social
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Social] Fwd: GNU/social legacy


From: Melvin Carvalho
Subject: Re: [Social] Fwd: GNU/social legacy
Date: Thu, 13 Dec 2012 11:29:42 +0100



On 13 December 2012 01:53, hellekin <address@hidden> wrote:
On 12/12/2012 07:46 PM, Melvin Carvalho wrote:
>
>
> I dont have a problem with 404s , they show that the web is constantly
> changing.  It is hard for a system as big as the web could survive
> without something like a 404.  I like a degree of fault tolerance.
>
*** As much as I'm not interested, as a user, in the format of an image,
whether it's GIF, JPEG, or PNG, I'm not interested either on having my
information flow broken by obsolescence. The HTTP protocol contains
error codes to support redirections, and those should be used whenever
possible. But most of the time, they're not, digging holes in the
hypertext.

In contrast, Xanadu's design comes with unbreakable links, whatever you
can think about Ted Nelson or his project. So, it cannot be regarded as
a feature, but as a bug.

I know exactly what you mean.  But beware of over engineering a solution.  Facebook got big by not inventing anything new.  If we can just clone that model first, lots of exciting things can be added later. 
 

>
> I would prefer:
>
> address@hidden <mailto:address@hidden> at the User Interface level
>
*** The door open to spammers! The very fact that it *looks* like an
email means it will be harvested, and used, even if the chance that it's
actually not an email is non-zero: when it is, spam works.

Tg (from the Secushare project) commented recently that he "doesn't have
any issue with the identifier being the hash of a public key, as long as
the UI can alias it to something meaningful for the user, such as
address@hidden" I tend to agree with that statement. Dereferencing a
unique key is not out of scope.

The most meaningful ID to a user is their name. 

Agreed, the email camp have a spam issue.

Hash of a public key is fine but you need a robust dereferencing tech too.
 

>
> An HTTP profile page listing data about the user using web standards.
>
*** It works, as you know, for FOAF. One issue I see is that it requires
an always-on connectivity to be useful. Otherwise, how can you
dereference a non-accessible resource? Relying on third-parties is not
exactly the objective of decentralization. The Briar project might be
useful in that sense, as it relies on a pre-existing network of trust,
and uses cryptographic keys.

Remember we are decentralized.  You can take copies just like in git, that's what caching is.  Always on is also good, which is what people do today, or with freedombox. 
 

>
> Yes.  The trust third party model and lock in are often deployed by
> businesses.  I actually think businesses would do better in the long
> term by allowing more user freedom.
>
*** I'm with Jacob Appelbaum and Dmitry Kleiner on that one: the
business model is surveillance[1], and trying to fix it requires true
decentralization. Hail StatusNet :)

You talk about megacorps.  Did you look at the list of companies behind the OStatus spec? 

One of the authors of webfinger recently wrote this publically

[[
You should call it "MicrosoftGoogleFinger", because those are the only relevant players here. The IETF is hopelessly fucked, because there is no meaningful engagement with startups. This specification is as unlikely as ever to be adopted
]]

Personally, I have no problem at all working with companies big or small, so long as they make great technologies.  
 

>
> I dont believe there is 'one social protocol to rule them all'.
>
*** I like diversity too :)

>
> Lorea is one of my favourite projects, is caedes still on the team?
>
*** Definitely yes. He's not so present in the public because he's busy
with the local life, and prefers spending his computer time coding.

Awesome, one of the smartest (and nicest) people I've had the privilege of working with. 
 

> You'll have to put me down as a heavy skeptic of OStatus.
> If it gets approved by any recognized standards body
>
*** Bittorrent didn't come from a standards body. I'm wondering why you
trust them to come up with the right stuff, especially as they're
dominated by megacorps, with obvious interests against decentralization.
The recent ITU conference in Dubai didn't show much progress in that
direction.

First and foremost the two big standards bodies (W3C and IETF) have a stellar record for keeping specs royalty free.

When specs are not under a standards body, and have to go through lawyers to get published (as with activity streams), that's just one extra layer of complexity.

A great tech does not need to come through a standards body, but for every one that makes it outside, there are 99 that fail. 
 

> or demonstrates good degrees of interop with systems other than
> itself, I'll be very happy to change opinion.
>
*** I prefer that part :)

Any system that is serious about interop and can demonstrate it should be considered. 

For example, to set up an HTTP POST enpoint that can send and receive messages from heterogeneous systems, should not take any longer than a day to program.  This seems like a logical starting point.
 

==
hk



reply via email to

[Prev in Thread] Current Thread [Next in Thread]