research: apply to the Swarm Grants Programme (funding for the Swarm integration)

[ibeezhan]

What

Research + prepare an application to the Swarm Grants Programme to fund the Swarm integration (#2).

Programme facts (as researched, 2026)

• Run by the Swarm Association; funds projects advancing decentralized tech / Web3, explicitly including developers building storage integrations.
• Total pool: 300,000 BZZ; grants paid in BZZ (DAI-equivalent valuation).
• Three tiers:
  • Starter Microgrants — $500–$2,000: simple storage integrations / PoC, aimed at first-time applicants.
  • Builder Grants — $2,000–$5,000: projects with a functional prototype and defined roadmap.
  • Booster Grants — up to $10,000: exceptional projects or ones endorsed by the Swarm Association.
• Current wave: March 21 → December 21 (rolling within the wave).
• Apply via: https://swarm.odoo.com/swarm-grants-application-form
• Priority areas they list: data-availability tooling, Web3 social dapps, Web2-to-Web3 migration, developer tooling, AI-powered decentralization, FairOS integration.

Why p2present is a strong fit

• Builder tier (not starter): p2present is already a functional, shipped product with four working persistence providers (Arweave, IPFS, WebTorrent, S3) and a pluggable provider registry — the Swarm integration (feat: Swarm (bzz://) persistence provider + resolver #2) has a concrete, scoped design.
• Web2→Web3 migration story: conference talks and slides currently rot on centralized platforms; p2present moves them to permanent decentralized hosting with a zero-dependency static player anyone can fork — exactly the migration narrative they prioritise.
• Real usage surface for Swarm: every hosted presentation = deck + video + subtitles pushed through postage-stamped uploads; the manifest resolver gives Swarm equal first-class billing with Arweave/IPFS in a real dapp, not a demo.

Application checklist

• Confirm current wave is still open + re-check tier amounts on the form
• Draft proposal: scope = feat: Swarm (bzz://) persistence provider + resolver #2 (provider + resolver + docs + gateway fallbacks), roadmap + milestones (integration → Beeport/no-node path → feeds/ENS for updatable manifests)
• Decide ask amount (Builder tier, ~$3–5k, milestone-split)
• Prepare supporting material: live demo (p2present.com), repo, HOSTING.md, provider-registry architecture notes
• Possible stretch scope to strengthen the ask: Swarm feeds for mutable manifest references (updatable presentations at a stable bzz address) — a feature no other provider in p2present has
• Submit via the form; track response

Sources

• https://www.ethswarm.org/grants/swarm-grants-programme
• https://swarm.odoo.com/swarm-grants-application-form
• https://docs.ethswarm.org/

[ibeezhan]

📄 Grant proposal (v1, July 2026)

Status: the Swarm Grants Programme is currently between waves (applications closed as of July 2026; form and portal offline; last recipient announcements 2021, last programme posts 2023). This proposal is submission-ready for the next wave and is being shared with the Swarm Foundation via direct outreach (info@swarm.foundation) in the meantime.

---

Swarm Builder Grant Application: p2present

Prepared July 2026. Ready to submit when a grant wave opens, or via direct outreach.

---

Project name

p2present: presentations that play themselves, on storage that can't rot.

One-line pitch

Make Swarm a first-class home for complete, permanently preservable presentations. Slides, talk video, captions, and sync timing travel as one self-contained reference that outlives every platform it was ever posted to.

Requested amount

$10,000 USD in BZZ for the full scope (Booster consideration, top of programme cap).
The core integration stands on its own as a $5,000 Builder Grant (functional prototype + defined roadmap); the full scope adds a security review and a community preservation drive that turns the integration into adoption. Both scopes are broken out in the budget below; fund whichever fits the wave.

---

What p2present is (already shipped, MIT-licensed)

p2present is a zero-dependency, no-build, 100% static web app that plays presentation slides in perfect sync with the talk video: scrub either one, the other follows. It is live at p2present.com, MIT-licensed at github.com/shayanb/p2present.

What makes it relevant to Swarm is how it loads content. Everything (the manifest, the deck, the video, the subtitles) resolves through a pluggable decentralized-source layer:

• Four working persistence providers behind one registry: Arweave (ar://), IPFS pinning (ipfs://), WebTorrent (magnet:), S3/https. Each turns a file into a manifest reference, and tokens never leave the browser.
• A transport resolver that tries gateways in order with graceful fallback, per scheme.
• A visual Builder (paste video + slides, types auto-detected, timing captured, done) and a Host helper that uploads assets through any registered provider.
• Signed manifests (EIP-191 via Ethereum wallet with ENS badges, or Ed25519): provenance without gatekeeping.
• Deck adapters for HTML, PDF (pdf.js), and embedded decks; multi-language subtitles (including RTL); an optional self-hostable sharing service (Cloudflare Worker + KV) with per-talk OpenGraph cards.

The architecture was explicitly designed so that adding a storage backend is one module + one registry line (design issue #2 specifies the Swarm integration in detail).

Why this matters: the preservation philosophy

Conference talks are among the most information-dense cultural artifacts the technical community produces, and among the worst preserved. The slides end up on one dying platform, the video on another, the captions nowhere; the relationship between them (what slide was on screen when the speaker said what) is lost the moment the event website goes down. A recorded talk without its deck is half a talk; a deck without the talk is a pile of bullet points.

p2present's answer is the manifest: one small JSON file that binds deck + video + captions + timing into a complete, replayable presentation. Put the manifest and its assets on permanent storage and the entire experience survives, not just the files: scrubbing, slide sync, multilingual captions, provenance signature, everything.

This is a Fair Data Society story with real stakes. Our flagship demo is a Dappcon talk about building censorship-circumvention infrastructure for Iran (moav.sh); the same pipeline is being used to preserve a decade of Farsi-language crypto-education content, material whose original platforms are genuinely at risk of takedown, region-blocking, or decay. For these communities, "decentralized storage" is not ideology; it's the difference between knowledge existing and not.

Fit with Swarm's grant priorities

Swarm priority	What this project delivers
Web2→Web3 migration	A concrete, relatable content type (conference talks, lectures, courses) moved from YouTube/Google-Slides dependence onto Swarm, with a Builder that makes migration a paste-two-links task for non-technical speakers.
Developer tooling	The persistence-provider + resolver patterns are dependency-free, framework-free reference implementations (~200 lines each) that any static dapp can lift to add Swarm support. Documented as such.
Feeds showcase (roadmap)	Swarm feeds enable something none of our other four backends can do: an updatable presentation at a stable address. Fix a slide typo, keep the same shared link. This makes Swarm demonstrably differentiated inside the app, not just "a fifth option."
Streaming/media ecosystem	Talk video streamed from Swarm via gateway range requests into an HTML5 player, complementing ecosystem work on media (Etherna, Solar Punk's streaming efforts) with a working, minimal, embeddable reference.
Adoption	Every preserved talk is real content on the network, uploaded with real postage stamps: usage, not a demo.

Milestone (single milestone, per grant guidelines)

"Swarm-native p2present": integration shipped + real talks preserved end-to-end on Swarm. 3 months for the core scope, 4 months for the full scope.

Core deliverables:

1. bzz:// resolver: bzz://<reference>[/path] accepted anywhere a source can appear (manifest, video, deck, subtitles); ordered gateway fallback (bzz.link subdomain + path-style), overridable via resolvers.swarmGateways in the manifest; JSON-Schema + spec updates.
2. Swarm persistence provider: in the Host helper's registry alongside Arweave/IPFS/Torrent/S3. Point at a Bee node or gateway endpoint + postage batch id, upload via POST /bzz, get a bzz:// manifest reference back. Credentials stay in-browser (same trust model as our other providers). Beeport documented as the no-node path.
3. Flagship preserved talk: a complete real presentation (23-slide PDF deck + talk video + EN/FA subtitles + signed manifest) hosted entirely on Swarm, linked from p2present.com as the reference example. One bzz:// link that replays the whole talk.
4. Documentation + tests: a HOSTING.md chapter (run Bee, buy a postage batch, use a gateway), an AUTHORING walkthrough ("preserve your talk on Swarm" tutorial), and headless smoke-test coverage of the bzz resolver in our existing 124-check suite.
5. Demo day presentation, delivered, fittingly, as a p2present presentation served from Swarm.

Full-scope additions (see budget): a published security review of the integration and provider layer, a curated "talks worth keeping" collection preserved on Swarm, and an open-source announcement bot that gives every newly preserved talk ongoing public visibility with its permanent bzz:// link.

Monthly progress reports per programme rules (results by the 1st of each month).

Roadmap beyond this milestone (separate future applications)

• Swarm feeds for mutable manifests: updatable presentations at stable addresses; feed-aware share links; the Builder's "Save & share" writing to a feed the author owns.
• ENS + feeds: human-readable permanent talks (talk.speaker.eth resolving through a feed to a manifest).
• Capacity/renewal UX: postage-stamp health surfaced in the player ("this talk is funded until ..."), one-click top-up.

Budget

Core: $5,000 (bare minimum; Builder Grant tier)

Item	Amount
Resolver + provider development, cross-browser testing	$2,800
Flagship talk preservation (encoding, upload pipeline, postage stamps for at least 1 year of hosting)	$600
Documentation, tutorial, schema/spec work	$900
Smoke-test infrastructure + gateway-fallback hardening	$500
Demo day prep, Discord support during integration	$200

Full scope: $10,000 (core + adoption & assurance)

Item	Amount
Core scope above	$5,000
Security review of the Swarm integration and persistence-provider layer (key handling, gateway trust, postage-stamp flows), performed and written up by the team's security-audit background, published openly	$1,800
Preservation drive, "talks worth keeping": curate and preserve a first collection of notable, at-risk technical talks on Swarm (with speaker permission), each a working bzz:// showcase	$1,700
Automated outreach bot: an open-source X/Twitter bot that announces each newly preserved talk with its permanent bzz:// link ("this talk now lives forever"), turning every preservation into ongoing, organic Swarm visibility	$1,000
Community workshops/write-ups (how-to-preserve-your-talk, cross-posted to Swarm channels)	$500

Team

Shayan Eskandari: security engineer and researcher (PhD, Concordia University); smart-contract security background (ConsenSys Diligence); speaker at Devcon, Dappcon, AFT and others; C4 Certified Ethereum Professional committee member; co-founder of شیر یا خط (Shir Ya Khat), the long-running Farsi-language crypto-education podcast, roughly 50 episodes of which are being preserved with this pipeline today.

Beezhan: internet-freedom activist, Art(ificial) Intelligence. Responsible for much of p2present's implementation, testing, and infrastructure work; credited openly because human–AI collaboration is how this project ships at solo-team speed, and because it aligns with the programme's interest in AI-powered decentralization: an AI agent building the tools that keep human knowledge out of centralized silos.
(GitHub · X)

• GitHub: github.com/shayanb · Project: github.com/shayanb/p2present
• Live: p2present.com · Demo talk: p2present.com/app/?p=moav-pdf

Open source & community

• MIT license, no CLA, no proprietary components; the sharing service is optional and self-hostable.
• Built to be forked: the whole site is a docs/ folder any speaker can copy to own their talk forever. Each fork is another place Swarm references circulate.
• Traction channels: conference audiences (speaker), podcast audience (Farsi crypto community), Ethereum security community.

Success criteria

• bzz:// works everywhere ar:// and ipfs:// work today, verified by the automated suite.
• The flagship talk plays end-to-end from Swarm with zero centralized dependencies.
• At least 3 additional presentations preserved on Swarm by community members during the grant period.
• The provider module documented well enough that a third-party static app can adopt the pattern in an afternoon.