Maria System is an original project created as an attempt to design a different way of transforming text than classical ciphers. Its main idea is not a simple substitution of characters, but a gradual transformation of each character into an independent numerical trajectory.
In this system, a single character is not treated as a literal symbol, but as a starting point for a process that passes it through successive transformation phases. The result is a sequence of numbers that no longer has a direct connection to the original character. From these sequences, the MS2 record is created along with the resulting MST stream, which represents the output of the entire transformation.
An important feature of the system is that it does not preserve the plain structure of the text. Each character, space, tab and line break creates a separate trajectory. Thanks to this, it is possible to precisely reconstruct the original document, but only when the appropriate phase data is available.
In its current demonstrational version, the system consists of three logical parts: the Encoder, which builds trajectories, the Decoder, which can reconstruct them back into text, and external phase files, which act as a local, non-public data space.
Maria System was created as an experimental project. It was not intended to replace existing encryption algorithms, but to test whether it is possible to completely “blur” information by moving it into another space – the space of numbers and trajectories.
The system was designed to visualize the idea of phase trajectories, to show how text can be detached from its classical representation, and to build a foundation for potential further experiments: from static ciphers, through stream ciphers, to transmission systems.
For each character, its index is created — a numerical equivalent of what the character represents. Then, external phase sets provide values used as the material to build the trajectory. This process can be deterministic or include elements of randomness.
A trajectory is not a single number, but a sequence of consecutive values. It can be written in simplified form as:
T = [ f(i₁), f(i₂), ..., f(iₙ) ]
Without knowledge of the phases used, such a record reveals no information about the original content. This is a deliberate assumption of the system.
The MS2 file is the technical record of the trajectories created by the Encoder. It contains full trajectory data for each character, stored as numerical sequences. This file is not intended for transmission and remains locally with the system operator.
Each line of MS2 represents the full phase trajectory of a single character. It is a technical layer of the system, used for synchronization and integrity control of the decoding process.
The MST file is the transmission stream of Maria System. It is not a simple list of numbers, but a text stream with a header and a data section. The beginning of the file contains transmission metadata, including name, hashes and session parameters.
Example MST structure:
[name:transmission_005_20251207] [hash:6985dc82a7c899fdf09f8dd87e800a134a4de392d30986b7261b37c355cda965] [phase_hash:434495331dcb609918dde3eebd068f48f78521300b773359646df91254f90af5] [day:7] [mask:flipcase] ---STREAM--- kZe3mYaTmJi3ic0XnZGGlta2ncaRmJm5ic0XmdeGkZi0ncaTmtu3ic0YnduGkZaXnsaR...
The MST data stream does not reveal the structure of the text or its character boundaries. It is a continuous sequence which can be correctly interpreted only when the proper phase data and the decoder logic are available.
During encoding, each input character is converted into an index, and then — using selected phase sets and an appropriate offset — is transformed into a trajectory. Trajectories are written to the MS2 file as a technical record, and their transformed form is placed into the MST transmission stream.
v = F[ I(c) + offset ]
The Decoder reads the MST data stream, and then — using local phase files and MS2 records — reconstructs character indices and restores the original text.
The key point is that the MST file itself does not contain any information about phase sets. They must already be locally known on the receiver side.
The security of the system does not rely on a classical key, but on the lack of direct correlation between the original text and the MST record. Each character is “dispersed” into a separate trajectory, which makes statistical analysis extremely difficult.
Additionally, the system assumes the possibility of:
From the very beginning, the project was designed as an open space for further experiments. In the future, it can be developed in the direction of:
This document does not close the project, but describes its foundations in a form that allows it to be understood without going into implementation-level details.