Re: <nettime> Direction of Travel

[Rich Kulawiec via Nettime-tmp <nettime-tmp@mail.ljudmila.org>]

On Wed, Jun 14, 2023 at 08:35:11AM +1000, paul via Nettime-tmp wrote:
> I definitely appreciate your concerns.  Having said that, i don't know if
> it's realistic for us to predict the future that way. Riseup could disappear
> tomorrow, the server set up by a whizkid volunteer could disappear tomorrow,
> and presumably even Panix could disappear tomorrow.  100% agree on running
> with our own domain name though, that's table stakes to be able to migrate
> seamlessly and maintain control of our destiny.

[ good content snipped for brevity; I agree with pretty much all of it ]

I dislike the term "whiz kid": running a mail system with a mailing list
manager is something that any/every competent sysadmin should be able to
handle, just as they should be able to compile sendmail from the source
code, install it, configure it, and run it.  These are baseline skills.
Same for BIND or Apache HTTPD, postfix or NGINX,  or other things.

There is an unfortunate and accelerating trend toward learned helplessness
and reliance on third parties.  This is short-sighted and foolish: it
leads to becoming a sharecropper on other folks' platforms and that
choice has inevitable consequences.

For example: note how many journalists/news organizations used Twitter
to communicate with their audiences.  Some of them invested a lot
of time and effort in attracting followers and interacting with them.
And now?  Now that sociopathic moron Musk has taken over and is clearly
intent on burning it down, all of that is being destroyed.  Similar things
are happening at Reddit and Substack.  They will happen soon enough at
Bluesky.  Of course they are, it's inevitable that their owners will all
race to the bottom in an attempt to maximize short-term profits.

The only way to win this game is not to play it, or to play it as little
as possible: minimize dependencies on third parties.  The two things you
can't avoid are (a) a registrar, for domain registration, and (b) a host,
to run a system.  But everything else can be ditched OR, if necessary,
used in a fashion that allows it to be quickly changed/discarded if/when
conditions change.  For example: there's a very good (and free) DNS
secondary service run at puck.nether.net.  I've used it in production for
years and it...just works.  (Unsurprising given that the guy running is a
very experienced network engineer.)  But suppose one day it goes down and
stays down: first, everything that's using it will keep working because
other DNS servers are available; second, removing it from the domains'
registration and zone files will take less than an hour; and third,
it's unlikely any end users will notice that any of this has happened.

Design/implementation like this takes some thought, but the payoff
is that services can be made quite resilient.

---rsk
#  distributed via <nettime>: no commercial use without permission
#  <nettime>  is a moderated mailing list for net criticism,
#  collaborative text filtering and cultural politics of the nets
#  more info: https://mail.ljudmila.org/cgi-bin/mailman/listinfo/nettime-tmp
#  archive: http://www.nettime.org contact: nettime@kein.org
#  @nettime_bot tweets mail w/ sender unless #ANON is in Subject: