Briefly, FYI, one of my little tasks is to add session affinity to
Continuity.  Given a pile of servers running Continuity hosted apps,
a connection to the wrong one should get you transparently proxied to
the correct one for your session where your persistent execution
context lives.  This doesn't do anything for giving you truly global
data across all nodes.  That's a seperate project.  I imagine
global, coherent data syncronized with forks.pm between nodes across 
an OpenSSI cluster, but that would be a hard sell.  

I have to say that "cloud computing" is a lame metaphore, a leaky
abstraction, and something appealing for romantic reasons -- running code
on Google's servers -- rather than practical reasons.  Building and
colocating servers is generally the least concern of a significant
web effort, and not having control, if it were a concern, would be
a large one.  Hitching your wagon to Google App Engine's horse
would be a major commitment with no easy exit stragetegy.  Two
programmers can't even agree on an ORM.  This is the modern counterpart
to Geocities.  Not to discourage you from your task too much -- do 
it if you want to, but as they say, beware the hype.

Hmm... adding some map/reduce crap to Continuity would be interesting.
And I guess I should be thinking about a non-blocking interface to
that distributed cache thing everyone is using...

Good luck...

-scott

On  0, Stephen Adkins <[EMAIL PROTECTED]> wrote:
> Hi,
> 
> I haven't sent anything out through the P5EE mailing list in a couple of 
> years.
> I'm using it now to raise a related issue.
> 
> "Cloud Computing" has arisen as a powerful new paradigm (and buzzword) [1].
> So here are my questions for discussion.
> 
>   1. Does anyone know if anyone in the Perl community is working with Google
>       to get Perl certified for use in Google App Engine? [2]
>       (It currently only supports Python.)
> 
>   2. If I were going to try to organize a project ("Cloud Computing on Perl"),
>      are there others who would like to participate? (I would imagine that 
> such
>      a project would address a variety of cloud computing platforms [3] [4]
>      including Free Software DIY clouds.)
> 
>   3. If I were organizing this, does anyone have any particular 
> people/projects
>       with a vested interest to suggest that I should coordinate this with?
> 
> Stephen
> 
> [1] http://en.wikipedia.org/wiki/Cloud_computing
> [2] http://code.google.com/appengine/docs/whatisgoogleappengine.html
> [3] http://www.amazon.com/gp/browse.html?node=201590011
> [4] http://en.wikipedia.org/wiki/Cloud_computing#Cloud_services

Reply via email to