Tim,
Welcome to the discussion. It is good to see you here.

Your note suggests that PSHB was motivated by Twitter... I'm not Brad or
Brett, so I don't know their specific motivations, however, I think it would
be fair to point out that PSHB is really just the latest in a long line of
similar efforts. Brad, at least, has been active in this area for years --
having first implemented what is now known as the "SixApart Update
Stream<http://www.sixapart.com/labs/update/>"
while he was developing LiveJournal. (The update stream delivers a stream of
atom-based Fat Pings.) He, and others on this list were also quite
supportive of the various efforts to get shared ping servers running (
Weblogs <http://weblogs.com/>, Pingomatic <http://pingomatic.com/>, etc.).
And, we should also remember the that FeedMesh effort has resulted in the
sharing of Fat Pings between Yahoo, Google, Pubsub.com, and others for quite
a few years. (The "Fat Ping" term became established during the FeedMesh
discussions I believe). Check out Blo.gs <http://blo.gs/> that was in Yahoo
for some years but was recently transferred to Automattic... Of course,
Winer will also want us to remember that he proposed rssCloud quite a few
years ago. And, there were also the various XMPP PubSub efforts such as
PubSub.com as well as systems built by Ralph Meijer in the Netherlands.

So, I think it is inappropriate to view PSHB as only a response to anything
recent. This is really the culmination of many years of work and many
different attempts to make this system work.

Also, hopefully the large number of responses to your posting that said
"PSHB isn't solving the last-mile" problem convince you that not being able
to reach desktop clients is an intentional design element -- not a bug or a
weakness. Previous efforts, like FeedMesh, have clearly demonstrated that
having a server to server network is exceptionally useful and doesn't do
anything to reduce the opportunities for creativity in solving the last-mile
problem. In appears that the server-to-server problem is relatively well
undestood and can be addressed with common solutions. By doing this, we
allow client software developers to experiment and innovate without having
to worry about the better understood backend problem. The intention is, I
hope, that PSHB will solve one problem and solve it well. Other problems
should be solved elsewhere.

bob wyman

On Mon, Oct 19, 2009 at 2:20 AM, Tim Bray <[email protected]> wrote:

>
> http://www.tbray.org/ongoing/When/200x/2009/10/18/PubHubSunday
>
>  -T
>

Reply via email to