Version en français

The title could be enough. But it could be considered as a too short format! So it needs to be developped.

If you know the excellent Software Heritage (SWH) and its indispensible work by its brillant team [1], the idea is the same... but instead of the software source code, it's about the file format source code.

If the previous paragraph with SWH is not clear enough and if you are still reading (I hope so!) then you want more details: here they come.

The softwares are the tools. The datas are what the tools manipulate (input) and generate (output).

The software with open source code can to be studied, re-used, upgraded, archived, even quoted [2] and that preserves our digital memory. The Software Heritage Web pages, documents or conferences explain that in a very clear and more detailled way [3].

The datas used by the softwares, as input or as output informations, have a file format: it is their structure, written in a source code. It's technically called the file format specifications.

With open specifications, it's possible to understand how the datas are structured, it's possible to use them in many softwares (not only one), it's possible to study them, to improve them, to archive them or to reuse them. Of course an open licence must allow that too.

With closed specifications, the datas are in a jail (usable by only 1 software, perhaps in only 1 version) or are dead informations (if the authors of the closed specifications decide to stop to use them, or decide to stop the only software wich uses their specifications or simply if they stop at all - how do you use Aldus Pagemaker 1.0 files [4] with the datas stored in its 1985 format?).

Open file formats are necessary, and inside a repository with powerfull possibilities to use them like Software Heritage framework, it would an excellent place.

Our world is more and more a world with digital tools, and with digital datas: they are our digital memory. That's why we need a Software Heritage, and a Format Heritage.

References:

Post-scriptum:

The title is a way of expressing an idea, it could be seen as a format. Like abstract in an academic article is perhaps the format of the modern TLDR :-) And this article, in English, on a blog, that's 3 non-digital formats too, and open! And you may prefer this last format: s/Software/Sotware+Format/g

I spoke about this idea in a free software meeting called Capitole du Libre in Toulouse (France) in november 2017. But in was a speech format, a part of a larger conference about open formats.

Last point, an example: imagine the possibilities with the Refer format or the Groff format in a FMH (Format Heritage) repository?