💾 Archived View for gemi.dev › gemini-mailing-list › 000370.gmi captured on 2024-08-25 at 11:20:06. Gemini links have been rewritten to link to archived content

View Raw

More Information

⬅️ Previous capture (2023-12-28)

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

HTML as an escape hatch

1. Nathan Galt (mailinglists (a) ngalt.com)

There?s a persistent worry on other threads that Gemini might become ?too 
complex?, for some definition of ?complex?.

This is a reasonable worry, considering the very complicated mess we?re all running from.

On the other hand, I think there?s a fairly, if not fully, general 
counterargument to all sorts of additions that we might want to make to the spec:

?There?s nothing wrong ? or even uncool ? about making a website with only 
HTML and, at most, 30 lines of CSS that looks great in Lynx.?

Thoughts? Counterarguments?

Link to individual message.

2. Stacy Harper (contact (a) stacyharper.net)

Hello,

> On the other hand, I think there?s a fairly, if not fully, general 
counterargument to all sorts of additions that we might want to make to the spec:

I think if you want additions, just use WWW. Gemini focus on content and
not formating.

> ?There?s nothing wrong ? or even uncool ? about making a website with 
only HTML and, at most, 30 lines of CSS that looks great in Lynx.?

Yes you can create cool website using some standard html tags and a
little bit of css. But Gemini choosed its own gmi format specially to
prevent some implementations to add this kind of cool features.

> Thoughts? Counterarguments?

WWW decadence was fun, let's never do it again :)

Link to individual message.

3. meff (meff (a) meff.me)

For me, I see nothing wrong with Gemtext and HTML living side by
side. This isn't a zero-sum game, as much as the Web wants us to think
it is with its focus on hyperscale. As Stacy mentioned, Gemtext is great
for content and not formatting. If you want formatting, then use HTML!

- meff

Stacy Harper <contact at stacyharper.net> writes:

> Hello,
>
>> On the other hand, I think there?s a fairly, if not fully, general 
counterargument to all sorts of additions that we might want to make to the spec:
>
> I think if you want additions, just use WWW. Gemini focus on content and
> not formating.
>
>> ?There?s nothing wrong ? or even uncool ? about making a website with 
only HTML and, at most, 30 lines of CSS that looks great in Lynx.?
>
> Yes you can create cool website using some standard html tags and a
> little bit of css. But Gemini choosed its own gmi format specially to
> prevent some implementations to add this kind of cool features.
>
>> Thoughts? Counterarguments?
>
> WWW decadence was fun, let's never do it again :)

Link to individual message.

4. Sean Conner (sean (a) conman.org)

It was thus said that the Great Nathan Galt once stated:
> There?s a persistent worry on other threads that Gemini might become ?too
> complex?, for some definition of ?complex?.
> 
> This is a reasonable worry, considering the very complicated mess we?re
> all running from.
> 
> On the other hand, I think there?s a fairly, if not fully, general
> counterargument to all sorts of additions that we might want to make to
> the spec:
> 
> ?There?s nothing wrong ? or even uncool ? about making a website with only
> HTML and, at most, 30 lines of CSS that looks great in Lynx.?
> 
> Thoughts? Counterarguments?

  It's one of the anti-Gemini arguments I've seen on sites like Hacker News
(https://news.ycombinator.com)---why not just use a stripped down HTML and
webservers that do what you want?  And yes, there is something to that
agument, but solderpunk has made it clear he doesn't want any crack that
will lead to the current web (and personally, I would love it if the
"application web" went via QUIC and the "original hypertext web" stayed on
TCP).

  The problem I see with gemtext (and having been on the mailing list from
the beginning, way over half the messages have been about the gemtext
format) is that people *want* their HTML, but it's mostly different subsets
of HTML.  About the only agreement is the absence of Javascript.

  Personally, I'm not a fan of the gemtext format as I find it restrictive
in a way that I don't find plain text or HTML, but it is what it is.  All I
do is put up pages with the proposed specification and see if there are
complaints (and there were---oh the surprise).

  Again, if you want formatting, HTML and Markdown are there for you to use
on Gemini.

  -spc

Link to individual message.

5. Sean Conner (sean (a) conman.org)

It was thus said that the Great Stacy Harper once stated:
> Hello,
> 
> > On the other hand, I think there?s a fairly, if not fully, general
> > counterargument to all sorts of additions that we might want to make to
> > the spec:
> 
> I think if you want additions, just use WWW. Gemini focus on content and
> not formating.

  If gemtext was focused on content and not formatting, it wouldn't have
headers, lists, or pre-formatted areas.

  My own websites:

	http://www.conman.org/
	http://www.conman.org/people/spc/
	http://boston.conman.org/

uses straightforward HTML (4.01 if anyone cares), and while I do use some
CSS for formatting, it still has a structure that is visible if you strip
away the CSS (on Firefox, if you go "View ? Page Style ? No Style" it is
still readable).  And you would have to really hunt about my websites to
find the Javascript (it's there, just a very little bit of it).

> WWW decadence was fun, let's never do it again :)

  I think I'm one of the few left that actually like hypertext qua
hypertext.  Sad.  It's a neat concept.

  -spc

Link to individual message.

6. Philip Linde (linde.philip (a) gmail.com)

On Thu, 10 Sep 2020 23:17:19 -0700
Nathan Galt <mailinglists at ngalt.com> wrote:

> On the other hand, I think there?s a fairly, if not fully, general 
counterargument to all sorts of additions that we might want to make to the spec:
> 
> ?There?s nothing wrong ? or even uncool ? about making a website with 
only HTML and, at most, 30 lines of CSS that looks great in Lynx.?
> 
> Thoughts? Counterarguments?

I agree with this notion. text/gemini has a very simple format, still.
I think there are several good reasons for that being the case:

1. It's easy to parse. You can write a parser in an afternoon, and a
   renderer for the terminal or your favorite GUI toolkit that same
   evening
2. It's easy to write pages. You can grasp every nuance of the document
   standard in an hour
3. It's human readable. You can easily read a text/gemini page as plain
   text because the markup is minimal
4. It has enough features to write very basic formatted documents, but
   not enough for the author to impose on the user

Similarly, for HTML, there are good reasons for it being structured as
it is. It supports nested documents that allow for greater control of
formatting, but at the expense of all the reasons I listed for Gemini.

If there is a good middle ground between HTML and Gemini, I'll happily
support that as its own MIME type, but I think the final most important
complexity that text/gemini should address is that it shouldn't be a
"living standard". That has a huge cost to site admins and developers.

-- 
Philip
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 488 bytes
Desc: not available
URL: <https://lists.orbitalfox.eu/archives/gemini/attachments/20200911/206a
58ba/attachment.sig>

Link to individual message.

7. Alex // nytpu (alex (a) nytpu.com)

> If there is a good middle ground between HTML and Gemini, I'll happily
> support that as its own MIME type
While I understand the desire to have something novel, there really is
no need to reinvent the wheel here. The whole point of having mimetypes
in gemini is so you can use whatever format you want to serve content.
If you want basic linking and sections, use gemtext. If you want rich
formatting (emphasis, boldface, etc) use markdown or asciidoc or any of
a million lightweight markup languages. If you want deep control over
the presentation of the document, use html. If you want no formatting at
all, use plaintext.  There's near certainty that there's an existing
format that generally suits your needs unless you're doing something
like math formulae, and even then you could just link to an image or use
a pdf and LaTeX.

The thing is, I think people have a misconstrued idea of what gemtext
is. The long and short of it is that it's a gophermap with headers and
code blocks. Sure, the syntax is different but at the end of the day
that's all it is. If you can't live without stronger control over the
styling and document structure you should just use a subset of html
(There's absolutely nothing wrong with that! You can still browse gemini
without having a gemini capsule yourself). You could even serve your
html over gemini if you so chose, although I'm not sure how well it'd be
supported by clients.

> but I think the final most important complexity that text/gemini
> should address is that it shouldn't be a "living standard". That has a
> huge cost to site admins and developers.
I think this should aim to be the highest priority over "making gemtext
better." Obviously there's an exception right now since it's in the
early stages and not finalized yet, but even now there shouldn't be
massive changes made to the spec at all, let alone changes that turn
gemtext into an ad hoc, impossible-to-parse implementation of half of
html.

-- 
Alex // nytpu
alex at nytpu.com
GPG Key: https://www.nytpu.com/files/pubkey.asc
Key fingerprint: 43A5 890C EE85 EA1F 8C88 9492 ECCD C07B 337B 8F5B
https://e-mail.is-not-s.ms/
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: not available
URL: <https://lists.orbitalfox.eu/archives/gemini/attachments/20200911/3d68
5ed3/attachment.sig>

Link to individual message.

8. Gary Johnson (lambdatronic (a) disroot.org)

It's true that the Gemini protocol is able to serve up a variety of file
formats given its explicit requirement of mime-types in response
metadata. I'm also happy with the simple stripped-down nature of the
Gemtext format and am not looking for any major changes to be made to it
at this time.

However, I do want to throw in a counterpoint to recommending people
abandon Gemtext and serve up more Markdown or HTML in Geminispace.

While it is true that they can all be served up by the Gemini protocol,
it is not currently the case that very many Gemini clients can render
anything other than Gemtext. If enough people start to publish in
Markdown and/or HTML, we may soon find our experience of surfing
Geminispace to largely be about downloading Markdown and HTML pages
whenever we click on links and then having to open the local files in
our format-specific renderers of choice. HTML pages also reintroduce the
issue of potentially requiring clients to download any number of
auxiliary pages in order to render them. While we could hope that Gemini
publishers will stick to a minimal subset of HTML (for each user's
varying definition of "minimal"), there is no way to enforce this.

This does not sound like much fun to me, and it would impose a much
higher development cost on each browser developer as they would have to
implement rendering engines for Gemtext, Markdown, HTML, etc. in order
to provide their users with a more seamless browsing experience. Given
uneven distribution of resources to maintain and extend clients, this
could drive us down a path in which a smaller and smaller number of
browsers come to dominate Geminispace as the number of requirements on
what they have to render continues to increase.

In a worst-case scenario, we ultimately end up recreating a post-web
monopoly of Geminispace.

Just my 2c, but I'd like to see more folks encouraging the use of
text/gemini or text/plain in Geminispace rather than punting to more
complex formats except in special cases. If anyone is feeling super
restricted by Gemtext, I'd encourage them to spend some time on
Astrobotany (gemini://astrobotany.mozz.us) for inspiration.

YMMV,
  Gary


Alex // nytpu <alex at nytpu.com> writes:

>> If there is a good middle ground between HTML and Gemini, I'll happily
>> support that as its own MIME type
> While I understand the desire to have something novel, there really is
> no need to reinvent the wheel here. The whole point of having mimetypes
> in gemini is so you can use whatever format you want to serve content.
> If you want basic linking and sections, use gemtext. If you want rich
> formatting (emphasis, boldface, etc) use markdown or asciidoc or any of
> a million lightweight markup languages. If you want deep control over
> the presentation of the document, use html. If you want no formatting at
> all, use plaintext.  There's near certainty that there's an existing
> format that generally suits your needs unless you're doing something
> like math formulae, and even then you could just link to an image or use
> a pdf and LaTeX.
>
> The thing is, I think people have a misconstrued idea of what gemtext
> is. The long and short of it is that it's a gophermap with headers and
> code blocks. Sure, the syntax is different but at the end of the day
> that's all it is. If you can't live without stronger control over the
> styling and document structure you should just use a subset of html
> (There's absolutely nothing wrong with that! You can still browse gemini
> without having a gemini capsule yourself). You could even serve your
> html over gemini if you so chose, although I'm not sure how well it'd be
> supported by clients.
>
>> but I think the final most important complexity that text/gemini
>> should address is that it shouldn't be a "living standard". That has a
>> huge cost to site admins and developers.
> I think this should aim to be the highest priority over "making gemtext
> better." Obviously there's an exception right now since it's in the
> early stages and not finalized yet, but even now there shouldn't be
> massive changes made to the spec at all, let alone changes that turn
> gemtext into an ad hoc, impossible-to-parse implementation of half of
> html.


-- 
GPG Key ID: 7BC158ED
Use `gpg --search-keys lambdatronic' to find me
Protect yourself from surveillance: https://emailselfdefense.fsf.org
=======================================================================
()  ascii ribbon campaign - against html e-mail
/\  www.asciiribbon.org   - against proprietary attachments

Please avoid sending me MS-Office attachments.
See http://www.gnu.org/philosophy/no-word-attachments.html

Link to individual message.

9. meff (meff (a) meff.me)

Gary Johnson <lambdatronic at disroot.org> writes:

> However, I do want to throw in a counterpoint to recommending people
> abandon Gemtext and serve up more Markdown or HTML in Geminispace.
>
> While it is true that they can all be served up by the Gemini protocol,
> it is not currently the case that very many Gemini clients can render
> anything other than Gemtext. If enough people start to publish in
> Markdown and/or HTML, we may soon find our experience of surfing
> Geminispace to largely be about downloading Markdown and HTML pages
> whenever we click on links and then having to open the local files in
> our format-specific renderers of choice.

I see this largely as a feature, not a bug. HTML desired to be just
this, the de facto and de jure landing spot on the web. It's now become
a vehicle for everything: for applications hooking into a Document
Object Model to modify state, for semantic markup, and for layout. This
has also largely driven innovation into HTML away from other formats,
because HTML is the format that everyone sees.

> This does not sound like much fun to me, and it would impose a much
> higher development cost on each browser developer as they would have to
> implement rendering engines for Gemtext, Markdown, HTML, etc. in order
> to provide their users with a more seamless browsing experience.

Lightweight markup formats like Markdown and ReST are readable without
your browser formatting anything. Lightweight clients can get away with
not parsing anything and dumping a set of MIME types directly to the
user (either on-screen, to stdout, or however the browser works). HTML
has high quality parser implementations on various platforms in various
languages. XML also usually has good quality parsers available on most
platforms/languages. HNGopher, a Hacker News Gopher mirror, uses w3m to
render HTML content into plain text. Nothing is stopping Gemini browsers
from shelling out to w3m on platforms that have it available and
rendering content this way.

- meff

Link to individual message.

10. Matthew Graybosch (hello (a) matthewgraybosch.com)

On Thu, 10 Sep 2020 23:17:19 -0700
Nathan Galt <mailinglists at ngalt.com> wrote:

> There?s a persistent worry on other threads that Gemini might become
> ?too complex?, for some definition of ?complex?.

The only problem I've had with Gemtext is that it so closely resembles
Markdown that sometimes I use the wrong syntax for URLs when I'm drunk.

This isn't a bad thing, incidentally. Gemtext lends itself well to a
Gemini-first/Web-second approach where you write your content as
Gemtext first, then copy the file to the same name with a .md
or .markdown extension, change the link syntax, and add frontmatter if
you're using a static site generator like Jekyll, Hugo, or Pelican.
Hell, you could probably just use pandoc and a makefile if you know
what you're doing.

Or, you could just take your gemtext, change the extension, and
manually wrap everything in HTML tags.

> On the other hand, I think there?s a fairly, if not fully, general
> counterargument to all sorts of additions that we might want to make
> to the spec:
> 
> ?There?s nothing wrong ? or even uncool ? about making a website with
> only HTML and, at most, 30 lines of CSS that looks great in Lynx.?
> 
> Thoughts? Counterarguments?

I think a lot of people concerned about web bloat would benefit from
being directed to the following websites even if their names might be
offensive:



The first is just plain HTML with only browser default styles. The
second adds seven lines of inline CSS to the former. The third builds
upon the second and adds some server-side goodies like HTTPS, caching,
and gzip compression.

It's still not that hard to create a static website if that's what
people really want to do. I just wonder if they've forgotten how.

-- 
Matthew Graybosch		https://matthewgraybosch.com
#include <disclaimer.h>		https://starbreakersaga.com
	 			gemini://tanelorn.city
"Out of order?! Even in the future nothing works."

Link to individual message.

11. Matthew Graybosch (hello (a) matthewgraybosch.com)

On Fri, 11 Sep 2020 08:59:38 +0200
Stacy Harper <contact at stacyharper.net> wrote:

> WWW decadence was fun, let's never do it again :)

This should be added to the fortune cookie databases used by the BSD
fortune utility. =^.^=

-- 
Matthew Graybosch		https://matthewgraybosch.com
#include <disclaimer.h>		https://starbreakersaga.com
	 			gemini://tanelorn.city
"Out of order?! Even in the future nothing works."

Link to individual message.

12. Wojciech Kwolek (me (a) irth.pl)

On 9/15/20 2:50 PM, Matthew Graybosch wrote:
> Hell, you could probably just use pandoc and a makefile if you know
> what you're doing.
I do that for my website (gemini://irth.pl / https://irth.pl).
Gemini is simple enough to write a converter for yourself in like, two
evenings:

 ? https://git.r23s.eu/wojciech/gm2html
 ? (I don't have a Gemini git front-end set up yet, excuse me :))

I run it with a Makefile to generate .html files, and I wrote some nice 
CSS to
style it: https://github.com/irth/website/blob/master/Makefile

For relative links it changes the .gmi extension to .html automatically, so
links work well on both versions.

If anyone wants, you can borrow my CSS, I'll probably add an MIT license 
and a
README to the repos when I find some time and willpower.


> It's still not that hard to create a static website if that's what
> people really want to do. I just wonder if they've forgotten how.
It is possible, but having a specification that forces you to do that allows
the reader to have some expectations before clicking a link. If I see a
gemini:// link, I know I won't have to sit through a few seconds of 
loading up
all the javascript before seeing the text :)

-- 
Wojciech ~irth Kwolek

- gemini://irth.pl
- https://irth.pl
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.orbitalfox.eu/archives/gemini/attachments/20200915/74ea
37f5/attachment.htm>

Link to individual message.

---

Previous Thread: Gemini proxy image display

Next Thread: Critique my setup, please