Hey Strata, I'm going to respond to a couple of points.
Strata Rose Chalup wrote:
> One man's "interference" is another man's "lower MTBF".
> The kind of P2P this author envisions is going to be a nightmare
> for support staff and result in lower productivity and more frustration
> for offices in general. An environment where desktop machines can
> contain critical data is one which is infinitely more difficult to
> back up-- there is a reason why there are file servers in an office.
Gawd Strata, you're so centralized. How does that backup
happen? Every file is collected and stored in a central,
fireproof place or every file is automatically bit-split, compressed,
and encrypted and stored on every other desktop? Invisible
backups with the only overhead of donating some spare computing
and network cycles to do the bit-splitting, compression,
and encryption. This seems like an easier way--they've been trying
to do central store and automated backup for years now. Not an
easy problem. Even Legato needs to install something at the edge.
> Desktops are also more susceptible to viruses and to users installing
> non-supported HW/SW on them (which usually causes trouble). The only
> nice thing about this currently is that only two people get screwed
> by this-- the desktop user him/herself and the poor bugger from IT or
> IS who has to untangle the mess. Just wait until critical documents
> are trapped on Joe Sixpack's machine when he nukes it with some lame
> drivers from who knows where.
What do you think is happening now? Where do you think the
information is already living? There's a big huge ugly truth
that IT people ignore and it's the fact that almost all of
the real data is living, breathing, and changing at the edges
of the network. Wouldn't you at least like to have some information
about what's out there? I suppose you are going to tell me ERP
systems are so successful because they centralize every piece of
data, business logic, and application logic into a big enterprise-scale
database that sits in a basement someplace.
> P2P is a great concept, but not necessarily for business-critical data.
> I think that the only feasible applications for P2P in key business
> areas are ones where individual desktops store A copy, not THE copy,
> of a file. Of course, then you get into locking and updating issues.
> I'll be surprised if there isn't a CVS pserver or similar lurking under
> the covers of the first P2P app to make it big in business.
The problem is, centralizing stuff is hard. P2P lets you do the
easy stuff first with only centralizing the metadata and leaves the
harder problem of migrating the rest of the stuff for later.
-- Gregory Alan Bolcer | firstname.lastname@example.org | work: 949.833.2800 Chief Technology Officer | http://www.endtech.com | cell: 714.928.5476 Endeavors Technology, Inc. | efax: 603.994.0516 | wap: 949.278.2805
This archive was generated by hypermail 2b29 : Fri Apr 27 2001 - 23:18:17 PDT