Ecosystem case studies — index
Tier: Public
Status: v1.6 — custom vertical code §4.4 SOLID/SRP + yeastEditor example (2026-06-05)
Audience: young community members first, then contributors, platform evaluators, vertical builders, learners repositioning in hard times
Related: MANIFESTO.md §2.2, GETTING-STARTED.md, LICENSING.md §5.3, PLATFORM-ARCHITECTURE.md §2.1–§2.2
[!NOTE] These documents are gratitude and lesson, not product reviews. They record what our reference network felt and observed building on or adjacent to other ecosystems — so Umbraculum can name structural commitments (free try, public docs, no certification gate) without polemic.
For young community members
You are not supposed to already know why Umbraculum looks the way it does. We wrote this series so you can see the why behind our initial structural choices — not from theory, but from painful experience: for some of us directly, for others watching friends and colleagues struggle when they should not have had to.
We are not asking you to join a priesthood or wait three years before your questions count. We are asking you — also for your own good — to try to understand us: read the map below, open one or two studies that match your background, and meet us on common ground. The RFCs, public docs, and apparatus exist so you can ship structurally similar work without oral tradition (MANIFESTO.md §2.2).
That is the strong point we keep coming back to: horizontal accessibility — no badge wall, no senior-contempt ritual. If something here does not land yet, say so in the forum; that feedback is part of the design (CORE-DEVELOPMENT-AND-COMMUNITY.md §4).
Experience is not superiority — and we are not trying to recruit you
The views in this series come from years of striving — wrong turns, repositioning when the market moved, spare-time learning when there was no spare time. That is why we write these lessons down. It does not mean we are smarter or better than peers who landed on paths that worked earlier. Some developers picked the right industry, stack, or employer on the first or second try. They have the houses, the cars, the family time. Many of us in this network do not. Modesty is the point: we are not asking you to admire that gap or to bet your life on closing it through Umbraculum.
Umbraculum is not a lottery ticket. Projects like this that grow into something that funds the people who built them are statistically scarce — including ones with thoughtful manifestos and disciplined apparatuses. We do not want to lure you here. If you read on, contribute, or try the stack, do it with eyes open: weigh time, money, and opportunity cost carefully. Do not let this project cause you hardship. We say the same elsewhere on purpose (MANIFESTO.md addendum, 2026-05-20; GETTING-STARTED.md §"Who this is for") — we are offering a tool, not a future.
Even if Umbraculum succeeds beyond our current hopes, most of us will still pay the bills with legacy projects and legacy code — the work clients already fund, the stacks we already know. That is normal, not failure. Learn here if the work itself is worth it on its own terms; do not contribute on the strength of an upside that statistics say is scarce.
What this series is about
We are not scoring ERPs. We are asking one question:
When a developer must reposition — at twenty-five or forty-five, willingly or because the market forced them — can they learn on spare time because the stack is clear, upstream tools are free, and the community is visible?
The positive anchor in our network is Magento (especially 1.x / Open Source): PHP, MySQL, Community Edition, forums. Most developers who earned a living on it were hired at agencies or product companies — not instant agency founders. Learning was often unpleasant; it was still possible.
Umbraculum's line: this makes learning possible for the ones who are willing to do so — this is the stack, try it. You do not need a customer's vertical to grasp the platform.
Students and university labs — free self-host experimentation, capstone-friendly project shapes, and an honest fit filter vs production ERP: ACADEMIC-AND-EXPERIMENTATION.md (also linked from the brochure).
Bright side for vertical builders: public try paths let domain experts knock on your door with real expertise. In our experience, five minutes of technical conversation beats a certification wall. Umbraculum does not operate a certification program (MANIFESTO.md §2.2, GETTING-STARTED.md §"No certification track").
Case study map
| Study | Ecosystem | Primary failure mode (for Umbraculum) | Umbraculum response (starting points) |
|---|---|---|---|
| Omnis | Omnis Studio | Good product; ecosystem never formed — Core vs verticals only, doc-site fragility (§3.4–§3.6) | Open backbone, forum, GETTING-STARTED |
| Business Central | Microsoft Dynamics 365 BC | Huge partner channel; opaque external API surface | @umbraculum/<code>-contracts, route tables, OpenAPI alpha |
| SAP | SAP ABAP / S/4 / BTP | One language, many verticals — but trials ≠ safe boat for repositioning | One try path; no cert gate |
| TeamSystem | TeamSystem / Polyedro (Europe) | Partner-only experiment path — ecosystem is the vertical for outsiders | docker compose up; public modules |
| Odoo | Odoo Community | Partial positive — CE learnability; we skip cert/partner ladder; §4.1 platform-shape comparison (not feature scorecard) | Copy ladder, not badges |
| Adobe → Magento | Magento Open Source / Adobe stewardship | Community lost through stewardship — dev docs scattered/404, cert churn, cloud integration pipelines failing, trademark; Mage-OS as fork counter-case | AGPL + DCO; fork rights; no cert gate; CI by default; doc redirects |
| Drupal + WordPress | Drupal, WordPress | Partial positive — stewardship + community; WordPress plugin hell (auth); Drupal core-scale debate unsettled | RFC-0001 Decision F; canonical modules; community proposals; no junior contempt |
| Custom vertical code | Unnamed platforms (agency delivery) | Delivery vs craft split; priesthood unmaintainability; custom code ~90% of upgrade pain; Shopify contrast §4.3; SOLID/SRP code shape §4.4 | Apparatus + CI + CODING-STANDARDS.md SOLID; yeastEditor worked example |
Pairings that help:
- Omnis + TeamSystem — capable product, fence around builders.
- Business Central + SAP — big ecosystems, hard entry (API maze vs trial maze).
- Magento (MANIFESTO §2.2) + Odoo — learnability ladders we respect.
- Adobe → Magento + Omnis — docs treated as disposable; Adobe also lost community through stewardship; Mage-OS proves fork path.
- Drupal + WordPress + Adobe → Magento — stewardship contrast: WP/Drupal longevity vs Adobe collapse; shared PHP-era lessons.
- Drupal + WordPress (RFC-0001) — modularity yes; parallel auth hell no.
- Adobe → Magento + Custom vertical code — two independent causes: stewardship/CI (indirect) and integrator custom code (direct); not one consequence of the other; Shopify contrast (§4.3); SOLID/SRP as the code-shape mechanism (§4.4).
- Custom vertical code + Drupal §3.2 — priesthood contempt vs common ground via apparatus.
Failure modes → commitments
| Failure mode | Where we saw it | Umbraculum commitment | Mechanism |
|---|---|---|---|
| Community never formed | Omnis, TeamSystem | Community first-class | CORE-DEVELOPMENT-AND-COMMUNITY.md §4 |
| Core vs verticals only (no independent developers) | Omnis (§3.4–§3.6) | Public forum + permissionless modules as third chair | ecosystem-case-study-omnis.md §4 |
| Single closed doc channel breaks | Omnis 2016 revamp (§3.5) | Redirect or explained removal — no silent 404s | docs/README.md; RFC-0005 |
| Integration docs by topic, not by job | Business Central | Integrator-first module docs + contracts | MODULES.md §5, API-OPENAPI.md |
| Trials without repositioning ladder | SAP | This is the stack, try it | GETTING-STARTED.md |
| Partner-only platform access | TeamSystem, Omnis (historical) | Free local evaluation | docker compose up |
| Certification as sales proxy | SAP, Odoo partners, BC badges | No Umbraculum certification program | MANIFESTO.md §2.2 |
| Expertise hidden in partner tenure | TeamSystem, Omnis | Expertise offers visible — learners can find vertical builders | Public repo, Tier 3/6 modules, forum |
| Community lost after acquisition | Adobe → Magento | AGPL + governance + no CLA; fork-friendly posture | ecosystem-case-study-adobe-magento.md; LICENSING.md §9 |
| Dev docs deprioritized / scattered URLs | Adobe → Magento (Experience League era) | Developer docs first-class; redirect or explained removal | ecosystem-case-study-adobe-magento.md §3.2; MANIFESTO.md §2.2 |
| Certification renewal while ladder narrows | Adobe Commerce / Magento certs | No Umbraculum certification program | ecosystem-case-study-adobe-magento.md §3.3 |
| Trademark enclosure after fork | Adobe → Magento → Mage-OS rename | Transparent trademark policy; AGPL fork rights | LICENSING.md §8–§9 |
| Cloud pipelines broken; official guidance to avoid integration | Adobe ece-tools deploy integration | Public-repo CI by default — not broken vendor deploy gate | ecosystem-case-study-adobe-magento.md §3.5; TESTING.md |
| Agency hidden integration cost vs Shopify | Adobe cloud commercial trap | Integrator sustainability (§2.1) | MANIFESTO.md §2.1 |
| Parallel auth/session per module | WordPress plugin ecosystem | Decision F consumption contract | RFC-0001 §8.2; ecosystem-case-study-drupal-wordpress.md §4 |
| Core PR volume vs small core dogma; senior contempt for juniors | Drupal-era PHP culture | Open governance shape — RFC canonical gate + permissionless modules + apparatus | ecosystem-case-study-drupal-wordpress.md §3.2, §6; CORE-DEVELOPMENT-AND-COMMUNITY.md |
| Delivery-first vs craft split; silent agency blame of platform | Many unnamed vertical stacks | Common ground — apparatus, docs, CI; ongoing public discussion | ecosystem-case-study-custom-vertical-code.md §2–§3, §6 |
| Unmaintainable mega-classes by senior leads | Rescue / handoff projects | Structural similarity before merge; empathy for next developer | ecosystem-case-study-custom-vertical-code.md §3; CONTRIBUTING.md |
| Upgrade pain blamed on core; custom code dominant cost | Agency estates (unnamed) | Honest bottleneck naming; vertical modules advised same discipline | ecosystem-case-study-custom-vertical-code.md §4; MODULES.md |
| Clients never evaluate supplier code quality | Software procurement norms | Respectful quality questions documented — Toyota supplier logic | ecosystem-case-study-custom-vertical-code.md §5 |
| Discarding unmaintainable work onto next developer | Handoff culture | Solidarity guardrail — tools were public; fix bill is fair | ecosystem-case-study-custom-vertical-code.md §5–§6; MANIFESTO.md §3.x |
How to read these docs
- Start here if you want the map.
- Read
MANIFESTO.md§2.2 for the policy in manifesto voice. - Read
GETTING-STARTED.mdif you want to try the stack. - Open individual case studies for experience and lesson — each ends with rewrite caution where relevant.
We respect other ecosystems' commercial choices (certifications, partner programs, trial strategies). We document why Umbraculum chooses differently — by experience, not to declare winners.
Acknowledgement
These case studies exist because maintainers in this network repositioned on open stacks when proprietary fences made that harder elsewhere — and because vertical builders deserve learners and experts who can find them without a badge economy in the middle.