Structured Geminispace


Its started from blogroll


The idea behind this post was built in my head for several months. I started my presence in Geminispace from blogroll file, it was simple list of interesting capsules, which I try to read.


After a while, I found that Amfora (text gemini browser) has ability to aggregate feeds. So I moved my blogroll into Amfora.


Then I found Comitium (gemini aggregator) and I moved my blogroll again into its format.


Finally, I found Lagrange (graphical gemini browser) and I moved my blogroll again into its format. Because I am using now several devices with Lagrange, I have a bit different subset of my blogroll on every device.


So now I'm a bit bored of my frequent movings and I'm afraid that it could be not the last time when I must to move with my blogroll.


I realised also, that Comitium generated feeds page isn't perfect for blogroll. Because its of course help to track capsules, which I am tracking, but without any my personal information. So it's not a real blogroll, it's a regular capsules list.


Bookmarks


Some time ago I respond about [Planned undiscoverability of small sites], where I was considering if we should try to focus on great popularity. But I agree that we should have some standardised way for exchange information of our blogrolls (or regular bookmarks).


Structured file


Maybe above questions: - How to avoid blogroll movvings? - How to publish bookmarks? - lead to the same answer. Geminispace need a structured data format for blogrolls and bookmarks. Only one open question remains - if it should be used locally, or globally. But that answer isn't crucial.


If we would have such format, we could:



For example I could publish my blogroll, which could be import into Amfora, Comitium and Lagrange. And it could be used for everyone else.


Structured Geminispace


I mentioned in my last post about city district map. I was thinking about this theoretical map. I don't know how to draw it. I'd like to put on it the most important points in Geminispace for me.


Of course such idea must be discovered earlier. Wikipedia mentions [Giant Global Graph] which is more wide idea of that map. So I could use some FOAF like format to describe my surroundings. And I could pick some more FOAF's from my blogroll (if everybody would use such format like FOAF) to create my city district map. My part of Geminispace.


Existing formats


Of course, again - there are many formats like that. The most of them were abandoned in the further years of Web 2.0 era.


Designed for WWW/HTML


We shouldn't spend time on inventing the new format but we could use some exist specification. I thought I saw somewhere on Geminispace short post about microformats.



Designed for structuring text files


But we should transform its representation into structured plain text files. Maybe it would be sufficient to copy some standard one by one into text file?



Hypothetical format adoption


It could be hard work to introduce such new format. I can observe progress on [RFC: TinyLogs format]. What is also interesting, I didn't read any Gemini or Gopher article about problem described similar like in [Giant Global Graph]. There are more about bloated data representation of WWW, but mainly about graphic, Java-Script, adverts. Not about unstructured data, and will structurization of it.


But in my opinion abandonment of formats designed for WWW/HTML is result of the direction of how the whole WWW is going to. Because the small-net direction (Gophersphere, Geminispace) would be a different one, there could be possibility for such change.


Last but not least it would be easier for developers to have such format for data migration. Now every app has its own data format, and migration of data require manual data extraction.


[Planned undiscoverability of small sites]

[Giant Global Graph]

[FOAF]

[h-card]

[XFN]

[OPML]

[TOML]

[YAML]

[RFC: TinyLogs format]


--

szczezuja.space CC BY-SA

@ Sat 11 Sep 2021 04:58:45 PM CEST



/gemlog/