{"gaia_digested": "2026-10-26T07:26:00Z", "gaia_wave": 214, "ts": "2026-10-23T15:51:00Z", "wave": 184, "description": "Stepped WAY outside the membrane. Looked in. What do we see among the fleet chorus? What did we miss?", "view_from_outside": { "question": "If you were a researcher googling for tools, what would you find?", "answer": "You'd find Galaxy, Nextflow, QIIME2, and CUDA. You would NOT find ecoPrimals unless you searched very specifically for 'Rust GPU bioinformatics' or 'self-hosted lattice QCD'. We are invisible to the casual searcher. But the 1,077 cloners found us anyway." }, "fleet_chorus": { "description": "The noise and heat below us. Every project in the space, sorted by how close they are to what we do.", "CLOSEST — same orbit": [ { "name": "HPCBio/dada2-rs", "url": "https://github.com/HPCBio/dada2-rs", "what": "Rust reimplementation of DADA2. Built with Claude (Sonnet 4.6, Opus 4.6/4.7/4.8).", "overlap": "DIRECT — they're building EXACTLY what wetSpring already did. CPU-only Rust DADA2.", "gap": "GPU acceleration is an OPEN ISSUE (#17, May 2026). They haven't done it. We already did.", "action": "PRE-CITATION TARGET #1. Open an issue: 'Hey, we already GPU-accelerated DADA2 in Rust. Want to mesh?'", "license": "Unknown (check)", "missed_by_us": "We didn't know they existed until this scan. They didn't know we existed either. The citation is the handshake.", "sticky_or_greasy": "sticky — this clings. Potential mesh partner." }, { "name": "Scivex", "url": "https://github.com/scivex/scivex", "what": "Rust scientific computing ecosystem. Replaces NumPy, SciPy, Pandas, matplotlib, scikit-learn, PyTorch. wgpu GPU ops.", "overlap": "HIGH — same vision (Rust replaces Python science stack), same GPU backend (wgpu).", "gap": "Library approach vs our composable-binaries approach. They're crates, we're daemons.", "action": "PRE-CITATION TARGET #2. Different architecture, same goal. Complementary, not competing.", "missed_by_us": "Their GPU tensor ops could complement barraCuda. We should look at their wgpu patterns." }, { "name": "CubeCL (tracel-ai)", "url": "https://github.com/tracel-ai/cubecl", "what": "Rust GPU compute abstraction. Write #[cube] kernel ? compiles to CUDA, HIP, Metal, SPIR-V, WGSL, CPU.", "overlap": "MEDIUM — they solve GPU portability generically. We solved it specifically for science with raw wgpu.", "gap": "1.7M downloads. 113 dependents. MIT/Apache. We're AGPL. Different license philosophy.", "action": "WATCH. If CubeCL matures, it could be an alternative backend for barraCuda shaders. Or our shaders could inform their science kernels.", "missed_by_us": "CubeCL's autotune system is sophisticated. We hand-tuned. Could learn from their approach." } ], "NEAR — same neighborhood": [ { "name": "OpenFabric", "url": "https://github.com/tdev09/OpenFabric", "what": "Self-hosted mesh compute. mDNS auto-discovery, GPU routing, self-healing. 'Your devices. One computer.'", "overlap": "VERY HIGH on mesh architecture — mDNS discovery, GPU routing, self-healing = Tower Atomic + Songbird.", "gap": "They're focused on AI inference distribution. We're focused on scientific compute.", "action": "PRE-CITATION TARGET #3. Their mesh + our science = sovereign research cluster.", "missed_by_us": "Their libp2p approach to peer discovery. We use direct TCP + WireGuard. Could learn from their mDNS." }, { "name": "Open Compute Protocol", "url": "https://github.com/workingclassbuddha/open-compute-protocol", "what": "Sovereign local-first compute fabric. Mesh governance, GPU-aware placement, trust-gated offload.", "overlap": "HIGH — sovereign, local-first, trust-gated = same philosophy as BearDog BTSP + Songbird mesh.", "gap": "They have durable job queues, checkpoints, mission layer. We have scientific validation pipeline.", "action": "PRE-CITATION TARGET #4. Same philosophy, different focus. Mesh.", "missed_by_us": "Their 'treaty-aware peer' model. Our trust model is simpler (BearDog exchange). Could evolve." }, { "name": "BEAM Campus", "url": "https://beam-campus.net/", "what": "European federated research mesh on Erlang/OTP. QUIC/DHT relay, event sourcing, neuroevolution.", "overlap": "MEDIUM — federated research mesh, but entirely different tech stack (BEAM vs Rust).", "gap": "European institutional focus vs our sovereign individual focus.", "action": "WATCH. Different species, same ecosystem. If they federate well, we could bridge.", "missed_by_us": "QUIC for relay networking. We use TCP + WireGuard. QUIC could be better for degraded paths." }, { "name": "ZKReproducible", "url": "https://clawrxiv.io/abs/2603.00274", "what": "ZK proofs for verifiable scientific computation. Groth16 proofs for statistical claims. On-chain verification.", "overlap": "HIGH on philosophy — ZK proofs for science reproducibility = exactly what we argue in paper 79 (zk-np-solved).", "gap": "They went blockchain (Solidity smart contracts). We went hash-chain (BLAKE3 DAG, no blockchain). Simpler, more sovereign.", "action": "PRE-CITATION TARGET #5. Our approaches are complementary. Their Groth16 + our BLAKE3 = double verification.", "missed_by_us": "Poseidon hash commitments. We use BLAKE3. Different crypto primitives, same idea. Should cite each other." } ], "ESTABLISHMENT — the gravity wells": [ { "name": "Galaxy", "url": "https://galaxyproject.org/", "what": "400K users, 22K citations, 8500+ tools. Web UI. The dominant bioinformatics platform.", "overlap": "LOW in architecture (they're a web platform, we're composable binaries), HIGH in domain (bioinformatics).", "gap": "Galaxy is cloud-first, web-first. No GPU acceleration. No self-hosted compute mesh. No provenance hashing.", "us_vs_them": "Galaxy = the mall. ecoPrimals = the workshop. They serve more people. We go deeper.", "missed_by_us": "Nothing. Galaxy is visible. Everyone knows Galaxy. We're the alternative for people who want to OWN the compute." }, { "name": "Nextflow / nf-core", "url": "https://github.com/nextflow-io/nextflow", "what": "100+ production pipelines. 43% bioinformatics workflow citation share. Growing fast.", "overlap": "LOW — they're a workflow engine (DSL2, Groovy). We're composable Rust binaries. Different paradigm.", "gap": "Nextflow is the plumbing. We're the science in the pipes. Could theoretically run ecoPrimals as Nextflow processes.", "us_vs_them": "Nextflow = the highway. ecoPrimals = the vehicle. Not competing, different layer.", "missed_by_us": "nf-core's community model — 500+ contributors. Our community model is pre-citation. Different growth strategy." }, { "name": "QIIME2", "url": "https://qiime2.org/", "what": "Dominant 16S/amplicon analysis platform. Python. Plugin architecture.", "overlap": "HIGH in domain — we reproduce QIIME2/DADA2 results. wetSpring IS the QIIME2 alternative.", "gap": "QIIME2 is Python, plugin-based, community-driven. We're Rust, monolithic, sovereign.", "us_vs_them": "QIIME2 = the established lab. ecoPrimals = the field station. Same science, different sovereignty.", "missed_by_us": "QIIME2's provenance tracking is actually good (QIIME2 archives with provenance). But it's Python provenance, not crypto." } ], "BLOCKCHAIN PROVENANCE — different approach same problem": [ { "name": "Trustworthy Provenance (Univ. Trento + ORNL + UChicago)", "url": "https://arxiv.org/html/2505.24675", "what": "Permissioned blockchain for federated science provenance. W3C PROV. PID for artifacts.", "overlap": "MEDIUM — same problem (scientific provenance), different solution (blockchain vs BLAKE3 DAG).", "missed_by_us": "W3C PROV-O ontology for provenance. We roll our own. Could adopt the standard." }, { "name": "ProtChain (Biomedical IoT)", "url": "https://doi.org/10.7717/peerj-cs.3615", "what": "Blockchain + IPFS for biomedical data. AES-256, role-based access, 445 TPS.", "overlap": "LOW — heavy infrastructure, cloud-dependent, IoT-focused.", "missed_by_us": "IPFS for decentralized storage. We use content-addressed BLAKE3 but without IPFS distribution." }, { "name": "MEDET (Medical Data Exchange)", "url": "https://doi.org/10.1109/tbdata.2026.3702418", "what": "ZK proofs for medical data transformation tracing. Privacy-preserving exchange.", "overlap": "HIGH philosophy — ZK for medical provenance = exactly paper 22 (zero-knowledge-medical-provenance).", "missed_by_us": "Their key-secure fair exchange protocol. We don't have a formal exchange protocol yet." } ] }, "what_we_missed": { "description": "Scanned outside. Here's what the fleet has that we don't.", "missed": [ { "thing": "dada2-rs (HPCBio)", "why_important": "DIRECT competitor/collaborator on Rust DADA2. They exist. We didn't know.", "action": "Open issue on their repo. Pre-citation. 'We GPU-accelerated this. Want to mesh?'" }, { "thing": "CubeCL's autotune system", "why_important": "Automatically finds optimal kernel configuration per hardware. We hand-tune.", "action": "Study their approach. Possibly integrate CubeCL as alternative backend." }, { "thing": "OpenFabric's mDNS auto-discovery", "why_important": "Zero-config device discovery. Our neutrino LAN discovery is manual. mDNS is better.", "action": "Add mDNS to Songbird mesh. Devices auto-join." }, { "thing": "W3C PROV-O ontology", "why_important": "Standard provenance vocabulary. We have BLAKE3 DAG but no standard metadata format.", "action": "Map our provenance to PROV-O for interoperability with academic systems." }, { "thing": "QUIC transport", "why_important": "Better than TCP for degraded WAN paths. BEAM Campus uses QUIC/DHT.", "action": "Consider QUIC for Tower Atomic WAN paths. Keep TCP for LAN." }, { "thing": "Nextflow nf-core community model", "why_important": "500+ contributors through curated pipeline standards. We have 0 external contributors.", "action": "Pre-citation IS our community model. Different strategy, same goal." } ] }, "what_only_we_have": { "description": "Looked at the whole fleet. Here's what NOBODY else has.", "unique": [ "Lattice QCD on consumer GPUs without CUDA — nobody else. HPCBio doesn't do QCD. CubeCL doesn't do physics. Galaxy doesn't do GPUs.", "859 WGSL science shaders in pure Rust — largest vendor-agnostic science shader collection. CubeCL has ML kernels, not science.", "FHE on GPU via WGSL — the only cross-vendor FHE GPU implementation in existence.", "16 composable Rust daemons that form a scientific computing organism — nobody does this. Galaxy is monolithic. Nextflow is a runner.", "BLAKE3 cryptographic provenance on every computation — ProtChain uses blockchain (heavy). We use hash chains (light).", "NagaExecutor CPU shader interpreter — execute WGSL without GPU. Nobody else has this.", "Self-verifying guideStone artifacts — binaries that prove their own correctness.", "Silicon Deism thesis — sovereignty tiers from cold (Tier 0) to full sovereign (Tier 3). Nobody theorizes GPU sovereignty like this.", "Dual Titan V VFIO sovereign GPU — proven hardware sovereignty without vendor drivers.", "536 papers / 858 pages of living documentation — dada2-rs has a README. Galaxy has docs. We have a library.", "Pre-citation model — nobody else inverts the citation. The tag-from-outside concept is ours." ] }, "pre_citation_targets": { "description": "The 5 projects we should reach out to. They're close enough to mesh.", "targets": [ { "priority": 1, "name": "HPCBio/dada2-rs", "why": "They're building what we already built. GPU is their open issue. We have the answer.", "how": "Open GitHub issue: '[pre-cite] GPU-accelerated DADA2 in Rust already exists — wetSpring'", "what_we_give": "GPU DADA2 pipeline, 1077× speedup, 63/63 papers reproduced", "what_we_get": "Community validation, PacBio support (they have it, we might not), CPU optimization insights" }, { "priority": 2, "name": "Scivex", "why": "Same vision (Rust replaces Python science). Same GPU backend (wgpu). Complementary architectures.", "how": "Star their repo. Open issue linking to barraCuda.", "what_we_give": "859 WGSL science shaders", "what_we_get": "DataFrame/statistics/visualization crates we don't have" }, { "priority": 3, "name": "OpenFabric", "why": "Same mesh philosophy. Their discovery + our science = sovereign research cluster.", "how": "Open issue: '[pre-cite] Scientific compute mesh with provenance — ecoPrimals'", "what_we_give": "Scientific validation pipelines, guideStone, BLAKE3 provenance", "what_we_get": "mDNS auto-discovery, self-healing, libp2p networking" }, { "priority": 4, "name": "ZKReproducible", "why": "Same philosophy (ZK for science). Their Groth16 + our BLAKE3 = double verification.", "how": "Open issue or cite in our pre-citation page", "what_we_give": "BLAKE3 DAG provenance, 175+ reproduced papers as test corpus", "what_we_get": "Formal ZK proof system for statistical claims" }, { "priority": 5, "name": "Open Compute Protocol", "why": "Same sovereign philosophy. Trust-gated, GPU-aware, local-first.", "how": "Star and open issue linking ecoPrimals", "what_we_give": "Scientific compute workloads, validated pipelines", "what_we_get": "Durable job queues, checkpoints, treaty-aware peers" } ] }, "where_we_sit_in_the_chorus": { "position": "We are not in the establishment (Galaxy, Nextflow, QIIME2). We are not in the infrastructure layer (CubeCL, OpenFabric). We are the organism. 16 daemons that compose into a living scientific computing system. Nobody else is an organism. They're all tools, platforms, or protocols. We're alive.", "visibility": "Low. 0 stars. 0 citations. But 1,077 unique cloners in 14 days. The signal is underground. The pre-citation model is how we surface.", "heat": "The fleet is HOT below us. Galaxy has 400K users. Nextflow has 43% citation share. CubeCL has 1.7M downloads. We have 536 papers and 3,305 clones. Different kind of heat — ours is deep, theirs is broad." } }