💾 Archived View for gemi.dev › gemini-mailing-list › 000460.gmi captured on 2024-08-19 at 00:46:58. Gemini links have been rewritten to link to archived content
⬅️ Previous capture (2023-12-28)
-=-=-=-=-=-=-
There seems to have been a lot of discussion recently about extending the protocol with various features. I propose that the spec be frozen again for some time to allow people to be creative within the existing limitations of Gemini instead of trying to extend the protocol.
On Tue, 10 Nov 2020 18:30:54 -0500 "Adnan Maolood" <me at adnano.co> wrote: > There seems to have been a lot of discussion recently about extending > the protocol with various features. > > I propose that the spec be frozen again for some time to allow people > to be creative within the existing limitations of Gemini instead of > trying to extend the protocol. I second that, it seems some people are invested in working within the limitations with their creativity, see Emery's post: On Wed, 11 Nov 2020 00:36:07 +0100 Emery <ehmry at posteo.net> wrote: > I've found a solution to the great content length dilema that I may > have unwisely contributed to, magnet links! > > magnet:?xs=gemini://gemini.spam.works/mirrors/lainzine/lainzine.org/txt-r eleases/lainzine-1.txt&xl=60132&xt=urn:sha256:9e7b437857cd3939e39e0a1c953f7 3da3f9045fb0b878014e9917110f8652afd
> On 11 Nov 2020, at 10:30 am, Adnan Maolood <me at adnano.co> wrote: > > There seems to have been a lot of discussion recently about extending > the protocol with various features. It is a truism of this list that if you pop in on any given month there will be at at least one fresh proposal for including content length in the MIME type, somebody else being cheeky with non-gemini URIs, a discussion about escaping the preformatting characters, and possibly a request to remove the TLS requirement. :) I'm not blaming anyone in particular - a lot of design discussion is buried in the mailing list archives - but many of the common themes are addressed pretty comprehensively in the FAQ and maybe we could save everybody's energy if we promoted it more: https://gemini.circumlunar.space/docs/faq.html Cheers, Tom
On 11/10/20 5:30 PM, Adnan Maolood wrote: > There seems to have been a lot of discussion recently about extending > the protocol with various features. > > I propose that the spec be frozen again for some time to allow people > to be creative within the existing limitations of Gemini instead of > trying to extend the protocol. Since no one voiced an objection to my proposal of November 7th perhaps it could still be considered for inclusion? Something in section 1.1 or 4 to communicate that servers should/must send a TLS close_notify message before closing the TCP connection.
> Since no one voiced an objection to my proposal of November 7th perhaps > it could still be considered for inclusion? > > Something in section 1.1 or 4 to communicate that servers should/must > send a TLS close_notify message before closing the TCP connection. FWIW I support that proposal. And if it can't be included in the spec, at least include it in the best practices document. Why use TLS at all, if we are fine with chance or a bad guy truncating a message ? Either use TLS properly or don't use it at all. regards marc
On Mittwoch, 11. November 2020 01:02:21 CET, Thomas Karpiniec wrote: > It is a truism of this list that if you pop in on any given > month there will be at at least one fresh proposal for including > content length in the MIME type, somebody else being cheeky with > non-gemini URIs, a discussion about escaping the preformatting > characters, and possibly a request to remove the TLS > requirement. :) I'm also in favor of freezing the spec. I think the meaningful experiments that haven't happened are higher-level, and the protocol discussions feel like a distraction. Emery
I'm against a protocol freeze for two reasons. 1. I don't see any feature creep as of yet. Many *suggestions*, but so far none that have been accepted. 2. There are parts of the specification that need clarification. I.e. the "Empty META" question a short while back. I would much rather see an engaged discussion about finding and clarifying ambiguity in the protocol specification than suggestions of new features, honestly. I think that's where the large gains are to be made. The community at large seems to have agreed that the protocol is very close to final, and feature suggestions are often met with "solve it with a different protocol" or "use a different filetype" (in the case of gemtext questions). Cheers, ew0k
On Wed, 11 Nov 2020 10:50:29 +0100 Bj?rn W?rmedal <bjorn.warmedal at gmail.com> wrote: > 2. There are parts of the specification that need clarification. I.e. > the "Empty META" question a short while back. I would much rather see > an engaged discussion about finding and clarifying ambiguity in the > protocol specification than suggestions of new features, honestly. I > think that's where the large gains are to be made. As far as I can remember if my memory serves me right, solderpunk said he'd update the spec for that, this is an acceptable change. This and the TLS related proposals.
Solderpunk proposed a plan for spec freezes in the beginning of March gemini://gemi.dev/gemini-mailing-list/messages/000463.gmi Permanent freeze gemini://gemi.dev/gemini-mailing-list/messages/001708.gmi Den ons 11 nov. 2020 00:40Adnan Maolood <me at adnano.co> skrev: > There seems to have been a lot of discussion recently about extending > the protocol with various features. > > I propose that the spec be frozen again for some time to allow people > to be creative within the existing limitations of Gemini instead of > trying to extend the protocol. > As far as I know, we are already in a spec freeze since the middle of June. If we go by the original plan, presented by Solderpunk when we entered the first freeze back in the beginning of March[1], the second freeze would end in the middle of December. However, it seems like Solderpunk changed his mind when we entered the second freeze in the middle of June[2], going for a permanent freeze instead. This permanent freeze is only on non-trivial new features. The way I interpreted this is that we will not add features and we have to wait until December to clear things up in the spec. Please let me know if I missed something. -- Katarina > -------------- next part -------------- An HTML attachment was scrubbed... URL: <https://lists.orbitalfox.eu/archives/gemini/attachments/20201111/2960 3208/attachment.htm>
Oops, I sent the reply before moving the footnotes to the bottom. [1] > gemini://gemi.dev/gemini-mailing-list/messages/000463.gmi > [2] > gemini://gemi.dev/gemini-mailing-list/messages/001708.gmi > -- Katarina > > -------------- next part -------------- An HTML attachment was scrubbed... URL: <https://lists.orbitalfox.eu/archives/gemini/attachments/20201111/d7b3 163d/attachment.htm>
---
Previous Thread: Statement of intent regarding document encodings
Next Thread: A hack to support content length and content hash