Eh, did I hear my name being used in vain? :)
The question is about whether there would be a market for
relatively low-cost systems, preinstalled with Postgres and
OpenACS, right? Possibly to be sold by openacs.org.
I'm sure there would be a market for such a thing, if the price
was right. But there are a few things we've learned that may
help add some perspective to this (these apply to systems
purchasing in general and aren't intended to cast aspersions on
Talli's systems!):
- When it comes to hardware, cheaper is not always better;
sometimes it's just junk.
- It doesn't really matter if your system came preinstalled with
the latest software if updates are easy to get; even if it does it'll
be out of date soon enough.
- No hardware vendor is perfect, and systems will fail. Then
you find out that the quality of the support you get from your
vendor is as important as anything else you can think of,
including cost. Sometimes a bargain is no bargain in the long
run.
When we were ready to upgrade our hardware, we ended up
leasing from Dell. Their prices are good, though not fantastic.
Their service and support has only been adequate, though, and
their accessory prices are astronomical (you want *how much*
for a second processor??? :). In retrospect I think we might have
done better to go with a Linux provider, but Dell is improving and
the systems themselves are outstanding.
On the other hand, we recently helped with an emergency Oracle
install which was hampered by a very poorly configured VA Linux
system. I won't go into the gory details since I didn't see it first
hand, but suffice to say that it was delivered with major packages
missing, and when he convinced them to fix it, it ended up in a
worse state than before. This shows that paying top dollar can
be as foolish as going cheap, I suppose.
I need to add a caveat, however, that all of this only applies to
businesses which are small but somewhat established, as furfly
is these days. For the individual developer, all of the "big guys"
are out of reach. That's a potentially lucrative market for
OpenACS boxes, but then you have to think about support - the
person who buys one of these will do it because s/he doesn't
know how to install all this stuff and doesn't want to have to
know; they want to get right to coding. That's fine until a new
version of something comes out... then the calls for help will
start. And what if there's a hardware failure? It's my
understanding that arsDigita includes hardware and software
support by TechSquare.com with the RocketStart boxes, as well
as colocation. Something like this would be required, I think.
So where does all this rambling leave us? The short version of
my opinion is that yes, there would be a market for these
systems. But there are issues of software and hardware
support that would have to be addressed. Even if you tell people
you provide no support or warranty at all they will still expect it in
some situations, and it could become divisive within the
community.
Personally I prefer Michael's original idea of trying to get the
companies who are shipping Linux systems to include
OpenACS; the PR benefits are much greater. :)
