Chuck Ryan <[log in to unmask]> wrote:
> So, instead of fighting for an enhancment to an existing database, that is
> generally regarded as extremely fast and stable, we should just shut up,
> quit whining and use an unproven open source database?
>
> Perhaps I am the only one who is disturbed by the trend from HP, to avoid
> using resources to fix or enhance the existing native 3000 apps, and
instead
> hope that someone will port an open source solution. This allows them to
say
> that an option is available but not require them to support it.

The trend will only continue.  Development costs are almost completely fixed
costs and so tremendous advantage accrues to popular systems.  Unpopular
systems will fall further behind.

We should now appreciate what a good move it was for HP to aggressively
pursue POSIX compatibility for the 3000.  We would be without Apache, Perl,
and other useful tools without POSIX.  It has extended the useful life of
the HP3000.


> To me this begs the question, why do I need the 3000 to run open source
apps
> when I can run them, in the environment they were developed and tested,
for
> a fraction of the cost of maintaining the 3000? Particularly when , due to
> limitations of the posix implementation on the 3000, the majority of the
> ported apps are less stable with many features missing compared to builds
> running on linux.

I think the decision is complicated.  Many of us have substantial
investments in hardware that would be expensive to duplicate on a Linux
system.  A more standard relational DBMS might be desirable for just a small
application that needs to be integrated into an already-existing Image
application.