💾 Archived View for oldest.gwit.site › log › _en › 20230802--tough-changes captured on 2024-05-10 at 10:37:15. Gemini links have been rewritten to link to archived content

View Raw

More Information

⬅️ Previous capture (2023-09-08)

-=-=-=-=-=-=-

Tough changes to keep gwit simple

Lately I've been updating the gwit specification to fix several issues that have surfaced during private mail exchanges and discussions with people who I consider relevant in the fields of Gemini and decentralized publishing. First, I want to thank each of them for their help and the time and effort that they have devoted to this, and for their very valuable input! 🖖 Also note that the "gwit-spec" mailing list is always open for further public discussion.

Discussion list for the gwit specification

While many of the changes are relatively minor, I've had to face some others with a deeper impact, where overall simplicity seemed at odds with author convenience. It's been very tempting to lean towards a (probably false) sense of usability that may attract more people towards gwit, at the cost of increased system complexity and fragility. These have been tough decisions, but in the end I feel that sticking to gwit's design goals will prove right in the long term.

gwit's design goals

I believe it's important to remind oneself of these goals. First the functional goals:

And the architectural goals:

It's not easy to stick to such radical principles when designing a software system, and not fall for endless flexibility and apparent friendliness. This reminds me of the latest updates to the Gemini FAQ and the reasons behind such peculiar but powerful decisions in protocol design.

Project Gemini FAQ - §4 Protocol design (see point 4.2)

With more changes to come soon (like splitting and reorganizing the spec), I hope that gwit's design goals will always help make the right choices.

🍃

🗒️ Back to log index