The enterprise resource planning system (ERP) is the operational heart of a company. It concentrates customer data, supplier data, HR data, accounting and financial data, and product data. Choosing an ERP is therefore not just choosing software: it is choosing whom you entrust with your information assets, under which jurisdiction and with what level of reversibility.
In 2026, in a context where the US Cloud Act, FISA 702, extraterritorial sanctions and geopolitical fragmentation are challenging the certainties of the past twenty years, sovereignty can no longer be treated as a secondary IT department topic. It has become a matter of governance and strategy.
The purpose of this article is not to promote a solution. Its aim is to lay out a rigorous analysis framework and to present each major ERP on the market with its strengths and limits, so that each company can make its own trade-offs based on its own context.
Transparency note. ANOR is an Odoo consulting and integration firm. We have nevertheless made the effort in this article to be as objective as possible, presenting the strengths and limits of each solution in a balanced way — including Odoo, whose limits and constraints we know from experience. Our position as an integrator does not prevent us from acknowledging that another ERP may be the right choice for a given context. The goal of this text is to help ask the right questions, not to push an answer.
Four axes for assessing an ERP's sovereignty
Sovereignty is not a box to tick. It is a combination of several dimensions that must be assessed together.
- The vendor's jurisdiction. Under which law can the vendor be compelled to hand over your data? A vendor subject to US law — even if it hosts in Europe — can be compelled by the Cloud Act to provide your data to US federal authorities, without you being informed.
- The nature of the code. Is the code proprietary and closed, or open source, auditable and modifiable? Closed code creates dependence on the vendor for security, maintenance, evolution and migration. Open code guarantees reversibility.
- The tech stack: language and database. An ERP relies on a programming language and a database, each with its own license and economic implications.
Some languages are free and universal: Python (PSF License, free), Java (OpenJDK under GPL+Classpath Exception), PHP (PHP License), JavaScript. Their developer pool is global, their tools are free, their documentation is abundant, and the skills are transferable from one project to another. Others are proprietary and specific to a single vendor: ABAP at SAP, X++ at Microsoft Dynamics F&O, SuiteScript at Oracle NetSuite, Harmony (4GL) at legacy Divalto. The skills pool there is limited, rates are high, tools are often paid, and developers trained on these technologies are captive to the vendor's ecosystem.
The same dichotomy exists for databases. PostgreSQL (PostgreSQL License, BSD-type), MariaDB and MySQL (GPL) are free and open. Oracle Database, Microsoft SQL Server, Azure SQL Database and SAP HANA are proprietary and paid. In some cases, the database license represents a significant share of the ERP's total cost of ownership, sometimes on a par with the application license itself.
A fully open stack (language + DB) guarantees portability, longevity, cost control, and access to an open skills market. A proprietary stack locks you in technically and financially and makes future migrations difficult. Changing the DB or the language means rewriting the ERP. - Hosting. Where is the data physically stored and who operates the infrastructure? A French host such as OVHcloud, Scaleway or Outscale offers different legal protection from AWS (Amazon), Azure (Microsoft) or Google Cloud Platform (Google), even when the latter offer European regions. An operator subject to US law remains exposed to the Cloud Act, regardless of the physical location of its data centers.
An ERP can be European but proprietary and therefore locked in. An ERP can be open source but hosted on an American cloud and therefore exposed. Real sovereignty is the intersection of these dimensions.
Overview
We present ERPs in two groups. First, solutions with an open source core, whose model fundamentally changes the relationship between the company and its vendor. Then proprietary solutions, whose functional strengths are often real, but which carry specific risks that must be faced squarely.
Open source core ERPs
An ERP that is open source is not simply software with a gentler license or a lower price. It is a distribution model that profoundly changes the relationship between the company and its vendor:
- The code is auditable. Any competent integrator or developer can read the code, verify its behavior, identify a bug or vulnerability, and contribute a fix.
- The data is directly accessible. In a standard PostgreSQL, MariaDB or MySQL database, you can extract, back up and migrate your data without asking the vendor's permission and without depending on its export tools.
- Reversibility is structural. If the vendor tightens its terms, changes strategy or disappears, the code remains usable and maintainable. The community or another provider can take over.
- The exit cost is under control. No license to buy back, no compliance audit, no proprietary format to decipher.
One nuance: the so-called open core model (a free Community part + a paid Enterprise part with accessible but non-redistributable sources), practiced notably by Odoo and Axelor, dilutes this "purity". The Enterprise part remains under a proprietary license, even if its code is readable. Strict GPL or AGPL models (Dolibarr, Tryton, ERPNext, Axelor Community), on the other hand, offer total, redistributable openness.
Odoo
Profile. Belgian vendor (Ramillies). All-in-one open core ERP suite. Community under LGPLv3, Enterprise under a proprietary license with source access. Three official hosting modes: Odoo Online (SaaS), Odoo.sh (PaaS on Google Cloud Platform — GCP, Google's infrastructure cloud, a US operator subject to the Cloud Act), On-Premise.
Strengths. Very broad functional coverage (90+ official modules, thousands of third-party apps). Modern UX. 100% open tech stack (Python/PSF, PostgreSQL). Global partner ecosystem. Moderate entry cost. Customization accessible to any Python developer. Official, fully supported on-premise mode, opening a path to complete sovereignty.
Limits. Centralized, imposed vendor roadmap. Short version cycle (annual release, 3-year support in Enterprise) that forces regular, costly migrations. Enterprise pricing policy that is tightening (price increases). Enterprise code not free despite source access — the open core boundary is often debated. Coexistence with the OCA (Odoo Community Association) community, which maintains a large catalog of complementary free modules: the relationship between the vendor and the OCA is dynamic, sometimes under discussion, and illustrates the duality of the open core model. Functional depth varies by module: some remain light compared with vertical specialists (complex manufacturing, EAM, specialized retail).
Sovereignty. Software: excellent (BE, open stack). Hosting: depends on the mode chosen — Online and Odoo.sh on GCP and therefore subject to the Cloud Act, On-Premise on a sovereign or own cloud = fully sovereign.
Dolibarr
Profile. French project backed by the Dolibarr Foundation (non-profit). GPLv3 license. No central commercial vendor — a pure community model.
Strengths. Maximum sovereignty on all four axes (FR, strict GPL, open PHP and MariaDB or PostgreSQL stack, free hosting). No license to pay. Active and lasting community. Excellent fit for very small businesses and SMEs. Dolistore marketplace with numerous third-party modules. Simple architecture, easy to maintain.
Limits. More limited functional coverage than Odoo, especially in manufacturing, multi-country, advanced supply chain, complex production. Less modern UX. Limited functional depth for complex mid-sized companies. Smaller, less structured ecosystem of professional providers. Non-profit governance = no predictable commercial roadmap. The community model is a strength for sovereignty, a weakness for buyers looking for a strong vendor commitment.
Sovereignty. Maximum, on all axes, without reservation.
Tryton
Profile. Historic fork of OpenERP (2008), backed by the Tryton Foundation (present in Germany, Belgium, Spain). GPLv3. Python/PostgreSQL stack.
Strengths. Rigorous, modular, stable architecture. Strong accounting depth. Clean code, ideal for developers. Maximum sovereignty (European foundation, open stack).
Limits. Austere UX, little investment in user experience. Very limited partner ecosystem. Documentation sometimes sparse. Limited job market. High initial expertise cost.
Sovereignty. Maximum.
Axelor
Profile. French vendor. Open source ERP/BPM/CRM suite under AGPLv3 (free Community, paid Enterprise). Java/PostgreSQL stack. Built-in low-code Studio.
Strengths. Low-code Studio allowing customizations without coding. Broad functional coverage (unified ERP + BPM + CRM). French vendor. Open stack.
Limits. Modest market share compared with Odoo. Limited partner ecosystem. AGPLv3 license is restrictive for proprietary integrations (any modification used over a network must be published). Low international recognition. Average functional depth in some specialized domains.
Sovereignty. Excellent on all four axes in the Community edition. Enterprise remains under a proprietary license (with source access).
ERPNext (Frappe)
Profile. Indian vendor (Frappe Technologies). GPLv3. Python/MariaDB stack. Broad functional coverage.
Strengths. Fully open source. Open stack. Broad functional coverage. Zero entry cost. Dynamic community. Frappe platform reusable for developing other applications.
Limits. Indian vendor — geographic and cultural distance, durability of the contractual relationship to be examined. Modest European provider ecosystem. French localization could be improved (accounting, tax, payroll). Average functional depth in some critical domains.
Sovereignty. Good (India outside the US extraterritorial regime, strict GPL, open stack). Reservation about the vendor's location, which can be an issue for some European organizations.
Proprietary ERPs
Before detailing each solution, the risks specific to the proprietary model must be named clearly. These risks are not theoretical. They materialize regularly in companies' lives, and they are structurally absent from the open source model.
The risk of data captivity. Your business data (customers, suppliers, transactions, sales history, accounting, inventory) is stored in formats and structures controlled by the vendor. In the event of contract termination, a commercial dispute, a pricing policy change or the vendor's acquisition, recovering your data in a usable format can be very expensive — and some vendors explicitly charge for a full export, or only offer partial, degraded exports, forcing the company to pay for extraction services to get its own information back. In a pure SaaS model, the situation is even more asymmetric: the data is at the vendor, on its infrastructure, and it technically holds the keys.
The risk of license audits. The major proprietary vendors (SAP, Oracle, Microsoft, IBM) conduct compliance audits that frequently result in significant retroactive charges, sometimes amounting to hundreds of thousands or even millions of euros, when the company is judged to be over-using its license rights — whether the over-use is deliberate or results from inextricable contractual complexity. These audits are a commercial lever as much as a legal mechanism.
The risk of imposed end of support. The vendor unilaterally decides the end of life of its versions. SAP has set the end of ECC 6.0 support at the end of 2027 (paid extension until 2030), forcing its customers to migrate to S/4HANA on a schedule they did not choose. Oracle, Microsoft, Sage or Cegid regularly carry out similar transitions. The customer is bound to the vendor's timetable, not to its own business pace.
The risk of captive price increases. Once the ERP is deployed and business processes are intertwined, the exit cost is such that the vendor can raise renewal or maintenance fees without fearing a mass departure of customers. Annual increases of 5 to 20% on proprietary SaaS suites have become common, and accumulate year after year.
The risk of disappearance, acquisition or strategic pivot. A vendor can be acquired (Infor by Koch, Talentsoft by Cegid, and many other cases), with immediate consequences for product strategy, commercial terms or roadmap. In the extreme case of bankruptcy or product discontinuation, customers are left with unmaintained software, no accessible source code, and no technical exit path.
The risk of the proprietary stack. When the language (ABAP, X++, SuiteScript) or the database (HANA, Oracle DB, SQL Server, Azure SQL) are themselves proprietary, the above risks are multiplied: the skills to maintain the system are rare and expensive, the underlying licenses add to the ERP's, and migrating to another foundation requires a complete rewrite.
None of these risks exists structurally with free software deployed on controlled infrastructure: the code is auditable, the database is standard and accessible, the data can be exported without permission, and operational continuity does not depend on the commercial health of a single vendor. The proprietary model, by contrast, rests on a fundamental asymmetry: the vendor holds the technical and contractual power, the customer pays to keep using what it has already paid for.
This does not mean proprietary solutions should be rejected wholesale. Their functional strengths, their maturity, the depth of their industry ecosystems can justify the choice in certain contexts. But these strengths come at the price of the risks described above, which must be honestly factored into the decision.
SAP S/4HANA and variants
Profile. Global reference for large-enterprise ERP. German (Walldorf), listed in Frankfurt and on the NYSE. Offered as S/4HANA (large enterprise), Business One (SME), Business ByDesign and GROW with SAP (mid-market SaaS).
Strengths. Unmatched functional depth in finance, supply chain, manufacturing. De facto standard in industry, aerospace, energy, pharma. Massive global ecosystem (consultants, training, certifications, specialists). Ability to support very large organizations with high regulatory or multi-country complexity.
Limits. Very high total cost of ownership (licenses, integration, maintenance at 22% per year, rare and expensive resources). Abrupt and costly migrations (ECC end of support set for end of 2027, paid extension to 2030). Proprietary ABAP code requiring rare expertise. Historically complex UX (Fiori improves but does not solve everything). Unilateral vendor roadmap. Proprietary HANA DB stack. License audits known to be aggressive.
Sovereignty. Partial. Favorable German jurisdiction and native GDPR, but NYSE listing, massive US operational presence and RISE/GROW cloud offerings largely running on AWS, Azure and GCP reintroduce extraterritorial exposure through hosting. The on-premise version on sovereign infrastructure remains possible, but runs counter to the vendor's current commercial strategy.
Oracle NetSuite
Profile. American pure-play SaaS (Oracle), dominant in the mid-market and finance-oriented mid-sized company segment, present in France.
Strengths. Pure SaaS with no infrastructure management. Native internationalization (multi-currency, multi-country, compliance). Robust financial reporting. Continuous updates. Short implementation time on standard scopes.
Limits. Weak manufacturing and production coverage. Customization constrained by SuiteScript (a proprietary JavaScript-like language). Cost that grows quickly with the number of users and modules. Strong technical and contractual lock-in on Oracle. Entirely proprietary DB stack. Oracle is historically known for its license audit practices.
Sovereignty. None. US vendor, Oracle infrastructure, Cloud Act and FISA 702 applicable.
Microsoft Dynamics 365
Profile. Modular ERP/CRM suite from Microsoft, hosted on Azure: Finance & Operations (large enterprise), Business Central (mid-market), Customer Engagement (CRM).
Strengths. Native integration with Office 365, Teams, Power BI, Power Platform. Modern UX. Massive partner ecosystem. Strong consistency with an already 100% Microsoft IT environment. Low-code customization capability via Power Platform.
Limits. High and rising license costs. Very strong Azure lock-in. Complex migrations between versions. Low portability outside the Microsoft ecosystem. Dependence on Power Platform for extensibility, which itself becomes a recurring cost. X++ is a proprietary Microsoft language.
Sovereignty. None. US vendor, Azure hosting, Cloud Act applicable. The Bleu offerings (Microsoft + Capgemini + Orange) aim to mitigate this risk but do not remove the legal uncertainty as long as the underlying technology remains under US control.
Infor CloudSuite
Profile. Industry-vertical suite (fashion, automotive, distribution, healthcare, industrial equipment). American, owned by Koch Industries since 2020.
Strengths. Very deep industry verticals (Fashion, Automotive, M3 discrete manufacturing, LN). SaaS on AWS. Approach based on industry best practices standardized by sector.
Limits. Less dense partner ecosystem in France than SAP or Microsoft. Customization tools being modernized. High cost. AWS lock-in. Platform modernization sometimes painful for legacy customers.
Sovereignty. None. US vendor, AWS, Cloud Act applicable.
Sage
Profile. British, long present in France. Fragmented range: Sage 100 (French very small businesses/SMEs), Sage X3 (international mid-market), Sage Intacct (finance SaaS, designed in the US), Sage Business Cloud.
Strengths. Product maturity. Very solid French tax and social compliance (accounting, payroll, FEC, filings). Large installed base. Dense ecosystem of French providers. Good reputation in payroll and accounting.
Limits. Technical heterogeneity between products — moving from 100 to X3 = changing software, not moving up a range. UX often dated. Commercial strategy pushing migration to Intacct (designed and operated in the US, which raises a sovereignty issue). Rising costs. Proprietary DB stack (SQL Server, Oracle).
Sovereignty. Good on legacy products (UK, outside the Cloud Act). To be examined product by product for Intacct (US-designed) and for SaaS offerings (hosting to be verified).
Cegid
Profile. French vendor, strong in certain verticals: retail (Cegid Retail), accounting firms and payroll (Cegid Loop), HR (Cegid Talentsoft), tax management.
Strengths. Very deep verticals. Impeccable French compliance. Strong foothold in the French market. Mature SaaS.
Limits. Poorly suited outside its targeted verticals. Heterogeneous range resulting from acquisitions. Capital largely held by foreign investment funds (KKR in particular), to be factored into the analysis of the vendor's capital sovereignty. Proprietary DB stack (SQL Server). Rising license and SaaS costs.
Sovereignty. Good on the jurisdictional axis (FR), with attention to the evolving capital structure and to hosting (often Microsoft Azure).
Divalto
Profile. French vendor specializing in manufacturing, trading and distribution SMEs. Divalto Infinity (ERP) and Divalto Weavy (mobile CRM) range.
Strengths. Very good fit for French industrial SMEs. Technical trading, construction and distribution verticals. Vendor proximity. French capital independence. Reasonable pricing.
Limits. Modest brand recognition and ecosystem size compared with the big players. Limited international coverage. Ongoing migration of the legacy Harmony stack (proprietary 4GL) to .NET, creating a technical transition for legacy customers. Proprietary DB stack (SQL Server).
Sovereignty. Excellent on the vendor axis (FR, independent capital). Hosting most often controlled by the customer. Tech stack partly proprietary.
IFS Cloud
Profile. Swedish vendor, a leading vendor in asset-intensive and project-driven industries: energy, defense, aerospace, equipment maintenance, on-site services.
Strengths. EAM (Enterprise Asset Management) functional depth, complex project management, field service. Strong presence in European defense and aerospace. Swedish jurisdiction. Modern UX (since IFS Cloud).
Limits. High cost. Less dense partner ecosystem in France than SAP or Microsoft. Tech stack partly proprietary (Java + Oracle DB). IFS Cloud often deployed on Azure, which reintroduces a hosting question. Capital majority-held by investment funds (EQT and Hg).
Sovereignty. Good at the vendor level (Sweden). To be examined on hosting (Azure is common) and on the DB (proprietary Oracle).
The "sovereign cloud" trap
Be wary of offerings labeled "sovereign cloud" or "trusted cloud". Several American hyperscalers have set up joint ventures with French operators (Bleu for Microsoft Azure with Capgemini and Orange, S3NS for Google Cloud with Thales) aiming for ANSSI's SecNumCloud qualification. These arrangements improve protection against extraterritoriality, but the legal debate remains open on their ability to fully shield European operators from US injunctions as long as the underlying technology remains controlled by the US vendors.
The truly sovereign clouds in the strict sense remain OVHcloud, Scaleway, Outscale (3DS Outscale, SecNumCloud-qualified), NumSpot, as well as qualified French regional hosts.
Comparison table
| ERP | Cost (license / integration / maintenance) | Usability | Functional breadth | Functional depth | Sovereignty | Language / DB (licenses) | Data ownership |
|---|---|---|---|---|---|---|---|
| Odoo | € / €€ / moderate | Excellent, modern | Very broad (90+ modules, thousands of apps) | Variable, can be extended by development | Software: excellent (BE). Hosting: Online and .sh on GCP, on-premise = sovereign | Python (PSF, free) + JS/OWL / PostgreSQL (free) | Depends on mode: Google (Online and .sh); customer (on-premise) |
| Dolibarr | 0 / € / low | Decent, less modern | Good for very small businesses/SMEs | Limited in manufacturing and multi-country | Maximum (FR, GPLv3) | PHP (free) / MariaDB (GPL) or PostgreSQL (free) | Customer (free self-hosting) |
| Tryton | 0 / €€ / moderate | Austere | Good, very modular | Deep (accounting) | Excellent (European foundation, GPL) | Python (free) / PostgreSQL (free) | Customer |
| Axelor | € (Comm. 0, Ent. paid) / €€ / moderate | Good (low-code studio) | Broad | Average-good | Excellent Community (FR, AGPLv3) | Java (OpenJDK, free) / PostgreSQL (free) | Customer |
| ERPNext (Frappe) | 0 / € / low | Good | Broad | Average | Good (India, GPLv3) | Python (free) + JS / MariaDB (GPL) | Customer if self-hosted |
| SAP S/4HANA | €€€€€ / €€€€€ / 22% per year | Average (Fiori) | Maximum | Very deep (finance, SCM, manufacturing) | Partial (DE but NYSE and RISE on US hyperscalers) | ABAP (SAP proprietary) + Java / HANA (SAP proprietary) | Customer if on-prem; SAP, AWS, Azure or GCP if RISE |
| SAP Business One | €€€ / €€€ / 18-22% per year | Average | Good for SMEs | Average | Partial | C# / SQL Server (MS proprietary) or HANA | Customer or partner host |
| Oracle NetSuite | €€€€ / €€€€ / SaaS included | Dated | Very broad | Good (finance), weak (production) | None (US, Cloud Act, FISA) | SuiteScript (Oracle proprietary) / Oracle Database (proprietary) | Oracle (US) |
| Microsoft Dynamics 365 | €€€€ / €€€€ / SaaS included | Good (Office integration) | Very broad (F&O, BC, CE) | Variable by module | None (US, Azure, Cloud Act) | X++ (MS proprietary) + C# / Azure SQL (MS proprietary) | Microsoft Azure (US) |
| Infor CloudSuite | €€€€ / €€€€ / SaaS included | Good | Broad (industrial verticals) | Deep (discrete manufacturing, fashion) | None (US since Koch Industries) | Java + Mongoose / multi-DB, AWS cloud | Infor / AWS (US) |
| Sage (X3, 100, Intacct) | €€€ / €€€ / 18-22% per year | Variable, often dated | Good, fragmented across products | Good (FR accounting and payroll) | Good (UK, outside Cloud Act); Intacct US-designed | Sage 4GL, .NET depending on product / SQL Server or Oracle (proprietary) | Customer (on-prem) or Sage (cloud) |
| Cegid | €€€ / €€€ / variable | Variable by product | Strong on verticals (retail, accounting firms) | Strong on targeted verticals | Good (FR), foreign fund capital | .NET / SQL Server (MS proprietary) | Customer or Cegid (often Azure) |
| Divalto | €€ / €€ / moderate | Decent | Good (manufacturing and trading SMEs) | Good on its targets | Excellent (FR, independent capital) | Harmony (proprietary) then .NET / SQL Server (proprietary) | Customer most often |
| IFS Cloud | €€€€ / €€€€ / included | Good | Strong (EAM, project management, manufacturing) | Very deep (process and asset-intensive industry) | Good (Sweden); Azure common | Java / Oracle Database (proprietary) | Customer or IFS Cloud (often on Azure) |
Cost legend: € very low · €€ moderate · €€€ high · €€€€ very high · €€€€€ premium large enterprise.
Solutions that are sovereign at every level
For an ERP to be considered sovereign at every level, four conditions must be met cumulatively:
- Vendor in a non-extraterritorial jurisdiction (ideally European).
- Free, auditable and modifiable source code.
- Language and database under a free license.
- Controlled hosting: self-hosting or a sovereign cloud not tied to a provider subject to US law.
Solutions that tick all four boxes without reservation:
- Dolibarr (FR, GPLv3, free PHP and MariaDB/PostgreSQL, free self-hosting).
- Tryton (European foundation, GPLv3, free Python and PostgreSQL, free self-hosting).
- Axelor Community (FR, AGPLv3, free Java OpenJDK and PostgreSQL, free self-hosting).
- Odoo Community (BE, LGPLv3, free Python and PostgreSQL, free self-hosting).
- ERPNext (India, GPLv3, free Python and MariaDB, free self-hosting) — sovereign in the strict technical sense, provided you accept a vendor outside the EU.
Solutions that can be sovereign under conditions:
- Odoo Enterprise on-premise on a sovereign cloud (OVHcloud, Scaleway, Outscale, NumSpot) or own servers: sovereign in practice, but with a proprietary Enterprise license (source access, not redistributable). Odoo Online and Odoo.sh are not sovereign (GCP).
- Axelor Enterprise on sovereign infrastructure: same logic.
Solutions excluded from complete sovereignty:
- All American vendors: Oracle NetSuite, Microsoft Dynamics 365, Infor — excluded by the vendor's jurisdiction.
- SAP, IFS, Sage, Cegid, Divalto — vendors outside direct US jurisdiction (Sage is UK, therefore outside the Cloud Act), but with proprietary code and most often a proprietary tech stack (HANA, Oracle DB, SQL Server). Sovereignty is partial, never complete.
Conclusion
Sovereignty is a matter of accepted risk. Choosing an American vendor means accepting the extraterritorial risk in exchange for product maturity, a dense ecosystem and often smooth integration with other US tools. Choosing a European proprietary vendor limits the jurisdictional risk but retains strong economic and technical dependence on a single vendor, with the specific risks of the proprietary model (data captivity, license audits, imposed end of support, captive price increases). Choosing a European open source solution means getting full control of the software, the data and the hosting, in exchange for a higher initial investment in skills and integration.
No solution is universally good. The right ERP is the one that simultaneously matches the company's business, size, culture, investment capacity, existing ecosystem and level of sovereignty requirements.
The grid proposed here is not meant to push an answer, but to make the criteria of the choice objective. An organization can, with full awareness, choose SAP for its functional depth despite its partial sovereignty, or Microsoft Dynamics for its consistency with its Office environment despite the Cloud Act exposure. What matters is that the choice is conscious, documented and revisable — and that the risks of the proprietary model are factored into the decision rather than ignored.
Choosing your ERP means choosing your dependence. You might as well choose it with clear eyes.