BBF develops software around a simple idea: preserve useful structure and reuse it when the conditions still apply. Each technology does a particular part of that work.
Start with the job you need done: store files, run recurring tasks, update an application, build a world or produce a view. Each card explains who it is for and what you can try or discuss today.
Byte avoidance measures how many bytes can be avoided for a defined job: stored, transferred or expanded, compared with a stated baseline. BBF reuses useful structure and reconstructs what is requested, retaining the information needed for the result.
GLYX · LOCAL FILE ARCHIVING
Keep related files. Restore the one you choose.
Related reports can share information. GLYX retains reusable structure and what differs, then reconstructs the exact selected original when you recover it.
01 · KEEP USEFUL INFORMATION
+
Reusable structure + file differences
The information required to recover each original stays available in the archive and its recovery dependencies.
02 · CHOOSE WHAT YOU NEED
✓
Recover Report B
Select one original file. You can leave the other files archived.
03 · RECONSTRUCT THE RESULT
The exact selected file
Recover the original bytes and check them. Keep the archive, the GLYX Recovery app and matching recovery key safe.
The saving comes from reusing what is already available and avoiding unnecessary copies or expansion. New or unrelated information still needs to be represented.
An illustrated overview of the idea. Explore each technology’s demonstrations and measured results below.
Developed by founder Ben Vivian, the BBF family connects structured representation in Glyphtronix with recoverable files, retained application state, interactive worlds and rendering. Each product applies the wider approach to a particular job.
Reconstruct selected content or working information from its compact representation when it is needed. In GLYX, choose the exact original files to recover. Expanding Worlds applies the idea to world state, HoloRenderer to scene or image information, and Seed State OS to application state.
Why does that matter?
A growing archive can reuse information across related files and versions. A world or application can keep a compact description and reconstruct the relevant state when needed. The potential benefits are lower storage, less repeated work and less unnecessary expansion; the evidence for each product shows what has actually been measured.
ONE THREE-PRODUCT WORKFLOW
A collection changes. What happens next?
A team keeps project revisions, updates its working data and runs recurring tasks. GLYX addresses recoverable file history. Seed State addresses the working state affected by changes. TokenHoloFold handles supported structured tasks locally.
Choose the products that suit the job. Each can be discussed separately; a business evaluation can establish the value of a combined workflow on your own data and platform.
KEEPRecoverable records
GLYX Archive
UPDATEWorking state ready to use
Seed State OS
EXECUTEA completed, checked task
TokenHoloFold
THE EIGHT TECHNOLOGY FAMILIES
One role at a time. A clearer view of the whole.
Eight UK patent applications are pending. Explore each technology’s current stage and the options for working with BBF.
Free Mac trial available
GLYX Archive
Mac users
Creators
Research teams
Compression that builds on what you’ve saved. More reusable structure can make later matching additions far smaller—giving you more room for files and versions, with exact recovery.
Keep files compact and recoverable.
A familiar problem
A growing collection of related files or versions.
Its job
Different files can share layouts, repeated sections and unchanged content. GLYX reuses supported structure and retains the additional information needed to recover each exact original.
What to evaluate
Compare the space needed to keep and restore your files, check that every original comes back exactly, and measure how long it takes.
Everyday example
A folder of monthly reports can contain repeated layouts and unchanged sections. Putting compatible versions in the same archive gives GLYX the opportunity to reuse that information; you can still restore the exact original report.
Related records and successive versions can create further reuse opportunities as the archive grows. The benefit depends on the data that actually matches.
Reuse applicable working state so supported operations can focus on the changes that matter.
Reconstruct the state needed for the work.
A familiar problem
An application needs to update retained state after a change.
Its job
Checks which retained state is applicable and reconstructs the parts required by a supported operation.
What to evaluate
Integration work establishes correctness and complete cost on the intended platform.
Everyday example
For a repeated operation, ask whether a saved, verified state can supply the unchanged parts and only the affected state needs reconstruction. The recorded studies show results for their particular CPU, GPU or FPGA operations.
Complete supported structured operations locally and return a checked result to the wider AI workflow.
Complete supported operations locally.
A familiar problem
An AI workflow repeatedly performs an eligible structured task.
Its job
Executes the supported operation locally with result checking and a defined refusal or fallback boundary.
What to evaluate
Compare with your best practical existing implementation on equivalent completed work.
Everyday example
For a recurring supported operation in an AI workflow, compare doing the work locally with making the existing remote request. Check the returned result, total time and resource use together.
Create varied animated creatures and linked or nested environments from compact seed data, reconstructing the detail needed for play or inspection.
Retain the rules and state a world needs.
A familiar problem
An interactive world grows and changes over time.
Its job
Develops compact, reconstructible world representations, with selected state materialised when needed.
What to evaluate
A development programme: retained state, runtime resources and interactive behaviour need application-specific evaluation.
Everyday example
In a game or simulation, a visitor moves between places and changes objects. A compact world description and saved changes can be used to reconstruct the selected state; the demonstrations show this development direction.
Carry useful scene information forward and reconstruct the view or supported image regions needed after a change.
Update the affected parts of an image.
A familiar problem
A scene changes while much of its structure remains useful.
Its job
Develops selective scene and image reconstruction using applicable retained information and checked dependencies.
What to evaluate
The useful reduction depends on the scene, changes and complete rendering path.
Everyday example
When a view changes but parts of a scene remain useful, examine which information can be retained and which must be reconstructed. Judge the finished image and the complete rendering cost together.
Connect compatible language tools through a shared representation of supported work, its relationships and its sources.
Different language tools. A common representation.
A familiar problem
Software tools need to exchange structured work while retaining its identity, relationships and context.
Its job
Provides a structured bridge through declared language and receiving-system interfaces.
What to evaluate
Specific applications and interfaces are evaluated separately.
Everyday example
A record needs more than a string of symbols: a system also needs to know which object it describes, how it relates to other records and when it applies. Glyphtronix makes those relationships part of the representation.
Check the required outputs and the delivered response so missing or mismatched work is visible before release.
Check the conditions before releasing a result.
A familiar problem
A result or proposed action needs evidence, current authority and a record of what was checked.
Its job
Develops governed checking and release control with evidence and dependency records.
What to evaluate
Qualification is specific to the operation and its release conditions.
Everyday example
If a process needs three independent checks, two completed checks should not be reported as three. Prooftronix investigates how to keep the evidence, dependencies and permission attached to that decision.
Investigate source-backed learning and how learned relationships apply when the surrounding context changes.
Develop an observer that can learn.
A familiar problem
An evolving system must relate observations, retained knowledge and the task at hand.
Its job
Researches a structured learning and observer architecture using selected reconstruction and checked proposals.
What to evaluate
Training and development continue. The programme is focused on practical language and learning milestones.
Everyday example
A system may need to learn a relationship, retain it and apply it in a different question later. The research records distinguish that demonstrated transfer from simply repeating a familiar answer.
Send a non-confidential outline of your application, current tools and target platform. Begin with a free written fit discussion. Agree any paid evaluation, licence, support and disclosure terms before work starts.
