TXT makes blockchain data more useful by adding context

Here’s something new from the workshop of Planaria Corporation and its lead developer, Unwriter. ‘TXT’ is a way of creating “semantic Bitcoin storage,” something that will make storing, organizing—and finding—transaction-based information on the Metanet far easier.

"Semantic Bitcoin": TXT lets you attach meaning and context to Bitcoin transactions, allowing machines and humans to filter and process them in various powerful ways never possible before with raw transactions alone.https://t.co/sF2AJY9ei0 pic.twitter.com/x5Zp1VmoyX

— _unwriter (@_unwriter) June 2, 2020

According to the TXT homepage, “TXT (Transaction Tape) is a portable Bitcoin transaction storage system which lets you store, manage, and share bitcoin transaction bundles with semantic metadata attached.”

TXT lets users add metadata to a Bitcoin transaction ID (or raw hash) with information about data what that TXID contains. Is it a payment? A photo or audio file? Customer information or a medical record? It all gives extra power to transaction explorers and search engines, which will become more important as Bitcoin’s Metanet grows into something as large as, or bigger than, today’s internet.

TXT will auto-sync with Miner API (M/API) and can read/write HTTP APIs. It includes instant data feed channels, portable backup storage… and more. Users can attach the metadata directly using TXT’s API, or generate it automatically using a confirmed TXID which will sync with transaction processors via M/API.

What is semantic metadata?

Metadata is “information about data,” or in other words, more descriptive information that explains what the data is, why it’s there, what it does, and anything else. Think of EXIF data on photos that contains information about type of camera, location the photo was taken, settings, etc. Users can use their photo management application to add things like star ratings and comments.

Giving data this extra context makes it more human-understandable and easier to find. For example, say you’re looking for that photo of your daughter you took about nine years ago—you can’t remember the filename (which is something like “IMG0000323445.jpg”) or the exact month, but you can remember it was at a birthday party and she was wearing a blue dress.

Now, say you never wrote “blue dress” in the notes, but “navy skirt” in the notes. That’s where the “semantic” part comes in. Semantics in technology allows for connections between words and concepts with similar meanings (think of the frustration when you spend ages searching for a file called “birthday-party” but never find it because it’s actually called “birthday_party”). Web search engines have improved over time as semantic technology evolves, serving up more relevant results… most of the time. It will also be useful for finding songs where you can remember every word of the chorus but the artist gave it a completely unrelated title.

TXT’s semantic layer can be used to upload (and encode) information like intent, filters, privacy, triggers, structure, and other data like “ephemerality,” “non-determinism” and others. It also lets you “bundle” transactions if they contain data related to each other. The information in the transaction can stay encrypted (Planaria gives the example of a description “these encrypted gibberish transactions are actually all my emails”).

There are two components to TXT: a portable storage format, which uses SQLite, and a Container (which it describes as a “tape player/recorder”). This allows the metadata to be shared across different containers (think of a music track which keeps all your ratings and notes even if you open it in a different player).

Current TXT demonstrations

There’s an implementation of TXT running on the BitPic network, to which users have uploaded thousands of images to the Bitcoin blockchain as transactions. Here’s an image I uploaded to BitPic last year (thanks to the blockchain, I can see it was on December 11, 2019).

As you can see from the current metadata, the information is pretty basic: mainly the transaction ID and the Paymail address I used when I uploaded it. I could find it thanks to that address, because there isn’t any other information to describe what the image is.

Using TXT, I could add more descriptive information that could help me find it decades into the future, or help others who might come across it (who knows, it could lead to more video views after I’ve passed on… hopefully also on the blockchain, since YouTube will probably have deleted everything by then).

An image file is pretty basic, and TXT can add metadata to any piece of data recorded in a Bitcoin transaction ID—meaning its uses are limitless. Someday it could be a full HD video, or some other vital record like an Act of Parliament, legal evidence, or important historic artifact.

Another implementation is running on C.txt, a Bitcoin file search engine Planaria developed. The more metadata users add to the data they upload in Bitcoin transactions, the more useful this will become.

How do I use it?

For now, using TXT means downloading the developer “container” app Docker (which allows apps to run in their own environment, where they can’t affect the rest of the system). You’ll then need to use its command-line interface to edit the data.

Over time, developers will build more beginner-friendly interfaces. However the backend and functionality is already there.

TXT is one more tool that will make Bitcoin far more useful now and into the future. Only Bitcoin BSV has the ability to store limitless amounts of data securely and immutably. And now, that data can be more searchable and manageable. That will be a huge relief to future users.