Running a business means juggling customers, invoices, inventory, HR, e-commerce, projects... The idea behind open source ERPs: offer a tool you can see, adapt and evolve with no black box or proprietary lock-in. In this landscape:
- Odoo — a complete, integrated ERP suite: broad coverage (CRM, sales, purchase, accounting, MRP, website/e-commerce, marketing), polished ergonomics, a sustained pace of innovation, a rich ecosystem (modules, integrators) and an open-core model, that is, a free core (Community under LGPLv3) + proprietary modules (Enterprise under Odoo Enterprise Edition License v1.0 / Odoo Proprietary License v1.0).
- Dolibarr — modular simplicity, lightness and accessibility (often chosen by SMEs, associations and freelancers).
- Apache OFBiz — a very flexible Java ERP and its lineage: SCIPIO ERP, opentaps, etc.
- Axelor — an ERP + BPM suite in Java/low-code, designed to model processes quickly.
- Compiere lineage (Compiere joined Aptean; the commercial branch has not been open source since 2010, after the acquisition by Consona): ADempiere, iDempiere, metasfresh, Libertya, OpenZ, etc.
- Odoo lineage: Tryton (derived from the Odoo project, a historical fork of TinyERP).
- Other families: ERP5, webERP/FrontAccounting, OpenConcerto, ...
In this ecosystem, Odoo plays a central role: it lets you start quickly on standard needs without closing the door to essential adaptations; it relies on a network of integrators present in many industries; it offers a cloud or on-premise offer suited to security and governance requirements; and it benefits from a structured community (among others, the OCA) that spreads good practices. The result: a controlled cost of change and a predictable evolution from the first pilot to large-scale deployments. Odoo also claims 13 million+ users worldwide, a sign of broad and continued adoption.
1- From origins to brand: TinyERP → OpenERP → Odoo
- 2005: birth of TinyERP in Belgium, with a simple idea: a modular, free ERP that you extend according to your needs.
- 2008–2014: growth, renamed OpenERP in 2008, then Odoo in 2014 to reflect the ambition of a suite beyond the classic ERP.
A nod – The short cooperation with Axelor (the very first web client)
As early as 2007, Tiny (the publisher of TinyERP) teamed up with Axelor to offer TinyERP in Web mode: a web client with its own interface and graphical functions (projects, inventory). In 2009, with the release of OpenERP 5.0, this client became OOweb 5.0 (Open Object Web), co-developed by Axelor and Tiny. This first web interface carried the Axelor logo; Axelor would later go its own way as a vendor with a 100% Java application.
2- The licensing shift: from strong copyleft to the hybrid model
2.1 The path (in brief)
- Beginnings: GPL licenses (strong copyleft).
- Server under AGPLv3: copyleft extends to usage over the network.
-
Since Odoo 9 (2015) :
- Community: LGPLv3 (more permissive than the AGPL)
- Enterprise: proprietary Odoo Enterprise Edition License v1.0 (sometimes "Odoo Proprietary License").
- This split remains in force through the current versions (principle unchanged since Odoo 9).
2.2 Scope of the Enterprise license - to avoid misunderstandings
- The clause "prohibition on publishing, distributing, sublicensing or selling copies (including modified ones) of the software without a valid subscription" applies to the Enterprise scope, not to the Community code nor, in itself, to third-party free modules.
- You can develop modules that build on Enterprise and distribute them under a license of your choice, as long as it remains compatible with the terms of the Enterprise license.
Quick summary: Community = LGPLv3 (free); Enterprise = proprietary license (use conditional on a subscription, with room to develop compatible modules).
3- The community, the OCA and the real-life compromises
3.1 The OCA (Odoo Community Association)
An independent association (created in 2013): it guides, structures and professionalizes the community ecosystem (module repositories, code conventions, reviews, PSC). At the time of the relicensing (the massive move to LGPL for the core), a public OCA ↔ Odoo SA agreement clarified the rules, notably on license compatibility and dependencies between modules.
3.2 The disagreements (and why they persist)
- The move to an open-core model fed the (recurring) idea of a "100% free" fork of the core. But maintaining a full ERP (migrations, security, QA, ecosystem) is colossal; no "core" fork has taken the lead.
- On licenses, the AGPL/LGPL debates remain lively: LGPL is seen by some as more pragmatic, by others as a weakening of the commons if there is no collective funding model in return. OCA discussion threads bear this out, even in 2025.
4- Forks and complementary paths
4.1 Tryton (the historical fork)
Tryton was born in 2008 as a fork of TinyERP 4.2. A 100% GPLv3 project, with a 3-tier Python/PostgreSQL architecture and foundation governance. Less of a "packaged product" than Odoo, Tryton moves forward with strict technical consistency, a calmer tempo, and serves as a platform (e.g. GNU Health).
Odoo vs Tryton, two philosophies:
- Tryton favors all-free and core stability.
- Odoo embraces open-core: fast cadence, wide distribution, massive investment, at the price of recurring community tensions.
Today, the Tryton project is rather niche
4.2 - Building "around" rather than "against": the Konvergo ERP example
Konvergo ERP illustrates an interesting strategy: not forking Odoo, but extending it with business modules and services (AI, automation).
- The web_konvergo_magic module (public) serves as an integration baseline with the Konvergo ecosystem (MAGIC / Allo).
- This path bets on business value + migration pragmatism when Odoo evolves (rather than rebuilding a core).
6- Two opposing visions... that feed each other
6.1 100% Open Source vision (strong copyleft)
- Everything (core + modules) stays free (GPL/AGPL).
- Monetization through services (integration, support, hosting).
- Pros: maximum transparency, independent longevity.
- Cons: less predictable R&D funding, slower "turnkey" adoption.
6.2 Dual-license / open-core vision
- Free core (LGPL) + proprietary Enterprise layer.
- Monetization through subscriptions, cloud, services.
- Pro: investment capacity (product, UX, infrastructure, go-to-market), accelerated distribution.
- Cons: tension with the community, risk of marginalizing the Community if governance is not clear.
6.3 What unites them / what separates them
- Unites: the value of an open foundation, the importance of modularity, the need for compatibility rules and an active community.
- Separates: the acceptable degree of openness, the vendor's power over the roadmap and the distribution of value (who pays what, for whom).
7- Down-to-earth practical advice (pragmatic and grounded)
- Choose the trajectory (not the "size"):
- Community = autonomy/sovereignty (more technical skills to mobilize, be careful not to let functionality erode).
- Enterprise = acceleration (SLA, time-to-value).
- Licenses & dependencies: check AGPL/LGPL/proprietary early to avoid lock-in and hidden costs. Frame your choices according to your business goals.
- Partner & team: industry references are a plus, not a deal-breaker; above all judge the people proposed and secure their assignment. Demand a short paid pilot with concrete deliverables.
- Outcome-based commitment without traps: accept it only after a scoping tied to a specific POC (SOW, acceptance criteria, assumptions). Otherwise, every deviation becomes a change order.
- Lean on the ecosystem (OCA): favor mature modules and, if possible, contribute; that means less TCO and less migration risk.
- Migrations & reversibility: version plan, test set, rollback gates. Ensure data export and a plan B if a component becomes blocking.
- Governance by value: measure business KPIs (cash, lead times, data quality).
- Adjust accordingly.
Conclusion: community, model... and direction (lessons learned)
I have had the opportunity to work on Dolibarr projects (from 2007 to 2014) and Apache OFBiz (2006-2007): the effective sobriety of one and the heaviness of the other taught me a lot. Originally, I had an affinity for OFBiz coming from the Java world. Since 2009, I have been implementing Odoo: varied projects, migrations, industry verticals.
Two truths can coexist without contradicting each other:
- Odoo would never have taken off without its community. The OCA, the contributors, the integrators and field feedback made the ecosystem hold up in real life.
- Odoo would not be where it is without the dual license. The Community (LGPLv3) + Enterprise (Odoo Enterprise Edition License v1.0) duo funded R&D, the product cadence and industrialization, and, indirectly, pulled the open source community along by broadening adoption.
Odoo is neither the Platonic ideal of "all-free" nor a proprietary citadel. It is a ridge line: keeping an open core, helping a community thrive and funding the ambition with an Enterprise layer, without letting the Community become an empty shell. As long as these safeguards hold, the bet remains a winning one: an ERP that is powerful, modular, durable... and, deep down, very faithful to the spirit that launched TinyERP.