FENYX : ALTER THESAURUS SCRIPTORUM MUSICORUM

No. By definition, a pirate site distributes copyrighted content without authorization. Fenyx, however, only hosts content that is in the public domain.

No. By definition, a leaks site distributes material obtained illegally. Fenyx’s content, however, was acquired through completely legitimate means.

All the treatises presented here come from copies of the Thesaurus musicarum latinarum (currently unavailable) preserved and made publicly available by the Wayback Machine. These copies have been accessible for many years. If no one has objected to the availability of these multiple copies, there is no reason to take issue with them being compiled by Fenyx.

The members of the Fenyx group are researchers frustrated by being deprived for many months of a resource essential to their work. None of them have any connection with the past or present TML team.

The 939 treatises made available here correspond to snapshots from spring 2025. They represent the entirety, or nearly the entirety, of the TML data at the time it went offline.

Fundamentally, it is the same data. However, all the files, as they were served by TML, have been reformatted according to the TEI standard, which has the advantage of consolidating, in a single structured electronic file, both the content of the treatises and their metadata, readable by humans and machines alike. We have also removed statements such as “Reproduced by permission”. Indeed, no permission was needed to reproduce raw transcriptions of public domain works. Moreover, these statements, likely related to TML’s history, could misleadingly suggest that third parties might claim exclusive rights over this content.

Inevitably. Any errors that were already present in the TML have not been corrected. Additionally, the files were processed in batches and automatically. It is therefore possible that certain irregularities in the original files led to processing errors. We were not able to check every file individually, consequently, we cannot provide any guarantee regarding the accuracy of the data. Nevertheless, if users report errors to us, we will make every effort to correct them.

The benefits are both personal and collective. Having a complete dataset locally allows direct access and querying without relying on a website’s availability and without being limited by its interface, which can offer significant advantages. Furthermore, TML has experienced several interruptions, sometimes long-lasting, over the past few years, raising concerns about the structural fragility of its historic host, or even potential neglect. Under these circumstances, the existence of multiple private copies reduces the risk of permanent data loss.

For the moment, there are no plans to restore them. On one hand, they are much less developed than TML. On the other, unlike TML, they are not released under a Creative Commons 4.0 license, which could create legal complications.

We began working on restoring the TML data at the end of summer 2025. The project was progressing slowly, but we were galvanized by the surge of support generated in November 2025 by the launch of a petition, which prompted us to double our efforts. Given this momentum, the idea of entering 2026 without a TML backup had become unbearable for us. At the same time, the historical hosting provider showed notable technical agility by substituting, in extremis, the traditional “404 Not Found” — which had been greeting visitors for months — with a customized “File Not Found” page. One may still wonder whether this counts as genuine progress or an unintentional piece of humor.