HiveCode
Before the first customer
A power-cycle bug shows up on a simulated hang rather than a customer's drive, and the first report says plainly how far the device falls short of professional tools.
Coming SoonFree editionRaspberry Pi & hardware
Recovery Proof Lab
How the device would earn the words 'world class': a repeatable test bench that measures it against ddrescue, HDDSuperClone and, where possible, professional results, using a library of donated failing drives and simulated faults, and publishes the recovery percentages, time taken and any drives lost.
The problem
Almost every recovery product claims excellent results, and few publish evidence, so nobody can compare them. Failing drives are also unpredictable, so without careful testing a design can look good on one lucky example.
Features
- Repeatable faults using Linux device-mapper: bad sectors, slow zones, hangs and sudden disappearance, before any real drive is risked
- A growing set of real failing drives, recorded and imaged in the same order by each tool under test where possible
- A results table: recovered percentage, time, power cycles, and how many drives died
- Clear wording rules: no recovery claim is published until the data supports it
How it works
How it works
- device-mapper targets (flakey, delay, error) for controlled faults
- Scripted runs with hashes against the known original data
- Public report generated from the raw results
Who it is for
Benefits
- Evidence instead of claims
- Bugs found on test data, not customer drives
- Honest public results
Built to be trusted
- Results come from the raw data
- Failures are published too
- Drives are donated or bought, with no customer data
Good to know
Simulated faults don't behave like real failing heads, so they can only test logic and safety. Real failing drives are all different and can only be read to death once, so results can't be perfect comparisons. Getting a statistically useful set of failed drives will take a long time and money, and comparing with lab tools needs a lab's cooperation. Claude Code can write and debug the software, but a person has to build the hardware and prove it on real failing drives.
Under the hood
For technical readers
- Linux device-mapper fault targets
- Hash comparison, report generator
- Python service and dashboard written with Claude Code; Raspberry Pi OS Lite
Related products
Works well with
ReadyPro edition
Clone & Deep Rename
Virtualization
Cold clone with verified copy or instant ReFS block clone, and a reversible rename of VM, folder and disks.
Coming SoonIn developmentFree edition
Licensing Engine
Platform
One licensing engine for every HiveCode app, the two hundred that exist on paper and the ones still to come.
Coming SoonPro edition
App Inventory & Behaviour Report
System & diagnostics
Know what an app does on your PC: pick an installed Windows program and get one plain-language report of its files, startup entries, services, scheduled tasks and browser add-ons, the…
Coming Soon
Notify me when Recovery Proof Lab is ready
Recovery Proof Lab is planned for a later release. Leave your e-mail address and we will write once when it becomes available. We ask you to confirm your address first, and you can unsubscribe at any time.
Questions now? Contact our team.