Skip to main content

Clinic Browser Compatibility: Test Real Tasks Before Keeping an Older Computer

Judge an older clinic computer by its supported operating system and browser, the vendor's requirements and evidence that the actual clinic tasks work. A successful login is not a compatibility test. Browser features such as Web Locks can coordinate local activity, but they do not replace server safeguards for two desks, uncertain writes or duplicate patient actions.

By Bractis · Updated · 8 min read

Indian clinic owners and reception leads deciding whether existing computers can remain in service during a software rollout or browser upgrade.

Record the device before deciding that it is too old or good enough

Start with the operating system, browser name and version, device architecture, available storage, relevant printer or scanner and the clinic software's stated support requirements. 'Old desktop' does not identify the reason a task fails.

Separate three questions: can the software execute, is the environment supported by its suppliers and does the intended workflow work reliably in the clinic? A page can load even when one of the other answers is no.

Google's Chrome requirements, checked on 25 October 2026, identify Chrome 109 as the last release supporting Windows 7 and 8/8.1. This is a specific vendor example, not a rule that every newer-looking browser installation is supported.

The same Chrome page distinguishes browser and operating-system timelines. A browser still accepting an operating system does not establish that the operating system has the security maintenance appropriate to the clinic's situation. Confirm the actual vendor and support arrangement.

Ask the clinic software provider to state the supported combinations and necessary peripherals in writing. If an unsupported combination happens to work in a demonstration, record it as unconfirmed or outside supported scope rather than silently treating the demonstration as a contract.

Source: Google Chrome Enterprise: Browser system requirements

What Web Locks can and cannot tell a clinic

Web Locks is a browser API that can coordinate work among relevant tabs or workers within its origin and browser-context scope. MDN documents its secure-context requirement and local coordination model. The feature name is not a claim that a clinic record is locked everywhere.

For example, an application may use a local lock to avoid two tabs starting the same browser-side synchronisation task. Another reception computer, a separate browser environment or an independently initiated server request is not protected merely because that first tab holds a lock.

The server still needs the appropriate rules for record versions, operation identity and concurrent changes. A browser capability check cannot decide whether two requests refer to one intended visit or two legitimate episodes.

If the API is unavailable, the product should use a documented supported alternative or explain that the affected action is unavailable. A fallback that simply removes coordination without preserving the operation's required correctness is not equivalent support.

Reception staff do not need to edit browser flags or disable security checks to make a clinical workflow appear to work. Ask the provider which behaviour is supported and what evidence verifies the fallback, including recovery after an interrupted request.

Keep feature absence distinct from a slow network or a server failure. Reinstalling the browser will not necessarily resolve a duplicate-write problem, just as a server lock cannot make an unreadable printer driver produce the correct document.

Source: MDN: Web Locks API

Test the tasks that will be performed, including their outputs

Build a compact acceptance matrix using the clinic's actual roles. Include finding the correct patient, creating a visit, saving or reviewing the relevant record, producing a document and confirming a billing or payment state. Add the peripheral-dependent work the clinic relies on.

Use synthetic data and record expected outcomes before the trial. The evidence should include the resulting stored record or output, not just a click that appears responsive. A spinner disappearing is not proof that the operation completed.

Test two tabs and two authorised devices where the workflow supports them. Repeat an action after a lost response through the configured recovery process and verify that the stored visit or payment is not duplicated.

Include logout and return, session expiry, a browser restart after saved work and a controlled interruption while a draft is unsaved. Distinguish what the product preserves from what it merely leaves visible in the original tab.

Test printed and downloaded documents on the actual intended device. Fonts, page breaks and printer settings can fail independently of the main application. Retain synthetic samples with the tested environment and date.

Make blocked operations explicit. If a required function is not supported, a clear limitation is preferable to silently recording an unverified success. The deployment decision must account for the effect on the clinic's intended service.

On a narrow screen, scroll the table horizontally to read all columns.

Browser acceptance matrix: inspect the result, not only the screen
Task familyEvidence to retainFailure worth distinguishing
Patient and visitCorrect synthetic identity and one intended visitWrong record, duplicate or unresolved submission
Clinical workStored content, authorship and versionLost draft, stale overwrite or unconfirmed save
DocumentsReadable actual print or downloaded fileMissing script, wrong page or older content
Financial actionCorrect bill and verified outcomeUncertain payment or repeated collection
Recovery and accessHonest return state and appropriate roleReopened stale task or another user's session
Unsupported capabilityDocumented limitation or tested alternativeSecurity bypass presented as compatibility

A percentage pass does not approve the workstation

In a fictional trial, four workstation/browser combinations each attempt five task groups, creating twenty matrix cells. Their passed results are five, five, four and three, for seventeen passed cells overall.

The other three cells consist of two clearly unavailable required tasks and one failed draft-recovery test. Seventeen of twenty is 85%, but averaging across computers does not make the weakest workstation ready for its assigned role.

The unavailable cells and the failed recovery cell also mean different things. An honestly blocked function may prevent bad data but still prevent the receptionist from doing the job. A silent draft loss can be unacceptable even when all other tasks passed.

Assign each device only to a scope whose dependencies have been verified and are appropriately supported. A computer suitable for a public information display may not be suitable for entering clinical notes; that distinction should not expose private records on the display.

After an upgrade or configuration correction, repeat the affected tasks and relevant boundaries. Preserve the prior finding and the new evidence. Do not turn an unresolved cell green because the browser version number changed.

This is an illustrative matrix, not a minimum certification suite. The clinic and supplier should agree the scope needed for their actual features, devices and consequences.

Upgrade deliberately and keep a usable route back to work

Arrange changes outside an active consultation where possible, with saved work verified and the clinic's appropriate backup or recovery process in place. A browser update can close sessions, change printing behaviour or require a new sign-in.

Prepare the information needed for support: device and browser version, synthetic reproduction steps, time of failure, expected outcome and actual stored result. Do not send a full patient chart or credentials simply to demonstrate a compatibility issue.

Keep a supported spare or an agreed alternative where the service depends on one workstation. An unsupported old browser is not a reliable long-term fallback merely because staff know how to use it.

Review the acceptance record after a substantial software, browser or peripheral change. The purpose is targeted verification of affected work, not repeatedly testing everything after a text-label update.

If the provider cannot explain whether a workaround preserves correctness, narrow the affected service or use the agreed alternative while the issue is resolved. Convenience should not turn an uncertain write into a presumed completed clinical or financial action.

Seventeen passed cells can still leave a required desk unusable

The fictional four-by-five matrix has seventeen passes, two unavailable required tasks and one failed draft recovery. Its 85% aggregate is descriptive only. The clinic must decide what each specific workstation can safely and reliably be used for under the actual supported scope.

Build a clinic browser acceptance matrix

Use these prompts in your clinic’s own notes. Record the result, responsible person and next action for each item.

Inventory support
Record the OS, browser, peripherals and the separate supplier support statements.
Define task evidence
Choose the stored record or output that proves each required task succeeded.
Test coordination
Exercise relevant tabs, devices and interrupted responses without assuming a local browser lock protects every desk.
Decide by consequence
Resolve each required blocked or failed task before approving the workstation's assigned scope.

Questions doctors ask

Does a working login prove a clinic computer is compatible?

No. Test the required records, documents, financial actions, peripherals and recovery, and confirm the suppliers' support requirements.

Does Web Locks prevent two reception computers creating duplicates?

Not by itself. It coordinates within its browser scope; cross-device correctness needs appropriate server-side rules and operation identity.

What should happen if a required browser API is unavailable?

The provider should document a supported, verified alternative or clearly limit the action. Removing the safeguard silently is not equivalent support.

Is browser support the same as operating-system support?

No. Check both timelines and the actual support arrangement, as well as the clinic software vendor's requirements.

Is an 85% acceptance-test pass rate enough?

A percentage alone cannot decide that. Inspect which tasks failed or were unavailable and whether the intended clinic service depends on them.

Sources and scope

Updated on . Original compatibility and acceptance model with a fictional matrix. Vendor documentation was checked on the stated date; no Bractis browser-support certification or universal hardware recommendation is implied. No independent clinical or legal review is claimed.

Try the workflow
with Bractis.

Ask Bractis to state its supported browser environments and demonstrate the required tasks on your intended devices, using synthetic records. Include an interrupted write and a second desk, then inspect the stored outcomes.

Request a clinic demo