On Thu, Jan 16, 2020 at 3:51 PM Aaron Janse <aaron at ajanse.me> wrote: > On Thu, Jan 16, 2020, at 12:26 PM, solderpunk wrote: > > If you're hosting shared gemini content for a group of non-technical users, > > it might make sense for your server to automatically reformat gemini files > > before sending them > > I don't have a problem with people "rendering" into Gemini. I just think that > the spec shouldn't make it necessary. Rendering/compiling into Gemini adds > a level of indirection that hides the transparency that makes me love the > protocol. > > Cheers! To be clear, I agree with you that a max line length shouldn't be enforced by the spec. I support the current wording of SHOULD as opposed to MUST. I'm more talking about establishing a recommended best practice. Gemini files with lines over N characters long should still be considered "valid" and must be accepted by any gemini client. They might just look a little funky depending on how sophisticated your client is. My belief is that if we don't establish a guidance on the line length, the community will informally gravitate towards one anyway. And that will almost certainly be 70-80 characters because that's what most people in this community are used to. Servers that return 500+ characters on a single line will look bad on many of the gemini clients that people write, and content authors will be pressured into conforming to maintain readability for those clients. The spec-spec as written right now says that all gemini lines should be formatted so that they can reflowed by clients if the client chooses to do so. That spec has been almost unanimously rejected by current gemini servers, in favor of hard wrapping at 70-80 characters. - mozz
---
Previous in thread (89 of 148): 🗣️ Aaron Janse (aaron (a) ajanse.me)
Next in thread (91 of 148): 🗣️ solderpunk (solderpunk (a) SDF.ORG)