MapGIS: modernize, integrate or migrate with evidence

Sources reviewed: July 17, 2026. GeoSAT is not affiliated with H&G Consultores or MapGIS. MapGIS belongs to its respective owner. This analysis separates published facts, vendor claims and technical inferences; it does not rate an installation that has not been inspected.

The short conclusion is that there is no single MapGIS architecture. An organization may retain a historical version with accumulated dependencies, operate MapGIS9 or be considering MapGIS10. The product name alone does not reveal the installed runtime, where data resides, which services it consumes, who can modify the code or how much it costs to evolve.

It would therefore be inaccurate to dismiss all of MapGIS because of the age of a historical version. A more useful conclusion is supportable: a model in which the map viewer concentrates data, integrations and processes belongs to an earlier generation of architecture. Before making new investments, an organization should establish whether its implementation still follows that model or has evolved toward decoupled data, APIs and components.

GeoSAT provides a MapGIS modernization assessment to answer that question before renewing, upgrading or migrating.

What “legacy” means in this analysis

Legacy does not determine utility, security or commercial status on its own. It describes an architectural and operational condition: changing one component is expensive because the system depends on accumulated technologies, contracts, code or knowledge that affect other layers.

A system may continue to provide a critical service and be legacy at the same time. A long-running system may also avoid that condition when its data is portable, interfaces are documented, components remain supported and different teams can operate it.

The test is not the purchase date. The test is how much control the organization retains over:

Data, history, metadata and permissions.
APIs, map services and consumers.
Business rules and integrations.
Source code, deployments and tests.
Upgrades, support and recovery.
The ability to engage a different team.
Total cost over the next several years.

2009–2024: the history documented by Medellín

A procurement document published by Colombia Compra Eficiente records the District of Medellín’s first MapGIS purchase as a MapGIS Flex 3.0 corporate license dated September 16, 2009.

The same document concerns a contract that began in March 2024 to subscribe to, support, maintain and upgrade MapGIS 9.0. It therefore documents technology continuity, not a platform frozen at Flex 3: for that procurement, Medellín described a Server/Java suite with an HTML5 and CSS3 front end integrated with ArcGIS.

The document also says the District held a perpetual license for unlimited users. It states that mission applications belonged to the District and could be modified by third parties as long as those parties did not use MapGIS platform-specific source code.

This source establishes a specific contractual and technology history in Medellín. It does not establish the version installed today, nor that other organizations have the same licensing, intellectual-property or maintenance terms. It also does not support using the Flex 3.0 reference to describe MapGIS 9.0 or MapGIS 10.

The code distinction does suggest a verifiable renewal question: which components can a third party maintain with the organization’s deliverables, and which require reserved platform code or knowledge? The answer must come from the contract, repositories and a technical test rather than a generic allegation of dependency.

MapGIS5: a historical listing hosted in Esri’s directory

The MapGIS5 listing in Esri’s partner directory describes a Java and Adobe Flex RIA web application using ArcGIS Server map services and data engines such as Oracle or SQL Server.

The listing remains available but shows no visible publication date. It should be read as a description of MapGIS5, not as documentation for MapGIS9, MapGIS10 or every active installation.

Adobe’s official product support matrix records the end of extended support for Adobe Flex 3 in February 2015. That does not prove that a current MapGIS installation uses the component. It means that if an inventory finds a real Flex 3 dependency, a documented path should exist to remove or isolate it.

Java does not establish an application's current technical condition either. It may run on current releases, receive patches and have a maintainable architecture. The exact runtime version, dependencies, support lifecycle and ability to deploy and test changes are what matter.

2024 and 2025: MapGIS9 remained in institutional deployments

The Municipality of Itagüí’s 2024 Circular 317 said its SIGMI Geoportal was based on MapGIS suite version 9.0. The source contradicts any claim that the product had stopped being used before that point.

Cornare’s first-half 2025 management report documented the development, configuration and integration of MapGIS modules within a new service architecture. It estimated that the new MapGIS9 version and its modules would be operational by the end of August 2025.

The wording matters: the first-half report describes work in progress and an estimated date. On its own, it does not prove that the full scope was completed by that date.

The Cornare intranet, updated May 6, 2026, separately identified an “Internal MapGIS 8.0 Geoportal” and a “(New) Internal MapGIS 9.0 Geoportal.” This confirms continuity and the public incorporation of V9, but it does not quantify savings or establish on its own that every planned module was complete.

2026: MapGIS10 changes the discussion

In an H&G publication reviewed in July 2026, the vendor presented MapGIS V10 as a move from the map at the center toward an ecosystem of data, processes and decision-making. The publication mentions integrated AI, multiple maps, split-screen analysis, performance improvements and integration with the latest ArcGIS SDK.

The current MapGIS suite page also claims web and mobile access, customization, compatibility with multiple platforms and interoperability between organizational systems.

Esri also published a Winter 2026 ArcNews case about MapGIS-POT in Medellín integrated with ArcGIS Enterprise, ArcGIS Pro, ArcGIS Online and ArcGIS Dashboards.

These publications show commercial activity and announced evolution in 2026. They do not independently verify performance, portability, cost, general availability of each capability or the effort required to upgrade a specific installation.

The V10 announcement strengthens the correct question: does the organization already operate the new architecture or retain an earlier deployment? A vendor announcement does not mean an installed system automatically gained those capabilities, nor that upgrading every existing application is free or transparent.

Cornare: hybrid modernization, not complete replacement

The Cornare case requires particular precision because it combines several layers.

In its 2024 management report, the organization documented databases migrated to PostgreSQL and reported a move from ArcGIS to QGIS. The contractual scope also included changes to MapGIS Data and a MapGIS CMS platform. The 2025 report described MapGIS9-module integration into the new service architecture, and the 2026 intranet published links to both the 8.0 and 9.0 environments.

The supportable reading is:

Cornare modernized its database and reported a move from ArcGIS to QGIS.
It retained MapGIS and was integrating MapGIS9 modules; by 2026 it published links to both environments.
The source does not publish realized before-and-after financial savings.
The contract combined support, maintenance, evolution and integration; it is not a universal migration price.

Cornare therefore does not show that every organization should replace MapGIS. It shows that a modernization can move ArcGIS to QGIS and databases to PostgreSQL while retaining modules that still add value. Read the full Cornare case analysis.

The architectural shift that should be assessed

A viewer-centered architecture tends to concentrate integrations, rules and operational expertise around the presentation platform. A data-and-process-centered architecture aims to make the viewer a replaceable consumer.

Installing PostGIS or publishing a WMS is not enough to declare a system open. Critical assets must operate through clear contracts between layers.

Data

Documented model and dictionary.
Authorized direct access without a specific interface.
Export of geometry, attributes, history, attachments and relationships.
Tested roles, audit, backup and recovery.

Services and integrations

Documented REST or OGC APIs.
Inventoried authentication, limits, versioning and consumers.
Business rules outside components that cannot be tested independently.
Ability to replace one consumer without migrating the entire database.

Applications

Identified code, dependencies and deployment process.
Regression tests for mission workflows.
Interfaces that do not hold the only copy of a rule or dataset.
Accessibility and mobile experience verified with real users.

Operations

Components within support periods.
Monitoring, alerts, SLA and accountable owners.
Tested restoration, not only declared backups.
Documentation sufficient to transfer operations.

Eight questions before renewing

Which exact MapGIS version and modules are in production?
Which Java, .NET, ArcGIS SDK, server, operating-system and database versions do they use?
Which data and rules can operate without the MapGIS interface?
Which services are documented, and which applications consume them?
Which code and repositories were delivered to the organization?
What can a third party maintain under the current contract?
What does an upgrade actually fix, and which applications must be adapted?
What is the five-year TCO of retaining, upgrading, decoupling or migrating?

Without these answers, a renewal repeats the current state without identifying the risk or cost being purchased. An immediate migration would be equally irresponsible: it could remove a critical capability before an accepted replacement exists.

Four paths, not one conclusion

1. Retain and document

If the implementation meets functional requirements, uses supported components and has a reasonable cost, retaining it may be the best decision. Investment can focus on documentation, tests, recovery and reducing tacit knowledge.

2. Upgrade with the vendor

A new version may remove historical dependencies and improve the experience. The assessment should separate licensing, infrastructure, module adaptation, data migration, testing and support. It should also compare promises with a pilot using the organization’s real workflows.

3. Decouple by layer

Data can move to PostgreSQL/PostGIS, services to APIs or OGC interfaces and selected rules to documented components while the organization temporarily retains the viewer or specialist modules. This limits risk and allows operational cost to be measured before expanding the change.

4. Migrate after validation

QGIS, GeoServer, QGIS Server, MapLibre, OpenLayers or a custom application can replace components once replacement workflows have passed acceptance criteria. An open-source label does not guarantee a maintainable architecture; support, security, infrastructure and evolution must still be funded.

What the public sources do not prove

The cited sources do not establish:

Which version every MapGIS customer uses today.
That MapGIS10 retains the MapGIS5 architecture.
That an installation is insecure or slow.
That data is locked in or completely portable.
That its cost is higher or lower than an alternative.
That upgrading is preferable to migrating.
That Cornare removed every MapGIS module.

Those answers require technical access, contracts, invoices, tests and interviews with accountable owners. Any conclusion before that work is marketing, not architecture.

Next step

For new investment, assess before renewing. GeoSAT’s MapGIS modernization assessment delivers a technical inventory, dependency map, portability matrix, total-cost scenarios and a path with acceptance criteria.

The conclusion may be to retain MapGIS, upgrade it, integrate it with PostGIS and APIs, or migrate components. The evidence decides.