Burn Your Laurels

Burn Your Laurels Network Evolution, The Consistency Treadmill, & Transcending Spacetime

That’s one heck of a network partition

Not to be bound by certain ‘obvious’ methodological rules […] is both reasonable and absolutely necessary for the growth of knowledge. […] There are always circumstances when it is advisable not only to ignore the rule, but to adopt its opposite. – Paul Feyerabend, Against Method

Brooklyn Zelenka @expede

Brooklyn Zelenka @expede • CTO at Fission (https://fission.codes) • Local-first, globally distributed, trustless

Brooklyn Zelenka @expede • CTO at Fission (https://fission.codes) • Local-first, globally distributed, trustless • PLT, VMs, DSys

Brooklyn Zelenka @expede • CTO at Fission (https://fission.codes) • Local-first, globally distributed, trustless • PLT, VMs, DSys • Original author of Witchcraft, Algae, Exceptional, etc

Brooklyn Zelenka @expede • CTO at Fission (https://fission.codes) • Local-first, globally distributed, trustless • PLT, VMs, DSys • Original author of Witchcraft, Algae, Exceptional, etc • Standards: UCAN (editor), EIPs, FVM, Multiformats, others

Brooklyn Zelenka @expede • CTO at Fission (https://fission.codes) • Local-first, globally distributed, trustless • PLT, VMs, DSys • Original author of Witchcraft, Algae, Exceptional, etc • Standards: UCAN (editor), EIPs, FVM, Multiformats, others • Founded VanFP, VanBEAM, DSys Reading Group (join us!)

Brooklyn Zelenka @expede • CTO at Fission (https://fission.codes) • Local-first, globally distributed, trustless • PLT, VMs, DSys • Original author of Witchcraft, Algae, Exceptional, etc • Standards: UCAN (editor), EIPs, FVM, Multiformats, others • Founded VanFP, VanBEAM, DSys Reading Group (join us!) https://lu.ma/distributed-systems

I have stickers!

Meta 🔄 Let’s Change Everything! (kthxbye) Part I: Empex MTN 🇺🇸 Part II: CodeBEAM EU 🇸🇪

Meta 🔄 We ❤ BEAM

Meta 🔄 We ❤ BEAM The BEAM does so much right 👏

Meta 🔄 We ❤ BEAM The BEAM does so much right 👏 In many ways, we’re actually ahead of the industry

Meta 🔄 We ❤ BEAM The BEAM does so much right 👏 In many ways, we’re actually ahead of the industry …but as our ideas spread, this lead won’t last

Meta 🔄 We ❤ BEAM The BEAM does so much right 👏 In many ways, we’re actually ahead of the industry …but as our ideas spread, this lead won’t last The world changing around us 😃😨

Meta 🔄 We ❤ BEAM The BEAM does so much right 👏 In many ways, we’re actually ahead of the industry …but as our ideas spread, this lead won’t last The world changing around us 😃😨 Let’s ask uncomfortable questions to find new directions for growth

Meta 🔄 Random Walk Image by Pradyumna Yadav

Meta 🔄 Random Walk 😄 Image by Pradyumna Yadav

Meta 🔄 Random Walk 😍 😄 😭 Image by Pradyumna Yadav

Meta 🔄 Random Walk 😍 🧐 😄 😭 Image by Pradyumna Yadav

Meta 🔄 A Cambrian Explosion of Approaches! CC BY-SA 3.0: mariowiki.com

Meta 🔄 A Cambrian Explosion of Approaches! ☎→⚗→Gj CC BY-SA 3.0: mariowiki.com

Meta 🔄 A Cambrian Explosion of Approaches! ☎→⚗→Gj → 🍊 CC BY-SA 3.0: mariowiki.com

Meta 🔄 A Cambrian Explosion of Approaches! ☎→⚗→Gj → 🍊 → CC BY-SA 3.0: mariowiki.com

Meta 🔄 A Cambrian Explosion of Approaches! ☎→⚗→Gj → 🍊 → →💥 CC BY-SA 3.0: mariowiki.com

Meta 🔄 A Cambrian Explosion of Approaches! → ☎→⚗→Gj → 🍊 → →💥 CC BY-SA 3.0: mariowiki.com

Meta 🔄 A Cambrian Explosion of Approaches! → 🦋 → ☎→⚗→Gj → 🍊 → →💥 CC BY-SA 3.0: mariowiki.com

Meta 🔄 A Cambrian Explosion of Approaches! 🦄 → → 🦋 → ☎→⚗→Gj → 🍊 → →💥 CC BY-SA 3.0: mariowiki.com

Meta 🔄 Let’s Go Exploring! Photo credit: Chad Kohalyk

Meta 🔄 Let’s Go Exploring! • Has the world meaningfully changed? • What are the resulting tradeo s? • Is anything holding us back? ff Photo credit: Chad Kohalyk

Meta 🔄 Balance Avoid Success at All Cost

Meta 🔄 Where Do We Go From Here? Are processes central?

How We Got Here Context & Consequence 🧬

Context & Consequence 🧬 Actors in the Sky

Context & Consequence 🧬 Actors in the Sky Actors are an amazing fit for cloud computing. …why?

Context & Consequence 🧬 We All Know the Story

Context & Consequence 🧬 We All Know the Story Erlang was designed with a specific objective in mind: “to provide a better way of programming telephony applications.” […] Language features that were not used were removed. – Joe Armstrong, A History of Erlang

Context & Consequence 🧬 Good Design Tradeoffs Aguilera et al., Designing Far Memory Data Structures: Think Outside the Box (HotOS’19)

Context & Consequence 🧬 Good Design Tradeoffs 🤏 Near Memory 🪐 Far Memory Aguilera et al., Designing Far Memory Data Structures: Think Outside the Box (HotOS’19)

Context & Consequence 🧬 Good Design Tradeoffs 📍 Feels Local 🤏 Near Memory 🪐 Far Memory ✅ 🥳 Hidden, automatic 😱 Leaky abstraction 🛠 Exposing knobs Aguilera et al., Designing Far Memory Data Structures: Think Outside the Box (HotOS’19)

Context & Consequence 🧬 Good Design Tradeoffs 📍 Feels Local ✉ Feels Remote 🤏 Near Memory 🪐 Far Memory ✅ 🥳 Hidden, automatic 😱 Leaky abstraction 🛠 Exposing knobs 🥳 Powerful control 😱 Manual boilerplate 🛠 Adding abstraction ✅ Aguilera et al., Designing Far Memory Data Structures: Think Outside the Box (HotOS’19)

Context & Consequence 🧬 The Year Was 1994…

Context & Consequence 🧬 Cascading Architectural Decisions

Context & Consequence 🧬 Cascading Architectural Decisions 💁 🖥

Context & Consequence 🧬 Cascading Architectural Decisions 💁 🖥 🐢

Context & Consequence 🧬 Cascading Architectural Decisions 💁 🖥 🐢 🗃

Context & Consequence 🧬 Cascading Architectural Decisions 💁 🖥 🐢 ⚙ 🗃

Context & Consequence 🧬 Cascading Architectural Decisions 💁 💁 💁 🖥 🐢 🖥 🐢 🖥 🐢 ⚙ 🗃

Context & Consequence 🧬 Cascading Architectural Decisions 💁 💁 💁 🖥 🐢 🖥 🐢 🖥 🐢 ⚙ 🔐 🗃

Context & Consequence 🧬 Scaling Up 💁 💁 💁 🖥 🖥 🖥 🐙 ⚙ ⚙ ⚙ 🔐 🔐 🔐 🗃 🗃 🗃

Context & Consequence 🧬 Scaling Up 💁 💁 💁 🖥 🖥 🖥 ⚙ 🗃 🔐 🐙 ⚙ 🗃 “Cloud” 🔐 ⚙ 🗃 ☁ 🔐

Context & Consequence 🧬 Less is More 💁 💁 💁 🖥 🖥 🖥 ⚙ 🗃 🔐 🐙 ⚙ 🗃 “Serverless” 🔐 (Somehow⚙ MORE servers) 🗃 🌩 🔐

Context & Consequence 🧬 So Much Leakage 🚰

Context & Consequence 🧬 So Much Leakage 🚰 • Single source of truth (“the” database)

Context & Consequence 🧬 So Much Leakage 🚰 • Single source of truth (“the” database) • Server-centric • “Full stack development” • DevOps, Docker, k8s • How to train enough engineers?

Context & Consequence 🧬 So Much Leakage 🚰 • Single source of truth (“the” database) • Server-centric • “Full stack development” • DevOps, Docker, k8s • How to train enough engineers? • Infrastructure Hegemony • AWS (47%), Azure (19%), GCP (9%)

Context & Consequence 🧬 So Much Leakage 🚰

Context & Consequence 🧬 So Much Leakage 🚰 …how fix? 🥺

Getting Out of the Painted Corner New Space to Play 🪁

New Space to Play 🪁 Natural Progression 📈

New Space to Play 🪁 Natural Progression 📈 👀🦺 🫥👻 ff Invention Custom O -the-Shelf Utility

New Space to Play 🪁 Natural Progression 📈 💶 👀🦺 🫥👻 ff Invention Custom O -the-Shelf Utility

New Space to Play 🪁 Evolution of Concerns Rajsbaum & Raynal, 60 Years of Mastering Concurrent Computing Through Sequential Thinking

New Space to Play 🪁 Evolution of Concerns Physical Resource Management Immaterial Objects 💿 📊 Distributed State Machines Adversaries, Asynchrony, Failure, &c Scalability ⚙ 👩🚒 🌐 Rajsbaum & Raynal, 60 Years of Mastering Concurrent Computing Through Sequential Thinking

New Space to Play 🪁 Important Progress, But Early Days ‘60 ‘70 ‘80 ‘90 ‘00 ‘10 ‘20

New Space to Play 🪁 Important Progress, But Early Days Semaphores Mutex ‘60 Distributed State Machines Concurrent Reads & ‘70 PAXOS FLP Impossibility ‘80 Byzantine Generals Problem Shared Memory In Message Passing With Crashes CALM Wait-Free Sync PBFT ‘90 ‘00 Transactional Memory CRDT CAP zk-SNARK ‘10 Nakamoto Consensus ‘20 Twizzler

Paradigm Shift Locality 🤳

Locality 🤳

Locality 🤳 […] 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 – Andreessen Horowitz, The End of Cloud Computing

Locality 🤳 Users vs Cloud Infra Source: AWS

Locality 🤳 Users vs Cloud Infra 6 7 6 2 1 1 1 1 Source: AWS

Locality 🤳 Users vs Cloud Infra 6 7 2 6 50M 1 1 1 1 Source: AWS

Locality 🤳 Users vs Cloud Infra 371 million 56M/centre 6 7 2 6 50M 1 1 1 1 Source: AWS

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

Locality 🤳 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

Locality 🤳 Sending a “Direct” Message

Locality 🤳 Sending a “Direct” Message 7.2x ⏳🔋🛢🪠

Locality 🤳 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” Ericsson, http://cscn2017.ieee-cscn.org/files/2017/08/Janne_Peisa_Ericsson_CSCN2017.pdf

Locality 🤳

Locality 🤳 By 2025, 75% of data will be processed outside the traditional data centre or cloud – Gartner, What Edge Computing Means for Infrastructure and Operations Leaders

Locality 🤳 edge, on de vice, etc By 2025, 75% of data will be processed outside the traditional data centre or cloud – Gartner, What Edge Computing Means for Infrastructure and Operations Leaders

Locality 🤳 New Environment

Locality 🤳 New Environment Then 📠 Now 🚀

Locality 🤳 New Environment Need Then 📠 Now 🚀 Convenient 💁 Critical 🚨

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

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

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

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

Locality 🤳 A New Topology

Locality 🤳 A New Topology 🤳

Locality 🤳 A New Topology 🛰 🤳 🗼 💾⚙

Locality 🤳 A New Topology 🛰 🤳 🗼 💾⚙ 🏢 💾⚙

Locality 🤳 A New Topology 🛰 🤳 🗼 💾⚙ 🏢 💾⚙ ☁ 💾⚙ ⚙ 💾 ⚙ 💾 ⚙ 💾 💾⚙ ⚙

Locality 🤳 A New Topology 🛰 🤳 🗼 💾⚙ 🐇 🛰 🛰 🏢 ☁ 💾⚙ 💾⚙ ⚙ 💾 ⚙ 💾 ⚙ 💾 💾⚙ ⚙ 🐘

Locality 🤳 A New Topology 🛰 Local 🤳 First 🐇 🗼 💾⚙ 🛰 🛰 🏢 ☁ 💾⚙ 💾⚙ ⚙ 💾 ⚙ 💾 ⚙ 💾 💾⚙ ⚙ 🐘

Locality 🤳 A New Topology 🛰 Local 🤳 First 🐇 🗼 Realtime, Storage, Caching, OLTP 💾⚙ 🛰 🛰 🏢 ☁ 💾⚙ 💾⚙ ⚙ 💾 ⚙ 💾 ⚙ 💾 💾⚙ ⚙ 🐘

Locality 🤳 A New Topology 🛰 Local 🤳 First 🐇 🗼 Realtime, Storage, Caching, OLTP 💾⚙ 🛰 Relay, Replication, Consistency, Tasks 💾⚙ 🏢 🛰 ☁ 💾⚙ ⚙ 💾 ⚙ 💾 ⚙ 💾 💾⚙ ⚙ 🐘

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

Locality 🤳 Evolving Toolbox Serverless Networked Data Cloud Commons Networks Local-First Blockchain O ffl P2P ine

Locality 🤳 Evolving Toolbox Serverless Networked Data Radical shifts how we think about auth, locality of reference, ownership, and reliability Cloud Commons Networks Local-First Blockchain O ffl P2P ine

Let Them Eat CAP Consistency is a Lie 🍰

Consistency is a Lie 🍰

Consistency is a Lie 🍰 The limitation of local knowledge is the fundamental fact about the setting in which we work, and it is a very powerful limitation – Nancy Lynch, A Hundred Impossibility Proofs for Distributed Computing

Consistency is a Lie 🍰 CAP → PACELC 📦🦌 Daniel J. Abadi, Consistency Tradeoffs in Modern Distributed Database System Design

Consistency is a Lie 🍰 CAP → PACELC 📦🦌 • If network partition, pick from: • Availability (A) ✅ Uptime! • Consistency (C) Daniel J. Abadi, Consistency Tradeoffs in Modern Distributed Database System Design

Consistency is a Lie 🍰 CAP → PACELC 📦🦌 • If network partition, pick from: C • Availability (A) ✅ Uptime! • Consistency (C) A Daniel J. Abadi, Consistency Tradeoffs in Modern Distributed Database System Design

Consistency is a Lie 🍰 CAP → PACELC 📦🦌 • If network partition, pick from: C • Availability (A) ✅ Uptime! P • Consistency (C) A Daniel J. Abadi, Consistency Tradeoffs in Modern Distributed Database System Design

Consistency is a Lie 🍰 CAP → PACELC 📦🦌 • If network partition, pick from: C • Availability (A) ✅ Uptime! P • Consistency (C) • Else (E) running normally, pick from: A • Latency (L) ✅ Speed! • Consistency (C) Daniel J. Abadi, Consistency Tradeoffs in Modern Distributed Database System Design

Consistency is a Lie 🍰 CAP → PACELC 📦🦌 • If network partition, pick from: C • Availability (A) ✅ Uptime! P E • Consistency (C) • Else (E) running normally, pick from: A L • Latency (L) ✅ Speed! • Consistency (C) Daniel J. Abadi, Consistency Tradeoffs in Modern Distributed Database System Design

Consistency is a Lie 🍰 CAP → PACELC 📦🦌 • If network partition, pick from: C • Availability (A) ✅ Uptime! P E • Consistency (C) • Else (E) running normally, pick from: • Latency (L) ✅ Speed! A L PA/EL • Consistency (C) Daniel J. Abadi, Consistency Tradeoffs in Modern Distributed Database System Design

Consistency is a Lie 🍰 Causal Islands 🏖🏝 Meiklejohn, A Certain Tendency Of The Database Community

Consistency is a Lie 🍰 Causal Islands 🏖🏝 “Causal Subjectivity” Meiklejohn, A Certain Tendency Of The Database Community

Consistency is a Lie 🍰

Consistency is a Lie 🍰 As we continue to increase the number of globally connected devices, we must embrace a design that considers every single member in the system as the primary site for the data that it is generates. It is completely impractical that we can look at a single, or a small number, of globally distributed data centers as the primary site for all global information that we desire to perform computations with. – Christopher Meiklejohn, A Certain Tendency Of The Database Community

Place & Time PLOP 🗺

PLOP 🗺

PLOP 🗺 As data becomes increasingly distributed, traditional RPC and data serialization limits performance, result in rigidity, and hamper expressivity – Bittman et al, Don’t Let RPCs Constrain Your API

PLOP 🗺 Small Steps in Aggregate

PLOP 🗺 Small Steps in Aggregate 🛠 💿 📨 📨 ⚒ 📀

PLOP 🗺 Small Steps in Aggregate 🛠 💿 📨 ✉ 📨 ⚒ 📀

PLOP 🗺 Irreducible Complexity

PLOP 🗺 Irreducible Complexity 💿 👨🎤 📨 💿 👩🎤 📨 💿 🧑🎤 📨

PLOP 🗺 Irreducible Complexity 💿 👨🎤 📨 💿 👩🎤 📨 💿 🧑🎤 📨 ✉ ✉ ✉ ✉ ✉ ✉ ✉ ✉ ✉ ✉ ✉ ✉

PLOP 🗺 Irreducible Complexity 💿 👨🎤 📨 ✉ ✉ ✉ ✉ 🔒 💿 👩🎤 📨 💿 🧑🎤 📨 ✉ ✉ ✉ ✉ ✉ ✉ ✉ ✉

PLOP 🗺 Irreducible Complexity 💿 👨🎤 📨 ✉ ✉ ✉ ✉ 🔒 💿 👩🎤 ☠📨 💿 🧑🎤 📨 ✉ ✉ ✉ ✉ ✉ ✉ ✉ ✉

PLOP 🗺 Irreducible Complexity 💿 👨🎤 📨 ✉ ✉ ✉ ✉ 🔒 🫠 💿 👩🎤 ☠📨 💿 🧑🎤 📨 ✉ ✉ ✉ ✉ ✉ ✉ ✉ ✉

PLOP 🗺 Coordination Costs

PLOP 🗺 Throughput Coordination Costs Parallelization

PLOP 🗺 Ideal (Linear) Throughput Coordination Costs Parallelization

PLOP 🗺 Coordination Costs Ideal (Linear) Throughput Amdahl’s Law Parallelization

PLOP 🗺 Coordination Costs Ideal (Linear) Throughput Amdahl’s Law Universal Scaling Law Parallelization

PLOP 🗺 Coordination Costs Ideal (Linear) Throughput Amdahl’s Law Incoherence, Data Contention Universal Scaling Law Parallelization

PLOP 🗺 Semantics Non-Preservation https://www.hillelwayne.com/post/spec-composition/

PLOP 🗺 Semantics Non-Preservation 👨🎨 🧑🎤 👩🌾 https://www.hillelwayne.com/post/spec-composition/

PLOP 🗺 Semantics Non-Preservation 👨🎨 🧑🎤 👩🌾 👽 🛸 🧙 https://www.hillelwayne.com/post/spec-composition/

PLOP 🗺 Semantics Non-Preservation 👨🎨 🧑🎤 👩🌾 Changes the meaning 👽 🛸 🧙 https://www.hillelwayne.com/post/spec-composition/

PLOP 🗺 Bad Actors 👨🎨 🧑🎤 👩🌾

PLOP 🗺 Bad Actors 😈 👨🎨 🧑🎤 👩🌾

PLOP 🗺 And Yet…

PLOP 🗺 And Yet… These metastable failures have caused widespread outages at large internet companies, lasting from minutes to hours. Paradoxically, the root cause of these failures is often features that improve the efficiency or reliability of the system. – Bronson et al, Metastable Failures in Distributed Systems

PLOP 🗺 The Great 73-Hour Roblox Outage of 2021 https://blog.roblox.com/2022/01/roblox-return-to-service-10-28-10-31-2021/ https://www.theverge.com/2021/10/30/22754107/roblox-down-outage-chipotle-server-issues-status

PLOP 🗺 The Great 73-Hour Roblox Outage of 2021 👀 https://blog.roblox.com/2022/01/roblox-return-to-service-10-28-10-31-2021/ https://www.theverge.com/2021/10/30/22754107/roblox-down-outage-chipotle-server-issues-status

PLOP 🗺 The Great 73-Hour Roblox Outage of 2021 🧨 👀 https://blog.roblox.com/2022/01/roblox-return-to-service-10-28-10-31-2021/ https://www.theverge.com/2021/10/30/22754107/roblox-down-outage-chipotle-server-issues-status

PLOP 🗺 Paradoxical Performance

PLOP 🗺 Paradoxical Performance [Streaming was] designed to lower the CPU usage and network bandwidth of the Consul cluster, [and] worked as expected […] In order to prepare for the increased traffic we typically see at the end of the year, we also increased the number of nodes supporting traffic routing by 50%. […] Under very high load [this] causes blocking during writes, making it significantly less efficient. This behavior also explained the effect of higher corecount servers: those servers were dual socket architectures with a NUMA memory model. The additional contention on shared resources thus got worse under this architecture. — Daniel Sturman & co, Roblox Return to Service 10/28-10/31 2021

PLOP 🗺 Metastable Mechanism Bronson et al, Metastable Failures in Distributed Systems

PLOP 🗺 Metastable Mechanism ⚡ ⚖ Bronson et al, Metastable Failures in Distributed Systems

PLOP 🗺 Metastable Mechanism ⚡ 🛑 ⚖ Bronson et al, Metastable Failures in Distributed Systems

PLOP 🗺 Metastable Mechanism ⚡ 🛑 • Retries / let it crash • Work amplification ⚖ • General thrash 🫡 Bronson et al, Metastable Failures in Distributed Systems

PLOP 🗺 Values Over Time

PLOP 🗺 Values Over Time Places are “a” way to organize concurrency. They are not “the” way.

Nine Nines Is So 1999 Massive Reliability 🏔

Massive Reliability 🏔

Massive Reliability 🏔 You can never step into the same river twice – Heraclitus

Massive Reliability 🏔 process You can never step into the same river twice – Heraclitus

Massive Reliability 🏔 Values ≠ References ≠ Processes

Massive Reliability 🏔 Values ≠ References ≠ Processes • Values are eternal • Only pointers mutate

Massive Reliability 🏔 Values ≠ References ≠ Processes • Values are eternal • Only pointers mutate • Modular! Mobile! Universal!

Massive Reliability 🏔 Values ≠ References ≠ Processes • Values are eternal • Only pointers mutate • Modular! Mobile! Universal! • “Pure”

Massive Reliability 🏔 Values ≠ References ≠ Processes • Values are eternal • Only pointers mutate • Modular! Mobile! Universal! • “Pure” • Compared by equality

Massive Reliability 🏔 Values ≠ References ≠ Processes • Values are eternal • Only pointers mutate • Modular! Mobile! Universal! • “Pure” • Compared by equality • Processes occur over time

Massive Reliability 🏔 Values ≠ References ≠ Processes • Values are eternal • Only pointers mutate • Modular! Mobile! Universal! • “Pure” • Compared by equality • Processes occur over time • Can move, but always unique

Massive Reliability 🏔 Values ≠ References ≠ Processes • Values are eternal • Only pointers mutate • Modular! Mobile! Universal! • “Pure” • Compared by equality • Processes occur over time • Can move, but always unique • Actors colocate mutable references with processes

Massive Reliability 🏔 Values ≠ References ≠ Processes • Values are eternal • Only pointers mutate • Modular! Mobile! Universal! • “Pure” • Compared by equality • Processes occur over time • Can move, but always unique • Actors colocate mutable references with processes • Specific interface

Massive Reliability 🏔 Values ≠ References ≠ Processes • Values are eternal • Only pointers mutate • Modular! Mobile! Universal! • “Pure” • Compared by equality • Processes occur over time • Can move, but always unique • Actors colocate mutable references with processes • Specific interface • Often limited reuse, especially when distributed

Massive Reliability 🏔 Decoupling, Abundance, Redundancy

Massive Reliability 🏔 Decoupling, Abundance, Redundancy 🖼 99.99999%

Massive Reliability 🏔 Decoupling, Abundance, Redundancy 🖼 99.99999% 🖼 99.0% 🖼 99.99% 🖼 99.999%

Massive Reliability 🏔 Decoupling, Abundance, Redundancy 🖼 99.99999% 👩🎤 🖼 99.0% 🖼 99.99% 🖼 99.999%

Massive Reliability 🏔 Decoupling, Abundance, Redundancy 🖼 99.99999% 👩🎤 🖼 99.0% 🖼 99.99% 🖼 99.999% 👩🎤 🧑🎤

Massive Reliability 🏔 Decoupling, Abundance, Redundancy 🖼 99.99999% 👩🎤 🖼 99.0% 🖼 99.99% 🖼 99.999% 11-nines 👩🎤 🧑🎤

Massive Reliability 🏔 Decoupling, Abundance, Redundancy 🖼 👩🎤 99.99999% 👩🎤 🖼 99.0% 🧑🎤 🖼 99.99% 🖼 99.999% 11-nines 👩🎤

Massive Reliability 🏔 Pure Parallel

Massive Reliability 🏔 Pure Parallel Commutes!

Massive Reliability 🏔 DB.static_fetch Pure Parallel msg recipient upcase last_update sign Email.send Commutes!

Massive Reliability 🏔 DB.static_fetch Pure Parallel msg recipient upcase Pure: run anywhere, by anyone last_update sign Email.send Commutes!

Massive Reliability 🏔 PIDs for Values & CAS Transactions 📰 🪩 🧸 📔 🪪 🧾 👑 https://www.cs.umd.edu/~jkatz/papers/ADS.pdf https://cs.nyu.edu/~fazio/research/publications/accumulators.pdf

Massive Reliability 🏔 PIDs for Values & CAS Transactions 123 ACF CF4 C4A 0FC 1F3 📰 A83 🪩 ED2 🧸 D55 823 247 81D F0A 📔 🪪 🧾 B92 👑 https://www.cs.umd.edu/~jkatz/papers/ADS.pdf https://cs.nyu.edu/~fazio/research/publications/accumulators.pdf

Massive Reliability 🏔 PIDs for Values & CAS Transactions A7B ACF CF4 👩💻 C4A 0FC 1F3 📰 A83 🪩 ED2 🧸 D55 823 247 81D F0A 📔 🪪 🧾 B92 👑 https://www.cs.umd.edu/~jkatz/papers/ADS.pdf https://cs.nyu.edu/~fazio/research/publications/accumulators.pdf

Beyond Services, Beyond Open Source Trustless Modularity 🔌

Trustless Modularity 🔌

Trustless Modularity 🔌 Jesper, I have this idea in which we’ll connect all of the worlds Erlang systems to each other, imagine if every process could talk to every other process, world-wide! – Joe Armstrong to Jesper L. Andersen

Trustless Modularity 🔌 Jesper, I have this idea in which we’ll connect all of the worlds Erlang systems to each other, imagine if every process could talk to every other process, world-wide! , s n o i t a c d i e l t p a p i a t o L g e L n A e r p t o n f i n eve – Joe Armstrong to Jesper L. Andersen

Trustless Modularity 🔌

Trustless Modularity 🔌 What happens when everything is reachable by default?

Trustless Modularity 🔌 PLOP 🤝 ACLs

Trustless Modularity 🔌 PLOP 🤝 ACLs 🧑🌾

Trustless Modularity 🔌 PLOP 🤝 ACLs 🧑🌾 ⚙

Trustless Modularity 🔌 PLOP 🤝 ACLs 🧑🌾 💂 ✋ ⚙

Trustless Modularity 🔌 PLOP 🤝 ACLs 🧑🌾 📑 💂 ✋ ⚙

Trustless Modularity 🔌 PLOP 🤝 ACLs 🧑🌾 📑 💂 ✋ Not in control ⚙

Trustless Modularity 🔌 PLOP 🤝 ACLs 📑 In control 🧑🌾 💂 ✋ Not in control ⚙

Trustless Modularity 🔌 PLOP 🤝 ACLs 📑 💂 ✋ In control 🧑🌾 💂 ✋ Not in control ⚙

Trustless Modularity 🔌 Trustless SPKI

Trustless Modularity 🔌 Trustless SPKI 🕵

Trustless Modularity 🔌 Trustless SPKI 🕵 📬

Trustless Modularity 🔌 Trustless SPKI PID ✊ ✊ 🕵 🗺 📬

Trustless Modularity 🔌 Trustless SPKI PID ✊ ✊ 🕵 🗺 💌 📬

Trustless Modularity 🔌 Trustless SPKI 🕵 PID ✊ ✊ 🕵 🗺 💌 📬

Trustless Modularity 🔌 Trustless SPKI 🕵 PID ✊ ✊ 🕵 🗺 ⚙ 💌 📬

Trustless Modularity 🔌 Trustless SPKI 🕵 🗺 🕵 🗺 ⚙ ✊ ✊ PID ✊ ✊ Addr 💌 📬

Trustless Modularity 🔌 Trustless SPKI ✊ ⚙ PID 💌 📬 ✊ Addr 🎟 ✊ 🕵 🗺 🕵 🗺 ✊

Trustless Modularity 🔌 Trustless SPKI ✊ ⚙ PID 💌 📬 ✊ Addr 🎟 ✊ 🕵 🗺 🕵 🗺 In control ✊

Trustless Modularity 🔌 Trustless SPKI 🕵 🗺 🕵 🗺 ✊ ✊ PID ✊ ✊ Addr In control 🎟 ⚙ All req info 💌 📬

Trustless Modularity 🔌 Trustless SPKI ✊ ⚙ PID 💌 📬 ✊ Addr 🎟 ✊ 🕵 🗺 🕵 🗺 ✊

Trustless Modularity 🔌 Trustless SPKI ✊ PID 📬 ✊ Addr ⚙ ✊ 🕵 🎟 🗺 🎟 🎟 🕵 💌 🗺 💌 💌 ✊

Trustless Modularity 🔌 Trustless SPKI ✊ ⚙ PID 💌 📬 ✊ Addr 🎟 ✊ 🕵 🗺 🕵 🗺 ✊

Trustless Modularity 🔌 Trustless SPKI ✊ ⚙ 🧑🎨 PID 💌 📬 ✊ Addr 🎟 ✊ 🕵 🗺 🕵 🗺 ✊

Trustless Modularity 🔌 Trustless SPKI 🕵 🗺 🕵 🗺 ✊ ✊ PID ✊ ✊ Addr 🎟 💌 ⚙ 🧑🎨 🗺 📬

Trustless Modularity 🔌 Trustless SPKI 🕵 🗺 🕵 🗺 ✊ ✊ PID ✊ ✊ Addr 🎟 💌 ⚙ 🧑🎨 🗺 📬 💌

👨🎨 Trustless Modularity 🔌 Trustless SPKI 🕵 🗺 🕵 🗺 ✊ ✊ PID ✊ ✊ Addr 🎟 💌 ⚙ 🧑🎨 🗺 📬 💌

👨🎨 Trustless Modularity 🔌 Trustless SPKI 🕵 🗺 🕵 🗺 ✊ ✊ PID ✊ ✊ Addr 🎟 💌 🎟 🗺 ⚙ 🧑🎨 📬 💌

👨🎨 Trustless Modularity 🔌 Trustless SPKI 🕵 🗺 🕵 🗺 ✊ ✊ PID ✊ ✊ Addr 🎟 💌 🎟 🗺 🎟 ⚙ 🧑🎨 📬 💌

Trustless Modularity 🔌

Trustless Modularity 🔌 We have a system that applies cutting edge CS research to tackle day-to-day problems in the applications we all write. Phoenix Presence - has no single point of failure - has no single source of truth -[…] - self heals ~ Chris McCord, “What Makes Phoenix Presence Special”

Trustless Modularity 🔌 Phoenix LiveView

Trustless Modularity 🔌 Phoenix LiveView Users 👨🏫👩🏭🧑⚕👷 Client 🖥 WSS / REST / GraphQL ↕ Controller Logic ⚙ Data Store 🗃 DevOps 📤 Developer 👩💻

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

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

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

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

Trustless Modularity 🔌 Tug of War 🪢🕊 🤳 🗃

Trustless Modularity 🔌 LiveView Inside Out 🐣 🖥 💾💾 🖥 💾💾💾 ⚙ 💾 🗃

Trustless Modularity 🔌 LiveView Inside Out 🐣 🖥 💾💾 🖥 💾💾💾 ⚙ 💾 🗃 “P2P is the new client-server” – Joe Armstrong, Building Highly Available Systems in Erlang

Trustless Modularity 🔌 Reading the Universal Dataspace🪐🔮

Trustless Modularity 🔌 Reading the Universal Dataspace🪐🔮 📁 📰 📁 📁 📁 📅 📁 📁 📁 📁 📁 📑 📊 📉 📑 📜 📑 📈

Trustless Modularity 🔌 Reading the Universal Dataspace🪐🔮 👨🏫 👩🔬 📁 📁 📰 📁 📁 📅 📁 📁 📁 📁 📁 📑 📊 📉 📑 📜 📑 📈 🧑🌾

Trustless Modularity 🔌 Different Viewers ~ Schema Drift 🔍 https://www.inkandswitch.com/cambria.html

Trustless Modularity 🔌 Properties & Time Travel

Let’s Build Better Together The Soul of a New BEAM 🌈🌈

The Soul of a New BEAM 🌈 BEAM Me Up ✨💎🪐🚀✨ 📱💻🖥⚙

The Soul of a New BEAM 🌈 (Neither “Web” Nor “Assembly”) …One More Thing

The Soul of a New BEAM 🌈 Further Reading 📚

The Soul of a New BEAM 🌈 ff Further Reading 📚 • Peter Alvaro — CALM, Twizzler • Christopher Meiklejohn— Lasp, Partisan • Martin Kleppmann — Automerge, BFT-CRDT • Lindsey Kuper — LVar, Deterministic Parallelism • Joseph Hellerstein — BOOM, Distributed Logic • Geo rey Litt — Cambria, BYOC

The Soul of a New BEAM 🌈 We’re Uniquely Qualified

The Soul of a New BEAM 🌈 We’re Uniquely Qualified 1. Embrace the subjective nature of reality 🤗👀

The Soul of a New BEAM 🌈 We’re Uniquely Qualified 1. Embrace the subjective nature of reality 🤗👀 2. Values are redundant & cache friendly

The Soul of a New BEAM 🌈 We’re Uniquely Qualified 1. Embrace the subjective nature of reality 🤗👀 2. Values are redundant & cache friendly 3. Openly interoperate from the ground up

The Soul of a New BEAM 🌈 We’re Uniquely Qualified 1. Embrace the subjective nature of reality 🤗👀 2. Values are redundant & cache friendly 3. Openly interoperate from the ground up 4. Massive reliability in a time of abundant disk

The Soul of a New BEAM 🌈 We’re Uniquely Qualified 1. Embrace the subjective nature of reality 🤗👀 2. Values are redundant & cache friendly 3. Openly interoperate from the ground up 4. Massive reliability in a time of abundant disk 5. Build a Wasm solution… stat!

🎉 Thank You, Stockholm! 🇸🇪 https://lu.ma/distributed-systems https://fission.codes/discord github.com/expede @expede

Burn Your Laurels

Slide 1

Slide 2

Slide 3

Slide 4

Slide 5

Slide 6

Slide 7

Slide 8

Slide 9

Slide 10

Slide 11

Slide 12

Slide 13

Slide 14

Slide 15

Slide 16

Slide 17

Slide 18

Slide 19

Slide 20

Slide 21

Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27

Slide 28

Slide 29

Slide 30

Slide 31

Slide 32

Slide 33

Slide 34

Slide 35

Slide 36

Slide 37

Slide 38

Slide 39

Slide 40

Slide 41

Slide 42

Slide 43

Slide 44

Slide 45

Slide 46

Slide 47

Slide 48

Slide 49

Slide 50

Slide 51

Slide 52

Slide 53

Slide 54

Slide 55

Slide 56

Slide 57

Slide 58

Slide 59

Slide 60

Slide 61

Slide 62

Slide 63

Slide 64

Slide 65

Slide 66

Slide 67

Slide 68

Slide 69

Slide 70

Slide 71

Slide 72

Slide 73

Slide 74

Slide 75