( ) merge merge

It’s All About that Data 📊 Abstract Associative Commutative ( merge merge • Network agnostic • Any number of replicas ) merge merge Idempotent ( ) merge merge

It’s All About that Data 📊 Abstract Associative Commutative ( merge merge • Network agnostic • Any number of replicas ) merge merge Idempotent ( ) merge merge fill

It’s All About that Data 📊 Abstract Associative Commutative ( merge merge • Network agnostic • Any number of replicas ) merge merge Idempotent ( ) merge merge fill fill

It’s All About that Data 📊 PNCounter

It’s All About that Data 📊 Syncing Tables Seems Inefficient