Re: Thoughs about communication

[Emmanuele Bassi <ebassi gmail com>]

Hi;

On 13 January 2017 at 12:45, George Barrett <bob bob131 so> wrote:
> On Fri, Jan 13, 2017 at 08:37:30AM +0000, Alberto Fanjul Alonso wrote:
> > Hi, hackers
> >
> > Do anybody though about trying new services for communication?
> >
> > - signal https://whispersystems.org/
> > - telegram https://telegram.org/
> > - matrix.org http://matrix.org/
> > - gitter https://gitter.im/
>
> None of these platforms have an official Telepathy protocol implementation.
> That people will have a lessened ability to access chat already seems like a
> non-starter.

Considering the state of Telepathy, this is not at all a blocker.
Actually, it may very well be the push needed to phase it out from our
platform.

> > pros/cons irc:
> >
> > pros:
> > <snip>
> > - is widespread
> > - integrated in gnome environment (bots, bugzilla)
>
> I would argue these two features are critical to any prospective chat
> platform. If people can't access chat in a way that suits their workflow, they
> probably won't. And it'd be a step backwards if automation suddenly became a
> stumbling block.

I guess the issue, here, is whether or not we care about reaching
other people than the existing pool of contributors.

> In terms of universality, the only chat platform rivalling IRC (that I can
> think of, at least) is XMPP. I don't know enough about it to seriously
> recommend it, though; does it support the requested features?

XMPP is somewhat a "nerd ghetto". In principle, it's awesome: a
federated protocol with capabilities negotiation that has free
implementations and a set of extensions.

In practice, it's a disaster zone — with terrible server
implementations, and a tendency for clients to either assume a certain
set of extensions and running against specific servers, or providing
the minimum common denominator because extension and capabilities
discovery is terrible.

All existing chat protocol silos — like Google's, WhatsApp, Signal,
Telegram — usually start off as some sort of XMPP protocol, and then
rapidly close it off because it's impossible to gracefully degrade
functionality and UI at the same time across a wide range of services
without giving the users something that only geeks, with their higher
pain threshold, would use.

> > Solutions by new technology:
> > - gitter can deal with integration
>
> So can IRC, thanks to an easy-to-parse protocol and an abundance of libraries.
> XMPP, too, has quite a few libs available.

The reason why IRC is even a contender is that it's a trivial
protocol. The problem is that it's *too* trivial, and does not provide
features without resorting to things like proxies, bouncers, or bots.

> > an syntax highlight
>
> Is syntax highlighting done in the browser, or baked into the protocol? I'm
> going to go out on a limb and assume it's the former. For many people (myself
> included), web chat is a non-starter. Assuming others in the community
> feel the same, is there any point moving to a new platform for extra features
> that many won't be able to use?

There's a point in moving to a new platform if people will be able to
use new features.

If you can use a browser with HTTPS then you're likely going to be
able to use IRC on a web form, or better user experiences.

If you cannot access anything outside of IRC then you're living in a
severely constrained network, managed by someone who doesn't really
want you to join online communities, and IRC is just a way to work
around those cases. Some platforms, like Slack, do offer IRC bridges
as well, in any case, with degraded functionality.

> > - signal is aware of privacy
>
> What does this mean? If the plan is that public chats are logged, is there any
> room for privacy considerations? I'm probably misunderstanding, but being
> privacy-aware seems moot in this instance.

Public channels can be logged, but private chats can be secured. These
are not conflicting goals.

> > - matrix.org can be bridged to gimpnet in full two-way communication
>
> I find this the most convincing argument, but if we're just going to keep
> using IRC anyway then is it worth the cost? That question isn't (entirely)
> rhetorical; it might be, but the benefit isn't immediately obvious to me.

Just because you found yourself happy with IRC it does not imply IRC
is without fault, or without features that are, indeed, required by
other people.

If it doesn't cost you anything to stay on IRC while other people move
to different standards — assuming there's a bridge, at the very least
as a transition mechanism — then asking "what is the cost" is just
stop energy.

Ciao,
 Emmanuele.

-- 
https://www.bassi.io
[@] ebassi [@gmail.com]