IPv6 in the gopherspace

Mateusz Viste

unread,

Dec 8, 2021, 10:17:09 AM12/8/21

to

Today, out of curiosity, I decided to look into IPv6 deployment
within the gopherspace.

According to OGUP's data[1], out of the 271 currently known active
gopher servers, 59 are IPv6-enabled. That is roughly 22%, which is
slightly more than the 20% of IPv6-enabled websites reported by
w3tech[2] this month. Isn't that cool?

A less positive information is that out of the 59 IPv6-enabled gopher
servers, over 10% are routed via the HE tunnel broker, meaning they
likely do not have native IPv6. Well, the world is not perfect I guess.

[1] gopher://gopher.viste.fr/1/ogup
[2] https://w3techs.com/technologies/details/ce-ipv6

Mateusz
--
gopher://gopher.viste.fr

Marco Moock

unread,

Dec 8, 2021, 2:14:52 PM12/8/21

to

Am Wed, 8 Dec 2021 16:17:07 +0100
schrieb Mateusz Viste <mat...@xyz.invalid>:

> According to OGUP's data[1], out of the 271 currently known active
> gopher servers, 59 are IPv6-enabled. That is roughly 22%, which is
> slightly more than the 20% of IPv6-enabled websites reported by
> w3tech[2] this month. Isn't that cool?

It definitively is.

> A less positive information is that out of the 59 IPv6-enabled gopher
> servers, over 10% are routed via the HE tunnel broker, meaning they
> likely do not have native IPv6. Well, the world is not perfect I
> guess.

It is much better than no IPv6 connectivity.

In future some networks will be IPv6 only. If every server is reachable
via IPv6 (regardless of native or tunneled) such networks can be
properly used, because the users mostly don't care how the data is
being transferred.

Mateusz Viste

unread,

Dec 9, 2021, 4:41:00 AM12/9/21

to

2021-12-08 at 20:14 +0100, Marco Moock wrote:
> > A less positive information is that out of the 59 IPv6-enabled
> > gopher servers, over 10% are routed via the HE tunnel broker,
> > meaning they likely do not have native IPv6. Well, the world is not
> > perfect I guess.
> It is much better than no IPv6 connectivity.

Nah. As of today, jumping through such hoops is only a nuisance. It has
exactly zero benefits compared to IPv4-only, but brings its share of
troubles, extra maintenance and performance hit.

> In future some networks will be IPv6 only.

Maybe, or maybe not. Who knows. Users don't need a public IP address,
and they mostly don't care about anything as long as y**tube and
f***book works.

> because the users mostly don't care how the data is being transferred.

Exactly, yes. Hence from a link provider, if CGNAT works well (and it
*does* work exceptionally well), why bother with IPv6? And again - I am
not advocating here against IPv6. On the contrary, I find it a nice
step forward. But that's technical puritanism, the real world ain't so
nice and easy.

Mateusz

Marco Moock

unread,

Dec 9, 2021, 1:40:41 PM12/9/21

to

Am Thu, 9 Dec 2021 10:40:59 +0100
schrieb Mateusz Viste <mat...@xyz.invalid>:

> Nah. As of today, jumping through such hoops is only a nuisance. It
> has exactly zero benefits compared to IPv4-only, but brings its share
> of troubles, extra maintenance and performance hit.

The benefit is that these are required steps to reach IPv6 only. There
is no way around it if we like to keep the internet as we know it -
user can run their own services and don't need to rely on big tech
companies to store their data etc.

> > In future some networks will be IPv6 only.
>
> Maybe, or maybe not. Who knows. Users don't need a public IP address,
> and they mostly don't care about anything as long as y**tube and
> f***book works.

90% of users don't need it, but I like to keep the possibility that I
can run my own server that I can control and not a "cloud" provider.