The durable output of Wave 0. Actor mapping is a project with an end; this is the standing system it leaves behind — the set of channels we follow so the knowledge base stays current instead of ageing from the day it is written.
The principle that keeps this alive
Every entry has a channel, a cadence, an owner, and a reason. An entry with no reason gets removed at the next review.
Watchlists die by accumulation. A hundred feeds nobody reads is worse than twelve that get read, because it produces the feeling of being current without the substance. Prune aggressively and treat removal as maintenance, not failure.
Tiers of attention
| Tier | Meaning | Typical count | Handling |
|---|---|---|---|
| Follow closely | Consistently produces things that change what we know or do | 5–15 | Read everything. Notify on publication. |
| Skim | Occasionally valuable, often not | 20–50 | Batch review on cadence; scan titles, open selectively |
| Periodic check | No feed, or very low frequency — visit deliberately | 10–30 | Calendar reminder, quarterly or semi-annually |
| Event-driven | Only relevant around a date | any | Conference programmes, annual reports, grant rounds, regulatory cycles |
Promotion and demotion between tiers is the main thing that happens at review.
Channel types and how each is captured
| Channel | Capture mechanism | Notes |
|---|---|---|
| Blog / news site | RSS | Still the best mechanism where available |
| Newsletter | Dedicated email address, not a personal inbox | Keeps the signal separable and searchable |
| Podcast | Podcast app, or RSS into the same reader | |
| YouTube / video | RSS (channels expose feeds) or app subscription | |
| Academic journal / author | Journal alerts, citation alerts on key papers and authors | The highest signal-to-noise of any channel type |
| Preprint / repository | Saved topic queries with alerts | |
| Official bulletin / regulation | Scheduled check — rarely has a usable feed | Regulation and grants are the fastest-decaying facts we hold |
| Grant and tender records | Scheduled check, tied to funding rounds | Shows where money is actually going |
| Social accounts | Lists, not the main feed | High noise; use only for genuinely high-value individuals |
| Events | Calendar entries for programme publication and registration dates | |
| People | Periodic conversation | The least scalable and often the most valuable |
Consolidate the reading surface. Most of the above can land in one RSS reader plus one dedicated email address. Two places to look is sustainable; nine is not.
Cadence
| Rhythm | Activity |
|---|---|
| Weekly | Clear the “follow closely” tier and the newsletter address. Short — 30 minutes. |
| Monthly | Skim tier. Capture anything worth keeping as a source note. |
| Quarterly | Periodic-check tier. Prune the list. Promote and demote. Review whether anything learned should change a domain brief. |
| Semi-annually | Regulation, grants and prices — the fast-decaying facts. Re-run a small snowball round to catch actors who have appeared since. |
| Annually | Re-read the landscape synthesis. Is the white space still white? |
Triage rule
For anything the watchlist surfaces, exactly one of four outcomes — decided immediately, not deferred:
- Changes a claim → update the relevant domain brief, log the source note.
- Is a source but not now → log in the source registry, do not read yet.
- Is an actor → add to the actor registry, triage per Wave 0 step 2.
- Is interesting but not actionable → content backlog, or discard.
There is no “read later” pile. A read-later pile is outcome 4 in disguise.
Keeping it machine-readable
The register below is a table for now. Once it exceeds roughly thirty entries it should
become a structured file (watchlist.yaml) with fields: name, actor, url,
feed_url, type, tier, cadence, language, domains, owner, reason,
last_reviewed.
The reason is practical: this repository already has agent tooling configured, and a structured watchlist is something a scheduled agent can actually work through — fetch, filter against our domain map, and produce a digest of what is worth human attention. That turns monitoring from a discipline that decays into one that runs. Worth doing once the list is large enough to be a burden, and not before.
Register
Seeded from Wave 0 round 1. Feed URLs still need collecting — that is the next mechanical task. Keep this list short enough to actually read.
Follow closely
| Name | Actor | Type | Cadence | Domains | Lang | Reason |
|---|---|---|---|---|---|---|
| GOB Mallorca | GOB | site / news | weekly | W, C2, C5, R | CA/ES | Most consistently active local voice on water and land policy; currently contesting the Hydrological Plan |
| Terraferida | Terraferida | site / press page | weekly | A1, W, C4, C5 | CA | Now explicitly research-focused; produced the pool census |
| DG Recursos Hidrics — water portal | Govern IB | official | monthly | W, C2 | CA/ES | Reservoir levels, aquifer status, Hydrological Plan documents |
| Consell de Mallorca + BOIB — rustic land and pools | Consell / Govern IB | official bulletin | monthly | C5, A1 | CA/ES | Live legalisation window closing Feb 2028; fastest-decaying facts we hold |
| Enertech — studies | Enertech | site / reports | monthly | P1, E4, H | FR | Closest methodological analogue to our evidence protocol |
| Low-tech Lab | Low-tech Lab | site / open-source docs | monthly | W1-W3, H3, H5 | FR | Documented, replicable solutions; the Site Zero model in practice |
| IBAVI / Balearic bioclimatic building | IBAVI | project pages, press | quarterly | H1, M1-M4 | CA/ES | Hyper-local bioclimatic work with quantified targets |
Skim
| Name | Actor | Type | Cadence | Domains | Lang |
|---|---|---|---|---|---|
| Chaleur Humaine | Le Monde | newsletter + podcast | weekly | P2, P3, E | FR |
| Bon Pote | Bon Pote | site | monthly | cross-cutting | FR |
| Reporterre | Reporterre | site | monthly | cross-cutting | FR |
| Sismique | Sismique | podcast | monthly | discovery | FR |
| negaWatt | negaWatt | publications | quarterly | E, P3 | FR |
| IDAE | IDAE | guides + incentives | quarterly | E, H, C5 | ES |
| PEP Passivhaus | PEP | site + conference | quarterly | H1, H4, H6 | ES |
| GBCe | GBCe | reports | semi-annual | M, H, P2 | ES |
| IBE — Balearic energy agency | IBE | programmes, grants | quarterly | E, C3, C5 | CA/ES |
| Cluster TEIB | Cluster TEIB | member news | quarterly | C6, T1 discovery | ES |
| Clean energy for EU islands | EU Secretariat | news + events | quarterly | C3, E3 | EN |
| Plan Bleu | Plan Bleu | publications | semi-annual | C1, C2, W | FR/EN |
| MedECC | MedECC | reports | semi-annual | C1 | EN |
| IOB | IOB | newsletter + congress | semi-annual | A1 | DE/EN |
| Maisons Paysannes de France | MPF | wiki + magazine | quarterly | H1, H2, M1 | FR |
Periodic check
| Name | Actor | Cadence | Why |
|---|---|---|---|
| ABAQUA open data portal | ABAQUA | quarterly | Check what datasets are exposed and whether they update |
| OSMM indicators | OSMM / MAPF | annual | Island-level environmental indicator series |
| UIB research groups and campus monitoring | UIB | semi-annual | 800-sensor campus; potential P1 partner |
| CIHEAM Zaragoza course catalogue | CIHEAM | semi-annual | Short courses on Mediterranean water — training, not just reading |
| CEBAS-CSIC irrigation publications | CEBAS-CSIC | semi-annual | Deficit irrigation, non-conventional water |
| Cyprus Institute EEWRC | CyI | annual | Island WEFE-nexus research |
| Grants and subsidies — national, regional, EU | various | semi-annual | High churn; see principle 10 |
Event-driven
| Name | When | Why |
|---|---|---|
| Mallorca construction, garden and pool fairs | check programmes annually | Cheapest R1 actor-discovery instrument available |
| Spanish Passivhaus Conference (PEP) | annual | Mediterranean-climate sessions; speaker list |
| IOB international congress | periodic | Natural pool technical centre of gravity |
| COAIB events agenda | rolling | Route to R1 architects |
| Clean energy for EU islands — Mallorca workshops | as announced | Direct local access to the island energy actor set |
Removed
Kept deliberately, so we do not re-add things we have already rejected.
| Name | Removed | Why |
|---|---|---|