[ih] the "cloud" (was Geek Terminology (was Re: Resource sharing))

[Miles Fidelman]

Bernie Cosell wrote:
> I think that the difference is in the software, rather than the hardware.
> Data is *always* stored somewhere.  We have several quite big MySQL
> servers with a lot of data stored in them and in the old-school sense,
> those servers act as the 'data center" for a lot of our applications.

as an aside.. yes, but not always in the sense of a specific instance of 
a datum, on a specific disk drive, attached to a specific machine - 
consider replicated data (e.g., usenet news, mirrored software 
repositories), or something like tahoe-lafs (where files are broken into 
blocks, with those blocks replicated and dispersed across lots of 
different systems, then reassembled on read)
>
> But "cloud" [to me] implies the invisibility of the data center -- it is
> "integrated" with the software on my local system so that (a) it
> *appears* to be local, and (b) I don't really know where it _actually_ is
> being stored.  Neither of those are characteristics of the traditional
> "data center".  Now, on the *other*side*, it is a huge distributed data
> center, but on _my_ side it looks like, well, part of my local system
> that happens not to be local.

I guess, to me, "in the cloud" implies something that's not local at 
all, to either my local machine, or to a specific data center.  For 
example, something (datum, process, service) identified and accessed  by 
URN rather than URI - with some invisible resolution process mapping the 
name to a specific instance (particularly where the instance is one of 
many equivalent resources, or where it's computed on the fly).

Or maybe that's just me.

Miles


-- 
In theory, there is no difference between theory and practice.
In practice, there is.   .... Yogi Berra