[SPEC] Backwards-compatible metadata in Gemini

Oliver Simmons oliversimmo at gmail.com
Wed Feb 24 17:58:03 GMT 2021


On Wed, 24 Feb 2021 at 14:36, aaronleonard <aaronleonard at risingthumb.xyz> wrote:
>
> Only benefit seems to be to bots/softwares involving gemtext such as
> clients, search engines or scrapers- and it would require creating a
> standard for the metadata that someone would have to maintain, and
> outreach to get people to adopt the standard in writing gemtexts and in
> writing their clients- plus it encroaches a little bit on how long a
> person should spend to write a gemini client of their own. That's a lot
> of stuff just to have metadata that should be included as data in the
> document.
>

As with most things in Gemini, this would be entirely optional.
Either the =: or my format wouldn't take much longer to code into
clients, as they are both based on existing features and follow all of
Gemtext's simplicity rules on line-by line etc.
Outreach luckily isn't much of an issue - most of the people that will
need to know are on this mailing list, and if/once it is a thing it
will be documented.
I don't understand your "maintaining" a standard point, it's a
standard, it should not be changed once it's been decided on (except
minor adjustments if required).

>
> For the reasons above, I'm relatively indifferent to metadata but think
> it's a fair lot of time to pursue(at least to the same degree as <meta>
> tags have gone). Personally, if a person was to create and adopt a
> standard, I'd use it if it's useful AND human-readable.
>

It should DEFINITElY be human-readable, Gemini (the protocol + text
format) is supposed to be a "human-style" thing.


More information about the Gemini mailing list