GUS database (was: Re: Ambiguity in unordered list item definition)

Natalie Pendragon natpen at natpen.net

Mon Jun 1 14:05:00 BST 2020

- - - - - - - - - - - - - - - - - - -

Hi Hannu,

On Mon, Jun 01, 2020 at 01:00:35PM +0300, Hannu Hartikainen wrote:

Just a thought: it would be great if the GUS database was openly available

(for protocol development purposes, at least). With the current amount of

content you could just grep through everything to see if anybody is using

`*listitem` syntax (my guess: no).

I'm not necessarily opposed to sharing GUS' index, and will thinkthrough it more. I do have a few concerns that come to mindimmediately though, about which I'm curious what people think.

PRIVACY: a lot of folks around here seem to care a lot about it. In the current state of Gemini, everything created in Geminispace has a fairly high degree of anonymity through obscurity without even trying. If I release the index to the world, then it's easy for people to do things like search it for email addresses. Or use it to create a Gemini archive. People have expressed concern about that in the past, because it doesn't give users the agency to take down their own content - and now, by whitelisting GUS in their robots.txt, they're also implicitly whitelisting those use-cases.

OVERCOMPLICATION: if I just made a way to enter regex in GUS itself, then we could probably all answer the great majority of these questions, on demand, without having to download anything, or learn the schema, or figure out how to grep it (it's not a standard "database" so figuring out how to grep it is actually non-trivial). Adding regex support feels pretty compelling to me, because GUS *is* essentially an open index

In the meantime, please do feel free to email me and chat about anyindex questions you're having trouble answering. In the case of theunordered list discussion, I just wrote some regex to collect relevantnumbers, and will post them to the other thread!

Yours faithfully,Natalie