[Logarion] On keeping human managed text files human readable

SP sp at orbitalfox.eu
Wed Feb 28 23:54:04 UTC 2018


On Wed, Feb 28, 2018 at 07:40:23PM +0000, refpga at disroot.org wrote:
>I think the machine generated data, should be kept away from the user
>centric parts, which adds to verbosity, and stands out as gibberish.

In Logarion a text file is a stand-alone independent unit. It doesn't
"belong" to the archive. It instructs and informs the archive. The
meta-data must be with the file, and because the file is human-readable,
the meta data is exposed to the person.

I'm not very fond of UUID format, but it's a good standard for the
purpose it's needed, which is to create unique identifiers across
disconnected systems.

Currently reconsidering its encoding. There is a balance between
interoperability with other programs and aesthetics.

>At best, it's unnatural to write by hand, and has no place within
>text files managed by users.
>Maybe generate uuid for new and renamed files while convert process.

There will be an `add` or `import`, or something the like, being
implemented soon, where one can start with a text that was produced
without any meta-information.

The author is not meant to write this meta-data by hand. Programs
either pre-prepare it for writing, or added to an existing text.

Also, once I get to editors, this meta-data can be hidden, or isolated
to special forms, same thing that happens to an email & HTTP: all the
data you input just generates a stream of human readable syntax.

>Maybe in the future, replace Author by username, if a federated
>network of archives, is what this turns out to be.

Logarion is a writing-centric environment and thus 'author' seems more
appropriate.

-- 
	SP


More information about the Logarion mailing list