Ecco Pro Replacement — Personal Information Manager (PIM) Alternatives

Question: Ecco Pro is the last Windows-only, desktop-only tool in daily use. What, if anything, should replace it?

Scope discipline: Core/Processes/Behavioural-Solutions.md measures tool consolidation — removing execution friction — at ≈0 behavioural effect (UK Current Account Switch Service: 12 years, 11.9M switches, no net change). A migration is only justified by a capability Ecco has that the vault cannot get. “One app to rule them all” is the failure mode here, not the goal.

Bottom line

Keep Ecco Pro. Do not migrate to any of the six candidates. Ranked: 1. Obsidian (incumbent — the only route worth trialling) · 2. SiYuan (fallback, only if the Obsidian route dies) · 3–7. Tana, Tolaria, Capacities, ClickUp, Dart AI — all disqualified on Talbot’s own criteria.

Inside Obsidian there are two ways to get Ecco’s per-item columns, and they trade against each other:

Route	Multi-level hierarchy	Grid editing	What it costs
Dataview + inline fields on bullets	✅ native folding, arbitrary depth, one note per outline	❌ the table is read-only query output	typing [abc:: A] into each bullet by hand
Bases + file-per-item (frontmatter)	❌ parent/child is a parent: property — one flat level, no folding	✅ confirmed 2026-08-20 — cells edit in place and the change persists to the note’s frontmatter	one note per item
Ecco Pro	✅	✅	Windows desktop only

The real gap is narrower than “hierarchy” — it is hierarchy inside a filtered view. The master outline keeps full depth; Obsidian Outliner folds it exactly like Ecco. It is the notepads that were the problem. Tested on screen, 2026-08-20:

Query form	Filters correctly?	Keeps the tree?
Dataview TABLE (+ computed Parent column)	✅	⚠️ one level of ancestry, as a column
Dataview LIST	✅	❌ matches render flush left
Dataview TASK	❌ no — matched only top-level roots, then dragged every descendant along (three thisWeek items leaked into a today view); a field on a subtask is ignored entirely, so a deep-match query returned “No results to show”	✅
dataviewjs (walk the tree, keep matches + ancestors)	✅	✅

The dataviewjs block is the answer — ~25 lines, parameterised by two constants at the top (FIELD, VALUE), one copy per notepad. It keeps every match plus its ancestor chain, italicises the ancestors so context is visibly distinct from matches, and drops non-matching children. Logic verified outside Obsidian against the same outline; render check in Focus/proofs/ecco-treeview-proof.md.

Be honest about what that costs. It is custom JavaScript living in the vault — precisely the “build-it-yourself flexibility over practical out-of-the-box functionality” trap named in SDC/Strategy/Identity/Internal View/Weaknesses.md. Ecco does this natively with no code at all. The mitigating facts: it is short, it is written once, and every view after the first is a copy with two words changed.

Before scoping any migration on this point: Ecco’s filtered-but-still-nested notepad may simply be unique. SiYuan’s attribute views are database tables, and tables are flat — this was not verifiable from public documentation and must be confirmed in the actual application before SiYuan is treated as the fix. Switching tools to solve a problem the new tool also has is the expensive version of this mistake.

Hierarchy depth is a hard requirement, not a preference (Talbot, 2026-08-20): “Only simplistic task items have a single parent. Real life project management involves tasks and subtasks that are multiple levels deep.” That single constraint knocks the Bases file-per-item route out of contention for outline work, despite it winning on editing — a parent: property models one level, not a tree. Bases stays useful for genuinely flat sets of notes; it is not an Ecco outline replacement.

Precise verdict on the hypothesis: Bases cannot do item-level fields inside a note — Obsidian’s docs are explicit that “each row is a file, and each column is a property of that file.” It works fine when the item is a file. So “frontmatter + Bases” is not dead; it is a different design with a different cost, and it is the variant that competes with Ecco on the one dimension Ecco actually wins: direct grid editing.

Both are proven in the vault: Focus/proofs/ecco-columns-proof.md (Dataview, 7 items / 5 columns / 3 views) and Focus/proofs/ecco-bases-variant-proof.md (Bases over proofs/items/, same seven items). Neither could be rendered headlessly — Obsidian has no headless mode — so expected row counts were computed mechanically instead and the render check is Talbot’s.

The honest verdict remains the one the task allowed for: keep using Ecco, keep migrating opportunistically — and trial exactly one real notepad in the vault first.

What Ecco actually provides (the bar any replacement must clear)

Capability	Detail
Hierarchical collapsible outlining	Decades of muscle memory; keyboard-driven
Per-item columns	ABC, Status, To-Do's, Action, Focus — attached to the item, not the file
Multiple views over the same items	13 notepads: Working On, Today, Now, Planning, This Week, Focus, Development, Finances, Research, Maintenance, Health, Recurring, Accounting
Local, private, free, no account	Runs entirely offline on the desktop
Not provided	Any-device access — Ecco is Windows desktop only

Two gaps therefore: per-item columns + views (gap 1), and any-device (gap 2).

Candidate comparison

Scored against Talbot’s stated ideal. ✅ = meets it, ⚠️ = partial, ❌ = fails.

	Obsidian + Dataview (incumbent)	SiYuan	Tana	Capacities	ClickUp	Dart AI	Tolaria
Local-first	✅ plain files on disk	✅ local-first, self-hostable	❌ cloud-only	⚠️ cloud store, offline editing	❌ cloud-only	❌ cloud-only	✅ plain files + Git
Data privacy	✅ never leaves the machine	✅ optional end-to-end encrypted sync	❌ vendor holds the data	❌ vendor holds the data	❌ vendor holds the data	❌ vendor holds the data	✅
No vendor lock-in	✅ Markdown	✅ open source, Markdown export	⚠️ Markdown/JSON export, live data is theirs	⚠️ import/export, format unstated	❌ API-only	❌ export not documented	✅ Markdown
Per-item columns	✅ Dataview inline fields (bullets) or Bases + file-per-item (frontmatter)	✅ block-level attributes + database views	✅ supertags with fields on any bullet	⚠️ object properties, not outline items	❌ task fields, not outline items	❌	❌
Multiple views over same items	✅ one query per view, unlimited	✅ attribute views	✅	⚠️	⚠️	⚠️	❌
Any-device	⚠️ needs Sync ($4/mo) or remotely-save	⚠️ needs sync purchase or self-host	✅ (but online-only)	✅	✅	✅	❌ desktop only, Git-based
AI / agent integration	✅ Claude Code operates directly on the files	✅ agent-oriented, MCP	⚠️ built-in AI, their models	⚠️	⚠️ ClickUp Brain add-on	⚠️	✅ MCP, Claude Code
Model-agnostic	✅	✅	❌	❌	❌	❌	✅
Cost	$0 (+$4/mo if Sync)	$0 local · $64 one-time PRO · $148/yr cloud	$10–18/user/mo	free tier + Pro	$7/user/mo (+$9 AI)	free tier · $8–12/user/mo	$0
Runs on Windows	✅	✅	✅	✅	✅	✅	✅ Windows x64 build

Notes per candidate

• Obsidian — the incumbent, and the recommendation. Obsidian Outliner (v4.9.0, enabled) closes the outlining half of gap 1: folding, indent/outdent, moving items and subtrees by keyboard. What it does not do is the half that matters — it has no per-item columns and no saved views over items. That is why the columns question is the whole question. Dataview (v0.5.68, enabled) answers it for bullets; Bases (core, enabled) answers it for file-per-item. See the trade table above. The residual loss against Ecco is ergonomic, and it is the “build-it-yourself flexibility” trap named in SDC/Strategy/Identity/Internal View/Weaknesses.md — which is exactly why Phase 1 carries a kill criterion.
• SiYuan — the only candidate that structurally matches Ecco and stays local: an outliner with real nested blocks, block-level attributes, database/attribute views with editable cells, open source, self-hostable, Windows. Once multi-level depth is a hard requirement it is the only option on this page that offers depth and per-item columns and editable views at once — the same three Ecco has. Cost: one-time $64 (PRO, third-party S3/WebDAV sync) or $148/yr for their 8 GB cloud; free for local-only use. Against it: adopting it means running a third system beside Ecco and the vault, which is exactly the ≈0-payoff move the behavioural evidence warns about. Escalation rule: it becomes the live candidate the moment the Dataview route fails the depth test — not before.
• Tana — conceptually the closest thing to Ecco built this decade (supertags attach fields to any bullet; every view is a query over those). Disqualified on Talbot’s own criteria: cloud-only, no reliable offline mode, subscription, not model-agnostic.
• Capacities — object-centric rather than outline-centric; offline support exists but the store is theirs. Does not match the per-item-column model.
• ClickUp — project-management SaaS. Cloud-only, per-seat, AI is a paid add-on with a thin API. Wrong shape entirely for an outline-with-columns workflow.
• Dart AI — cloud task manager. Its own pricing page documents no local storage, no self-hosting, no offline mode, and no export. That alone disqualifies it on lock-in.
• Tolaria — genuinely aligned in philosophy (files-first Markdown + YAML frontmatter, Git-backed vault, first-class Claude Code / MCP support, free, open source), and the 2026-08-19 release does ship a Windows x64 build — several secondary write-ups still say Mac/Linux only, which the release assets contradict. But it fails on the capability that matters: its properties and Views are note-level, exactly like Bases, and it is not an outliner at all. It is a candidate to replace Obsidian, not Ecco — and replacing a working vault buys nothing here.

Gap 2 — any-device

Separable from gap 1, and cheaper to close.

Option	Cost	Notes
Obsidian Sync	$4/mo billed annually, $5/mo monthly	Vault is 57 MB across 2,022 notes — small enough that no published storage tier is a constraint (the pricing page itself no longer lists per-tier storage; confirm at signup). One subscription covers desktop + Android + iOS. Simplest working answer.
remotely-save plugin	$0 + storage	Already installed in the vault but not enabled. Syncs to S3 / WebDAV / Dropbox / OneDrive. Free, but a moving part to maintain.
Git-based sync	$0	Collides with a standing rule — all repos are LOCAL-ONLY, never pushed (AGENTS.md). Mobile git sync needs a hosted remote. Do not adopt without an explicit decision to break that rule.
Deployed read-only view	$0	Already exists — web-deploy <brand>-focus ships vault content to the web. Read-only, so it answers “see it anywhere,” not “edit it anywhere.”
Dynalist (status quo)	already in use	Already provides any-device outlining for capture and reference. Gap 2 is therefore partly closed already — worth confirming what is actually missing before paying for anything.

Recommendation — phased, reversible

Phase 0 (30 seconds, decides Phase 1’s shape): open ecco-bases-variant-proof.md and try to edit a status cell in the table. If the cell edits in place and writes back to the item note’s frontmatter, Phase 1 trials the Bases variant; if it is read-only, Phase 1 trials the Dataview variant. The kill criterion below is “does entering the fields annoy him into quitting” — so if one variant removes the typing, that is the variant worth the week.

Phase 1 (hours, abandon cheaply): pick one real Ecco notepad — Today is the obvious candidate, ~20–40 items — and rebuild it in the vault using whichever variant Phase 0 selected, with two or three views. Run it in parallel with Ecco for a week. Both proof notes are working templates.

• Kill criterion, decided in advance: if maintaining the fields is annoying enough that Talbot stops doing it within that week, the route is dead — delete it and stop. That is a legitimate result, not a failure to try hard enough.

Phase 2 (only if Phase 1 survives): close gap 2 with Obsidian Sync at $4/mo — after first checking whether Dynalist already covers what is actually needed on the phone.

Phase 3 (only if Phase 1 dies): evaluate SiYuan as the single Ecco successor, judged on one question only — does its block-attribute model let Talbot edit item columns as directly as Ecco does?

Do not: attempt a bulk migration of the eleven .eco files (≈48 MB, Main.eco alone 15 MB and modified today). Continue the opportunistic per-project migration already underway.

What would be lost, honestly

• Nothing matches Ecco on hierarchy and grid editing at once. Each in-vault route gives up one of the two: Dataview keeps the folding outline but the table is read-only; Bases gives editable cells but flattens the hierarchy into a parent: property. SiYuan is the only candidate that claims both, and it is a third system. For the specific act of re-prioritising twenty nested items in one pass, Ecco is still the fastest thing on the machine.
• Decades of keyboard muscle memory — unmodellable, and a real switching cost that no feature table captures.
• Zero-risk status quo. Ecco works, costs nothing, and is not going to be discontinued any harder than it already has been.

Notes

• PIM — Personal Information Manager. The category Ecco Pro belongs to: outline, contacts, calendar, and task data in one desktop application.
• Bases — Obsidian’s built-in no-code database feature (core plugin, enabled in this vault). Rows are files.
• Dataview — a community plugin (v0.5.68, enabled) that queries notes and their inline list items. FLATTEN file.lists is what turns bullets into rows.
• MCP — Model Context Protocol. The standard several 2026 tools use to expose their data to AI agents.
• SSOT — Single Source of Truth.
• CASS — UK Current Account Switch Service. The 12-year natural experiment behind the ≈0 finding cited at the top.

Sources

• Obsidian Bases help — “each row is a file, and each column is a property of that file”
• Obsidian pricing / Sync
• SiYuan · SiYuan pricing
• Tana pricing
• Capacities pricing
• Dart AI pricing
• ClickUp — developer perspective, 2026
• Tolaria · releases