Re: Warning of major NHS IT overspend

[Peter Tomlinson <[EMAIL PROTECTED]>]

Dave Howe wrote:
> Roger Hird wrote:
>
> > Phase simply two proved too difficult, none of it ever happened,
> > and for years afterwards the central databases continued in use,
> > accessed through windows on the PCs emulating dumb ICL terminals.
> > Ex colleagues I've spoken to still think that trying to merge data
> > kept on a range of local systems, even if theoretically possible,
> > is extremely difficult and unreliable.
>
> Actually, it can be quite simple and reliable - how simple and
> reliable depends on how often the distributed data needs to be
> updated, how conflicting updates are resolved, and how often the
> schema which indicates what is stored where is changed. DNS (as used
> by the internet) is a good example of a distributed system - the
> schema sometimes changes, you can have multiple copies and cached 
> results stored all over the internet - but because only a single copy
> is permitted to be changed (and the official copies and cached copies
> take their cue from that) the changes ripple cleanly across the
> internet only rarely causing conflicts.
DNS is a disciplined (OK, self-disciplined) method where data is used 
all the time by messages coming from all over the world. Errors and 
service outages are quickly obvious and we all as users expect them to 
happen from time to time and be quickly fixed by the professionals who 
operate the nodes in the internet.

Health centre databases are complex, managed by people of varying 
degrees of competence and usually held in systems maintained by external 
suppliers. The outages and errors normally affect only the local service 
and cannot be tolerated by the remote A&E that in the distributed data 
architecture will be trying to access them.

No contest - but in the end its a political and (I hope) social decision 
which way to go.

Peter