FiveTech Support Forums

Dear All ,

Just out of curiosity, does anyone know roughly how many Harbour or xHarbour developers there are around the world?

Hi ,

Thanks for pointing this topic!

It seems that it very low I was expecting at least 20K around the world. And looking at next generation developers they learn python, java & Javascript / HTML.

 I believe that the Harbour VM should be accepted as similar Java VM. But need to know what is major difference between Harbour VM & Java VM.

Thanks

Shridhar

CHATGPT

Question: How many DBF-based applications are still out there?

Answer (short and candid): Nobody has hard numbers—DBF often runs “under the radar.” But based on industry experience, forum/association activity, and migration projects, a plausible range looks like this:

Order of magnitude (worldwide, 2025)

• Actively used core line-of-business systems (daily use):

50,000–150,000 installations

(typical: ERP/retail, POS, logistics, manufacturing, municipal, hospitality/food service, trades)

• Still in operation but only sporadically updated (maintenance mode):

100,000–300,000 installations

(stable, few releases, yet productive)

• “Embedded DBF” or tools/subsystems that use DBF internally:

hard to count—additional tens of thousands to hundreds of thousands of instances

Why the wide range?

• Invisibility: Many systems are in-house/on-prem with no public footprint.

• Regional clusters: LATAM, India, Eastern Europe/DE/AT/IT show disproportionately many legacy installs.

• Application longevity: DBF solutions often run 10–20+ years as long as hardware/OS permit.

• Migration path: Many modernize the frontend (Web/REST) while keeping DBF as a stable data hub for now.

DEEPSEEK

Of course. Based on the provided forum thread and general industry knowledge, here is a consolidated answer addressing all your questions.

1. How many Harbour/xHarbour Developers are there around the World?

This is the central question from the forum, and the consensus, including the user shri_fwh's own realization, is that the number is significantly lower than 20,000.

Evidence from the Thread: The user Maurizio linked to a Harbour Users Google Group. The activity level on such groups, forums, and other community hubs is a strong indicator of a smaller, niche community. If there were 20,000 active developers, these forums would be far more active.

Realistic Estimate: It is impossible to get a precise number, but a reasonable estimate would be in the low thousands globally for active developers, and perhaps tens of thousands who still maintain legacy systems but are no longer doing new development.

Reasoning: As shri_fwh correctly observes, new developers are learning modern languages like Python, Java, and JavaScript. The Harbour/xHarbour community is primarily composed of developers who started with Clipper/dBASE in the 80s and 90s and have ported their applications forward.

1. How many applications do you think still exist on a DBF basis?

This number is extremely high. While new commercial software is rarely built on DBF files today, the number of legacy systems still in operation is vast.

"Invisible" Legacy Systems: Countless small to medium-sized businesses worldwide still run critical, internal applications built with Clipper, FoxPro, Harbour, or xHarbour that rely on DBF files. These are often accounting, inventory, manufacturing, and custom database systems that "just work" and are too critical or expensive to replace.

Data Interchange Format: The DBF format is also still used as a common export/import format for data exchange between various systems because of its simple, tabular structure.

Conclusion: It's safe to say there are millions of applications and systems still using DBF files in some capacity around the world, the majority of which are legacy systems maintained by the Harbour/xHarbour developer community.

Dear Otto ,

Thanks for big informative reply..!!

It seems that no new Projects on xBase Technology.



  I believe that the Harbour VM should be scalable and provide rock solid REST API using embedded load balancing http server.  And GUI side the Harbour Server Pages using Javascript HTML GUI. Then it becomes the Harbour Fusion Solution all in one.

Thanks

Shridhar

Hello Shridhar,

Pragmatic “all-in-one” setup — would this meet your “Fusion” goal?

We’ve had good results with a single Windows VM (Virtual Machine) running:

-Apache (Reverse Proxy + Static Files + Logs)

-Harbour microservice — long-running REST (Representational State Transfer) API on 127.0.0.1:9000 that keeps DBF (dBASE-style files) and indexes warm

-cloudflared — Zero-Trust tunnel providing TLS (Transport Layer Security), optional WAF (Web Application Firewall) and rate-limits

Why this hits your objectives:

-Scalable enough for SMB LOB apps — here SMB = small and mid-sized businesses (not „Server Message Block“), LOB = line-of-business — without the complexity of embedded LB (Load Balancing).

-Rock-solid REST API: the service stays warm (no per-request VM spin-up), short lock windows, optional WAL (Write-Ahead Logging)/Audit.

-GUI path: serve HTML/JavaScript from Apache; the browser calls /api/.... Use SSR (Server-Side Rendering)/HSP (Harbour Server Pages) only where it truly helps (e.g., login, emails, PDFs).

Ops-simple: one VM to back up, one set of logs, fast DR (Disaster Recovery) via VM snapshots.

-Side note: cloudflared doesn’t get much “echo” in our circles, but in practice it’s a game changer: secure exposure to the internet, DNS (Domain Name System) routing, certificates, and rate-limits with minimal admin effort.

For EU (European Union) SMEs (small and medium-sized enterprises), data sovereignty often beats convenience. A single VM (virtual machine) with Apache → Harbour microservice (127.0.0.1 = local loopback, port 9000) and, optionally, a Zero-Trust tunnel gives us control over storage location, retention, and access—the things GDPR (General Data Protection Regulation) actually cares about.

We keep logs minimal (IP = Internet Protocol addresses anonymized, short retention), encrypt at rest (stored data) and in transit (network traffic), and back up only within the EU. If using a US-based tunnel, we add a DPIA (Data Protection Impact Assessment), SCCs (Standard Contractual Clauses), and technical measures (pseudonymization, header minimization).

Net result: simple ops (operations), fast DR (disaster recovery), and real compliance—without turning hosting into “rocket science.”

For privacy-compliant AI use, run a self-hosted AI gateway on your VM that anonymizes prompts before forwarding them to ChatGPT and injects context from your private knowledge base (RAG), so only a minimized payload ever leaves your stack.

Question: Would this VM + Apache + Harbour microservice (+ cloudflared) approach satisfy what you envision as the “Harbour Fusion Solution”? If not, which concrete capability do you still miss (e.g., built-in load balancer, templating, session model, etc.)?

May I ask what kind of software you work on?

Best regards,

Otto

shri_fwh wrote: Dear Otto ,

Thanks for big informative reply..!!

It seems that no new Projects on xBase Technology.

I believe that the Harbour VM should be scalable and provide rock solid REST API using embedded load balancing http server. And GUI side the Harbour Server Pages using Javascript HTML GUI. Then it becomes the Harbour Fusion Solution all in one.

Thanks
Shridhar

Dear Otto , Bayron ,

I’ve read several posts about developing web applications using mod_harbour, and it’s great to see the progress being made with Harbour + Apache HTTP.

Looking at broader frameworks such as Java Spring, we can see how they support smart microservices, a JDBC framework for database connectivity, and robust user authentication mechanisms. These features make it possible to develop scalable, high-performance web applications such as e-commerce platforms, core banking systems, and payment solutions.

To make Harbour more powerful and relevant for the next generation of developers, we should consider introducing new features inspired by these modern frameworks. By adopting similar architectural patterns and enhancing Harbour’s capabilities, it can evolve into a competitive platform for building large-scale, high-performance web applications.

While converting or translating XBase commands into another language may help to some extent, true progress will come from innovating and extending Harbour’s core features to meet modern development needs.

Antonio Linares wrote: Dear Shridhar,

This solution is much easier and powerfull than mod_harbour:

https://runnerxbase.app/

Is there a description of RunnerXBASE in English? (the installation package is in Spanish only)

Natter wrote: Is there a description of RunnerXBASE in English? (the installation package is in Spanish only)