💾 Archived View for gemi.dev › gemini-mailing-list › 000806.gmi captured on 2023-11-14 at 09:45:33. Gemini links have been rewritten to link to archived content

View Raw

More Information

⬅️ Previous capture (2023-11-04)

➡️ Next capture (2023-12-28)

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

[tech] Plaintext Gemini extension

Artur Honzawa <arturh (a) gmail.com>

It probably has been proposed before, but this is how I could see it
working:

Serve rot13 encoded stream in port 1964 as fallback for clients without TLS
support.

Add gemini-plaintext: schema for servers without TLS support.

Low implementation burden overall, old hardware support. Thoughts?

Link to individual message.

Stephane Bortzmeyer <stephane (a) sources.org>

On Wed, Mar 10, 2021 at 09:18:14AM +0100,
 Artur Honzawa <arturh at gmail.com> wrote 
 a message of 25 lines which said:

> Add gemini-plaintext: schema for servers without TLS support.

Each time you have two security levels (encrypted and unencrypted),
besides added complexity, you have the problem of downgrade attacks
<https://en.wikipedia.org/wiki/Downgrade_attack>. These attacks have
plagued all protocols with both an encrypted nd unencrypted version
(SMTP?), that's why HTTP/3 (and Gemini!) only have one version.

Link to individual message.

Artur Honzawa <arturh (a) gmail.com>

Makes sense. Would security be impacted if clients applied TOFU and refused
to downgrade if they had previously established a secure connection?

On Wed, Mar 10, 2021 at 9:33 AM Stephane Bortzmeyer <stephane at sources.org>
wrote:

> On Wed, Mar 10, 2021 at 09:18:14AM +0100,
>  Artur Honzawa <arturh at gmail.com> wrote
>  a message of 25 lines which said:
>
> > Add gemini-plaintext: schema for servers without TLS support.
>
> Each time you have two security levels (encrypted and unencrypted),
> besides added complexity, you have the problem of downgrade attacks
> <https://en.wikipedia.org/wiki/Downgrade_attack>. These attacks have
> plagued all protocols with both an encrypted nd unencrypted version
> (SMTP?), that's why HTTP/3 (and Gemini!) only have one version.
>
>

Link to individual message.

Björn Wärmedal <bjorn.warmedal (a) gmail.com>

On Wed, 10 Mar 2021 at 09:18, Artur Honzawa <arturh at gmail.com> wrote:
>
> It probably has been proposed before, but this is how I could see it working:
>
> Serve rot13 encoded stream in port 1964 as fallback for clients without TLS support.
>
> Add gemini-plaintext: schema for servers without TLS support.
>
> Low implementation burden overall, old hardware support. Thoughts?

All current gemini servers and clients support TLS, and the presence
of encryption is a cornerstone of gemini. If you don't want it you can
use gopher instead :)

Supporting old hardware is a noble goal, but I'd say that's better
achieved by building a proxy service somewhere that serves gemini
capsules over http.

Cheers,
ew0k

Link to individual message.

Artur Honzawa <arturh (a) gmail.com>

Thanks Bj?rn, I totally get that. Nevertheless, isn't the protocol still
vulnerable to first connection MITM attacks?

On Wed, Mar 10, 2021 at 9:54 AM Bj?rn W?rmedal <bjorn.warmedal at gmail.com>
wrote:

> On Wed, 10 Mar 2021 at 09:18, Artur Honzawa <arturh at gmail.com> wrote:
> >
> > It probably has been proposed before, but this is how I could see it
> working:
> >
> > Serve rot13 encoded stream in port 1964 as fallback for clients without
> TLS support.
> >
> > Add gemini-plaintext: schema for servers without TLS support.
> >
> > Low implementation burden overall, old hardware support. Thoughts?
>
> All current gemini servers and clients support TLS, and the presence
> of encryption is a cornerstone of gemini. If you don't want it you can
> use gopher instead :)
>
> Supporting old hardware is a noble goal, but I'd say that's better
> achieved by building a proxy service somewhere that serves gemini
> capsules over http.
>
> Cheers,
> ew0k
>

Link to individual message.

Björn Wärmedal <bjorn.warmedal (a) gmail.com>

On Wed, 10 Mar 2021 at 10:13, Artur Honzawa <arturh at gmail.com> wrote:
>
> Thanks Bj?rn, I totally get that. Nevertheless, isn't the protocol still 
vulnerable to first connection MITM attacks?

Yup! :)

Link to individual message.

---

Previous Thread: [users] New capsule from sunny Barcelona: gemini://bcn08012.ddns.net/

Next Thread: [TECH] HtmGem v1.0.0, Php program to view Gemini through HTTP