f/f george <george@yourplace.com> writes:

On Mon, 30 Aug 2004 04:45:34 GMT, Johan Kullstam
<kullstj-nn@comcast.net> wrote:

fairwater@gmail.com (Derek Lyons) writes:

Johan Kullstam <kullstj-nn@comcast.net> wrote:

fairwater@gmail.com (Derek Lyons) writes:

Johan Kullstam <kullstj-nn@comcast.net> wrote:

fairwater@gmail.com (Derek Lyons) writes:

"John Rehill" <iwp52@hotmail.com> wrote:

I support any
new venture but this is ridiculous.  Didn't they have a trial period before
telling everyone to switch to BOINC?

What period do you think they are in *now*?

Something that comes before "trial period".

Must be a nice planet you live on.

It is a nice planet.  Won't you join us?

No.  Because on my planet they are in a trial period, having not told
everyone to switch to BOINC, and are not on drugs.

Perhaps SETI will find you then.  FWIW, I am on planet earth.

With a "trial period" I expect things to mostly work.  The past two
months have had things sometimes working but often not.  A rough guess
is about 50% working.  This is still development and debugging period
in my opinion.  Once it starts to work 90% it is time to call it
trial.

Your opinion is a valid one to you!
But I have a question....when you scale a working project up and it
fails due to hardware problems, what do you call it then?

I would call it perhaps "alpha test".  I think a "beta" would need to
be much closer to a final product.  Beta and trial are fairly
synomymous to me for this project.

Seti under Boinc is not failing due to programming problems, it is
failing due to hardware issues.

It is hard to say exactly what is failing.  I am glad that berkeley
reports on the website messages about progress -- I would have given
up long ago if they hadn't.  However, the messages are a bit terse and
cryptic and it is hard to know exactly what is what.

What is a hardware issue?  I haven't seen reports of CPU burnouts or
harddrive failure.  AFAICT, it is rather in the flow and coordination
of the various subsystems and how that scales up under loading.  It
may also be insufficient hardware in that they need more boxen to do
the work.

If the project is NEVER scaled up how
can anyone know that there will be problems? WE users were not advised
to switch, we advised the program was open for public business!

Sure enough.  We users are helping with development, scaling and
debugging of the system.

A HUGE
difference! We were told that as soon as it was working the way they
wanted Classic would close down. No time frames were mentioned. Some
were definately "assumed" but none mentioned! And no "SOON" is NOT a
time frame!

All of which is fine.  I just think it is a bit pre-mature and hasty
to call it a trial.  It's just not arrived there yet.  This should not
be considered a negative judgement on the value of the system any more
than saying a 3-year-old is a child is an insult to the 3-year-old.
It would still be unsuitable to call the 3-year-old an adolescent.

-- 
Johan KULLSTAM