Source-based product comparison
Posnic vs Open Source Point of Sale: desktop or web stack?
Posnic packages a desktop runtime around the till. OSPOS documents a browser, PHP/CodeIgniter and MySQL or MariaDB stack. Compare who can operate each dependency and recover it under the shop's real failure modes.
Method and evidence boundary
Reviewed 18 August 2026. Posnic facts are pinned to public stable-channel release v1.3.0 and source commit b531ef4308c4dc3a25f250551a54fc5616e3b8d9. OSPOS facts are pinned to official master commit 8bd2a51cb52abbb6bd0103211f47101ad5f16bc9, fetched on 18 August. Posnic did not build the source, deploy its database or assign it a runtime score.
Posnic did not install or run Open Source Point of Sale for this review. Official project statements, source inspection, Posnic runtime evidence and a buyer's physical site acceptance are different evidence levels. The buyer must run both selected builds against the same written deployment profile.
Compare operating models before feature lists
| Decision point | Posnic published boundary | Open Source Point of Sale official boundary | Buyer acceptance test |
|---|---|---|---|
| Application shape | Electron desktop package with a local API and MongoDB on the till. | The pinned README describes a web-based POS written in PHP, using CodeIgniter 4 and MySQL or MariaDB. | Draw every browser, web server, proxy, database, DNS, TLS and counter-network dependency. |
| Source and licence | Public v1.3.0 source under AGPL-3.0-only, pinned to the package evidence commit. | The pinned README describes MIT terms with an important addition; the exact license includes visible-footer and ownership conditions. | Retain the exact build/source match, dependency notices and accepted legal decision for modification, hosting and distribution. |
| Release installation | Published desktop packages have exact names, sizes and hashes plus disclosed signing limits. | The install guide says to use a release archive for normal installation; a repository clone requires a build process. | Preserve the actual archive, hash, source commit, build provenance, configuration and rollback package. |
| Production stack | The application starts its local runtime on the shop computer; security and update ownership still need assignment. | The pinned guide supports PHP 8.2 to 8.4, MySQL 5.7 or MariaDB 10.x and requires allowed-hostname configuration in production. | Harden and monitor the exact PHP, web, proxy, TLS, database, secrets, backup and host configuration. |
| Network and host failure | One cash sale completed while external hosts were blocked inside Electron; no LAN, OS or host-failure result is claimed. | Availability follows the chosen placement of browser, web server and database; the docs do not make every failure equivalent. | Inject WAN, LAN, web-host, database and browser-session failures separately and reconcile all records after recovery. |
| Docker boundary | Posnic publishes native packages rather than a public production container path in this evidence set. | The pinned OSPOS guide provides Docker setup but warns that the basic setup is not suited for production and default passwords must change. | Record the production deployment design, secrets, image versions, TLS, persistence, backups and update ownership. |
| Hardware | 35 Posnic protocol tests passed without physical devices. | The buyer's browser/server deployment still needs exact printer, scanner, drawer, scale and terminal compatibility evidence. | Run the same named physical equipment, fault and reconnect cases against both final deployments. |
| Restore and exit | One small synthetic restore is reproduced; full production recovery and exit remain unproved. | The operator owns MySQL/MariaDB backup, restore, schema migration and web-stack recovery for the selected deployment. | Restore into a clean environment, reconcile data, then export and import all required business-owned records. |
What the current official sources establish
Pinned web architecture
OSPOS README at 8bd2a51c identifies PHP, CodeIgniter 4 and MySQL or MariaDB.
Supported runtime range
Pinned installation guide lists PHP 8.2 to 8.4 and MySQL 5.7 or MariaDB 10.x.
Production host validation
The same guide marks allowed-hostname configuration as required for production and says the application fails to start when it is absent.
Release versus source clone
The guide directs ordinary installation to a release archive and states that a repository clone requires the separate source-build path.
Docker warning
Its basic Docker setup is documented for local use with an explicit warning that it is not suited to production and that default passwords must be changed before exposure.
Exact licence conditions
OSPOS licence at 8bd2a51c must be reviewed in full; this page does not reduce it to a generic MIT badge.
What Posnic v1.3.0 evidence establishes
| Area | Recorded evidence | Do not infer |
|---|---|---|
| Release and source | The public stable channel identifies v1.3.0, manifest maturity beta, exact package sizes and SHA-256 values. Source is pinned to b531ef43 under AGPL-3.0-only. | Windows and Linux packages are unsigned; macOS packages are signed but not notarized. Public source and a checksum do not establish an independent audit. |
| Local sale | One Windows x64 cash sale, S-O2MA-000001, completed for INR 125 and reopened while external hosts were blocked inside Electron. | This was not an operating-system disconnect, router failure, payment-terminal run, power-loss test or production shift. |
| Backup and restore | One synthetic restore completed with 20 collections and 53 documents. | It did not establish a production schedule, off-site retention, representative database size, replacement computer, RPO or RTO. |
| Source tests | The curated API Jest run recorded 7,953 passing, 0 failing and 13 skipped tests. | It was not a hosted end-to-end run, competitor test, customer acceptance, penetration test or independent security review. |
| Hardware protocols | 35 printer-layout, receipt, report and scale-parser protocol tests passed with no physical equipment connected. | No named printer, scanner, drawer, scale, customer display or payment terminal has a public physical acceptance result in this evidence set. |
| Payments and compliance | The recorded local transaction used cash and did not connect a payment terminal. | No card authorization, capture, settlement, PCI validation, country-specific fiscal acceptance, tax approval or customer outcome is established. |
When each operating model belongs on the shortlist
Shortlist Posnic when
The business wants a packaged desktop runtime on each till, can own local computer/database recovery and accepts the current installer, hardware, platform and independent-security evidence limits.
Shortlist Open Source Point of Sale when
The technical team wants a browser-based PHP and MySQL/MariaDB POS, can secure and maintain the complete web/database stack and accepts the exact source, licence, release and production configuration path.
Stop the selection when
No one owns exact-version licensing, package provenance, physical hardware, payment/fiscal acceptance, backup restore, security updates, support escalation and usable data exit. Source availability cannot replace those decisions.
Run the same acceptance work on both products
| Test | Required result | Evidence to retain |
|---|---|---|
| Package or build provenance | The installed Posnic package and deployed OSPOS release each trace to retained source, licence and integrity evidence. | Release/archive, hash, commit, build record, dependency notices and legal decision. |
| First sale and exception | Representative staff complete/retrieve a sale, then run return, void, discount and denied-role cases with correct records. | Transaction IDs, receipts, roles, stock effects, audit entries and defects. |
| Dependency failures | Required tasks have clear results under WAN, LAN, browser, web-host, local-service and database failures, followed by clean reconciliation. | Failure injected, timing, errors, queued/local records, duplicates, gaps and recovery time. |
| Physical counter | Exact printers, scanners, drawers, scales, displays and terminals pass normal, fault and reconnect cases. | Models, firmware, drivers, browser/runtime path, samples, repetitions and failures. |
| Security and update | Named owners can patch the desktop package or full web stack, rotate secrets, preserve data and roll back safely. | Supported versions, dependency report, configuration, update rehearsal, rollback and contacts. |
| Restore and exit | A separate clean environment restores within target and every required data export reconciles and imports elsewhere. | Backup/hash, environment, RPO/RTO, row totals, field map, import result and exit owner. |
Primary sources and snapshot
Posnic release and source
OSPOS architecture
OSPOS installation
OSPOS licence
Source snapshot
Official master resolved to 8bd2a51cb52abbb6bd0103211f47101ad5f16bc9 on 18 August 2026; moving branch links are not used as evidence.
Acceptance method
Twenty-four-test benchmark and research and correction policy.
Continue the comparison
Open Source Point of Sale comparison questions
What is the main architecture difference between Posnic and OSPOS?
Posnic v1.3.0 packages Electron, a local API and MongoDB for the till. The pinned OSPOS source describes a web application written in PHP with CodeIgniter 4 and MySQL or MariaDB. The web server, database, browser network and production configuration are therefore part of the OSPOS operating model.
Does OSPOS use a standard MIT license without extra conditions?
The pinned OSPOS README says MIT terms have an important addition, and the license file contains visible-footer and ownership conditions. Read the exact selected version's license and obtain an accepted legal position instead of relying on a short label.
Will either product keep every task working during an outage?
This page does not establish that. Posnic reproduced one local cash sale under Electron-level external-host isolation. OSPOS behavior depends on where the browser, web server and database run and which dependency fails. Inject WAN, LAN, host and database failures separately.
Did Posnic install and benchmark OSPOS?
No. OSPOS facts come from its official source at pinned master commit 8bd2a51c. Posnic evidence covers Posnic only. A buyer must deploy the selected OSPOS release and retain its own runtime, hardware, restore and security results.
How should a shop compare Posnic and OSPOS fairly?
Use the same representative items, users, sales, returns, stock events, devices, outage injections, backup restore, security review and exports. Include the OSPOS web/database operations and the Posnic desktop/package limitations in the decision.