Hi, citing the specs, 5.4.2:
Clients can present links to users in whatever fashion the client author
wishes, however clients MUST NOT automatically make any network
connections as part of displaying links whose scheme corresponds to a
network protocol (e.g. gemini://, gopher://, https://, ftp://, etc.).
So, mime is useless as human has the final say, not the client
Nicol?
Tom <tgrom.automail at nuegia.net> writes:
>> 3) I am a mathematical physicist and use a lot of equations
>> and data visualization. As a scientist, how can I use gemini?
>> For equations, I usually use MathJax. Talk about bloating!!!
>> It is great but it is huge. There are alternatives, but none
>> is near close to a plain text solution. We also have a set
>> of utilities like tex2mail, which converts LaTeX into ascii
>> or unicode art, but... it is really ugly and inconvenient.
>> We could just use raw LaTeX code, which for some expressions
>> would be fine:
>>
>> \frac{a+b}{x^2+x-2}
>>
>> or
>>
>> (x+y^2)^n
>>
>> but what about a matrix or a more involved code?
>> Is this up to us or is there a way to integrate this without
>> adding excessive complication?
>>
>> What about data? The only thing that comes to me now is to
>> use Python's Sympy, which can render LaTeX equations in
>> ascii and also data graphics.
>>
>> 4) What about musicians, or artists? I am not talking about
>> music score, which even in html has never been offered a nice
>> solution (that I know of). But audio? Is it possible to
>> have audio? I mean, no need for the browser to play it, but
>> to link it? And what about artistic expression? As much as
>> I love plain text and black and white, many Sapiens consider
>> design and style a very important thing. Will these communities
>> feel excluded from gemini because of extreme (I like extreme)
>> simplicity? I am sure I still have a lot to learn here.
>> Sometimes I use ncurses just to have some nice colors in my
>> terminal. My question here is: is the style and design of
>> the page enforced by the protocol, the client or the page
>> itself? IMHO a plain text browser in which each page could
>> define background and text color would be really awesome.
>>
>> 5) I love design and colors, but most of all I love content,
>> and I think that gemini is a refocus on content. And I love
>> it, with or without colors, with or without math. But my dream?
>> A minimalistic browsing with page-defined colors and with
>> some easiness to integrate math and graphics with plain text.
>>
>> Thanks for all the work. I will keep learning from you.
>> And I will also write less words next time (`:
>>
>> fluxion
>>
>
> Hello,
>
> I understand your concern completely and share some of it. One of the
> biggest reasons not to use gemini over XHTML is the lack of MathML. And
> a reason to use MathML over others such as just exporting a scalable
> vector graphic is so that MathML can be read as Nemeth code so that
> disabled people may still view the math. Talking about accessibility
> this super important.
>
> Now I'm not a fan of not being able to insert objects inline in gemini,
> but the compromise of just having a link on a single line to do not
> feel serves this side particularly well. Take a more general thing like
> images. Clients that do want to display images the only way they have
> of determining which links contains images is be following each one of
> them, which in a form can trigger unwanted actions and certainly
> increase overall network load. Some clients look at the extension at
> the end of the file name, for example .jpg or .png and preload if the
> filename ends in that. I do not think this is a good option. For one we
> have MIME, and secondly using filenames for determining file type is a
> relic of DOS we really should leave behind, and we have left behind on
> every operating system besides DOS and NT. Replacing with magic(5).
>
> Many files to not even have a .something at the end of their name as
> magic(5) takes care of type recognition.
>
> I don't think it would be practical to have Gemini clients need to
> check the magic of every link and instead think a much more robust
> solution would be simple to optionally (or mandatory) add an expected
> MIME for links in Gemini. This way clients could decide what they are
> capable of rendering and decide to go get that content if it wants it.
>
> This however I feel is still only a half-solution for expressing
> mathematical and scientific formulas as it is still not inline, which
> is how scientific papers often use formulas.
>
> Looking for more commentary on this topic.
---
Previous in thread (4 of 5): 🗣️ Tom (tgrom.automail (a) nuegia.net)