response 181 of 271:

Jul 8 02:26 UTC 2002

A lot of people here have nearly as much experience as Joseph does.
I've been involved with the BBS community, in various capacities, for 19
years.  I'm glad m-net has found a solution that works well for them.  I
also think that, over the years, m-net has evolved into a distinctly
different system, and certainly in the past they've made some
horrendously bad decisions.  So I don't think they have any claim to
having the "best" answers, nor do I necessarily even agree that what
works well for them today is the best answer for us.  I respect their
decisions, and I think they're well worth watching.  I don't think that
means we need to jump ship hastily, even if we think they've got some
good ideas.

Just to take the "colocation" example; with m-net's paradigm, if a
hardware or serious software failure occurs, one staff member goes to
the site, removes the hardware, takes it home, and hacks on it.  When he
has working hardware, or a suitable replacement ready, he then tranports
this back to the colocation, plugs it in, and hopes that it all works.
This seems to translate into a guaranteed outage of 48 hours and, to my
recollection at least, has happened several times in the past year.
Obviously this has not been a big problem for m-net, and they seem happy
with the results.  Grex's antique hardware & software has in fact
managed to deliver superior reliability.  There is a reason we like sun
hardware.  Nevertheless, if a serious problem happened, we have the
ability to fix it "in situa", without any need to remove it elsewhere,
and we also have the ability for multiple staff people to easily and
conveniently work on it in a "common" spot, on a 24 hours around the
clock basis, if necessary.  We've done pretty much exactly this with
failing disk drives sometime "recently", ie, in the past few years.
Going to m-net's paradigm would seem to mean we need to switch to
smaller, more expensive, and less reliable hardware, might get better
network connectivity, might save on costs, but might lose on staff
access & increased downtime.  All of these things could be wrong; grex
staff might be smart enough to select even more expensive but reliable
small colocation requipment.  We might get a "deal" on space or access
in which case it might be even cheaper and come with "repair desk"
access.  Still, the tradeoffs are many and complex; I don't think we
need to feel we've made the wrong choices in the past or that what
worked for m-net would have worked for us.

response 189 of 271:

Jul 8 07:14 UTC 2002

Good grief-- an item that's usually mundane suddenly becomes the most 
pathetic and drawn out argument.  Pitiful.

response 191 of 271:

Jul 8 14:22 UTC 2002

Re #188:  Maybe I don't understand some aspects of how Grex works.  If the
          Arbornet treasurer had posted #24 (go back and look), that would
          have looked like an official fundraising effort to me.  Seriously.

          But this is really a side issue, and I'm willing to drop it.

response 193 of 271:

Jul 8 14:44 UTC 2002

Oh, come on.  It's been done once or twice.

response 195 of 271:

Jul 8 14:49 UTC 2002

Not when I was President.

response 197 of 271:

Jul 8 15:13 UTC 2002

That last sentence was a lame attempt to drag that response on topic.

response 199 of 271:

Jul 8 16:03 UTC 2002

Let's not discuss M-Net policy here, okay?  Other than what pertains to 
Grex, anyway.

response 205 of 271:

Jul 8 21:47 UTC 2002

Thanks Dan.

My question now is:

What is the question that staff is divided on?  Can a staff member write a
relativley simple statement of that question for the rest of us.