[LUAU] Hello - My Name is HOSEF

[Jim Thompson]

On Mar 26, 2008, at 10:59 PM, Angela Kahealani wrote:
> On Wed, 2008-03-26 22:16:23 Jim Thompson wrote:
>> I don't think the current setup with the Executive Director is
>> working, nor is healthy for the organization.
>
> ...and your proposed alternative is?

to eliminate the position, put Scott on the Board (if he and the other  
two board members agree) and proceed forward.

>
>> I'll repeat what I said earlier; you can fix a watch by replacing its
>>   parts, but replacing parts of HOSEF is unlikely to effect repair of
>> what ails it.
>
> what ails it? what is the proposed solution?

There are many things that ail HOSEF.

It is underfunded for its current mission.
It faces increased competition for its existing mission from much  
larger, better capitalized competitors.
Its "Executive Director" ignores its board of directors, putting the  
organization in peril.

this is not an exhaustive list.

>> To that end, I have also proposed that the mission of HOSEF needs to
>>   change away from its focus on ecycling and "charity" work.  I
>> honestly like the suggestions that the group focus more on business
>> and I'd like to add government to that suggestion.
>
> so, it sounds like what is needed is to recruit motivated people to
> start-up those interests, which doesn't have to mean Scott stops doing
> what he does, does it? I seem to recall some recent discussion about
> law, lawyers, etc... but am at a loss to understand what the problem
> is... HOSEF is a 501(c)3 corporation, and that as far as I know  
> doesn't
> prevent members/board from getting salaries/wages, only that the
> corporation not be "for profit"... have I missed something?

What you have said here is accurate, but yes, you've missed several  
points.

>
>> I'm not excited at the though of a LUG, but some kind of local
>> 'networking' would be great, as Hawaii doesn't seem to have much of
>> that.
>
> so what do you propose beyond the HOSEF and LUAU mailing lists?

Events like TPOSSCON/PFOSSCON on a more periodic / more intimate basis  
would be a start.

>> I'd like to see more Solaris and BSD in the mix.  I'd like to reach
>> out to the Apple community (Jordan Hubbard is an acqaintence, and he
>>   was spun up to speak at TPOSSCON/PFOSSCON this year, though that
>> didn't happen.)
>
> So, education via conferences is good. or via websites... or...?

Perhaps "unconferences"  http://en.wikipedia.org/wiki/Unconference

>
>> I actually arranged for all three speakers for the last PFOSSCON.
>
> Thank you!
>
>> After nearly 30 years in the Unix/Linux community, I probably know
>> others who would be willing to travel to Hawaii to speak.  Perhaps
>> you   do too.
>
> I've only been playing with UNIX for 26 years, but have lost contacts.
> Isn't it interesting that *all* the FLOSS is based on UNIX?

Not all, no, but most.

> MacOSuX, BSD, Linux... all be cousins. Meanwhile I'm playing with
> setting up the KLH-10 emulator on Linux to run the Panda distro of
> TOPS-20, which is now freely licensed for hobbyist use. Not quite
> as FLOSSy as *NIX (no commercial use), but interesting nevertheless.

Funny.  I run the Symbolics Virtual Lisp Machine (VLM) port to Linux/ 
x86_64

> Many of the ideas embodied in *NIX came from PDP-10/TOPS-[12]0/TENEX,
> and there is actually some current development work for it (FTP).

Welll... not really.  Most of Unix came from Multics, and Multics also  
heavily influenced TENEX, which became TOPS-10.

TOPS-10/TOPS-20 were eventually replaced VMS, and VMS sucked so hard  
that most of the TENEX/TWENEX hackers ended up on Unix.

See: http://www.opost.com/dlm/tenex/hbook.html

>> I'd like to reach out to honolulu-ciders, and the fledgling local
>> Ubuntu groups.  Some cross-polination could be interesting.
>
> Are there links to these on the HOSEF website?

Both have been mentioned on mailing list.

>> I'd like to approach Bishop Estate, OHA, and Kamehameha Schools about
>>   restarting kuokoa, a proposed Hawaiian language Ubuntu
>> distribution. (I've been building contacts on this front for two
>> years now.)
>
> Excellent! I understand that UH has done some work in the direction of
> Hawaiian fonts. Aren't we still lacking an ISO code for O'lelo  
> Hawai'i?
> Meanwhile, don't leave out the Hawaiian Charter Schools or Immersion
> Schools.

Those too, as I explained, it was painful typing with my thumbs (while  
eating dinner).  This was all discussed back
when I first proposed kuokoa.

>> I'd like to see if we can't activate a bit more scientific computing
>>   at UH. I'm aware of a few small clusters, but none of these seem to
>> reach the sophistication of the ones I've used on the mainland.
>
> Does the Maui SuperComputer Center have any spare CPU cycles to  
> donate?

Given that its about DOD activities, thats highly unlikely.
See: http://www.mhpcc.hpc.mil/about.html

you may wish to view last year's research subjects: http://www.mhpcc.hpc.mil/research/appbriefs/2007 
  and note that 'tracking Humpback whales' isn't about studying them  
as much as it is avoiding them so the Navy can continue to use its  
high-energy sonar, never mind "Theater UnderSea Warfare" (cough cough).

> Is it time for grid computing at UH to crunch science during off  
> hours?

latency rears its ugly head for the second time in this message!

>> I'd be very interested in discussions and demonstrations of "greener"
>>   computing, especially as it leverages the "open" nature of FOSS.  I
>> think it possible that computers that use less power could be better
>> for the environment than recycling existing machines, simply because
>> there wasn't much emphasis on power saving in earlier generation
>> chipsets, and we burn oil here for nearly every kilowatt, and will
>> continue to do so for many years.
>
> Excellent! But I thought you wanted to get away from hardware?

No, I want to get away from eWaste as a HOSEF activity.

I am somewhat concerned that Scott is now focused on "P4" generation  
hardware as the current donation stream, and that the availably of
some has fulfilled some dream.  Quoting a recent message of his:

"The average donation to HOSEF is a high-end PIII, and lately we seem  
to get more P4s and LCD monitors than I have counted. "

Pentium-m and Centrino M CPUs support C2 "Deep Sleep" states, P4  
'desktop' CPUs, to my knowledge, do not.
See: <http://download.intel.com/design/intarch/designgd/25131902.pdf>

C2 is pretty deep—for example, the chip can't detect or respond to  
interrupts, and current draw is cut down to about 3.4 watts on a Core  
2 Duo. (Mode C0 is normal full-speed, full-power operation. Mode C1 is  
enabled when one core hits a halt, and stops processing temporarily;  
the other core watches its bus. C1 mode runs the chip at full voltage/ 
frequency, but it consumes less current.)
The new Enhanced Deeper Sleep mode on Core 2 Duos takes the power down  
even farther with two more modes; without getting into the details, C3  
and C4 keep dropping the voltage/frequency down as far as they can go.  
The C4 mode even shuts down the L2 cache. C3 represents a processor  
current draw of 2.2 watts, and the C4 mode is a mere 1.8 watts.

When you consider all the time that these machines spend idle in the  
labs that HOSEF has setup, you start to see a real impact on the  
environment.  Putting a bunch of desktop P4s in the mix would be even  
worse.

For comparison:
The Mobile P4 used 70 W of power
The full Pentium 4 used about 82 W
The Mobile Pentium 4 M used about 35 W

A full listing of Thermal Design Power for various chipsets can be had  
here: <http://en.wikipedia.org/wiki/CPU_power_dissipation>, read it  
and weep once you understand what an extended life P4 might accomplish  
in terms of actual power consumed.

So you have a choice to make:

1) P4 computers in landfills (bad for the environment)
2) recycled P4 computers in classrooms (short-term: good for the  
environment (reduced landfill pollution) long-term bad for the  
environment (increased CO2 emissions)
3) P4 computers that enter a well-funded eWaste program (long-term: OK  
for the environment, short-term: still have to "pay back" the  
environmental costs associated with building the computer)

funny how this isn't advanced on-list, or reflected in any of Scott's  
verbiage, eh?

I like choice three, if it isn't obvious.

Now lets do a little math.

Lets say the computers are used 8 hours / day.  I think this is  
generous given that most school days aren't 8 hours long.

Watts used by a Core 2 Duo CPU during a 30 day period (idle 16 hours /  
day in a deep sleep mode.
8 * 30 * 31W + 16 * 30 * 1.8W = 8.304 kWh
24 * 30 * 82W = 59.040 kWh
59040/8304 = 7.1

The P4 desktop machine is 7X as expensive to feed electricity, in part  
because it can't enter any real "power save" modes, but mostly just  
because its a pig.
Run the numbers again without the Core 2 Duo entering *any* sleep modes:
24 * 30 * 31W = 22320 kWh
59040/22320 = 2.645

Its still 2.6X as expensive to feed power as a more modern CPU (which  
is faster at computation as well.)

Now multiply by the number of CPUs found in a typical lab (lets say  
20), and you end up with a P4 lab using 1181 kWh .vs a C2D lab using  
166.080 kWh during the month.

At $0.20 / kWh (the residential rate in 2006, I don't know what the  
schools pay), thats $33.22 .vs $236.20 per *month* just in power bills  
for that 20 station lab.  In a nine-month school calendar year, (I'll  
assume the lab is shut-off during school breaks) you're looking at a  
$1826.82 delta just for power use.   Do those computers look "free"  
now?  Each one costs $91 more to power...

never mind the effects on the environment.

For more dramatic effect, multiply by the several hundred (if not  
thousand) P4 CPUs that HOSEF might put back into use, rather than  
directing them to effective ecycling.

Two word summary:  "false economy".

And I have yet to touch the other power savings to be had in more  
modern systems (that run a lower voltage bus, support chips that can  
sleep, etc.)

>> I think recycling and HOSEF are not compatible.  Touching hardware
>> isn't scalable, making copies of software (distros, c++ source code,
>>   bash and perl scripts, etc), is and is as well more appropriate for
>> an organization  of HOSEF's size and financial resources.
>
> ...but is HOSEF necessary for that? anyone can download an .iso of a
> variety of Solaris, BSD, or Linux distros and burn them to a CD or
> DVD... and HOSEF/UH used to have an ftp server to download from,
this still exists

> but  even lacking a local server, what's the need for that?

You appear to not understand how latency affects TCP.

> Actually, downloading and burning an .iso to optical media is  
> trivial even for a
> Windblows user, the challenging part is doing the install of FLOSS  
> from
> the optical media onto their computer, backing-up their data first,  
> and
> now you're back into non-scalable support of individual computers
> (hardware), and trying to educate users (the alleged agenda of HOSEF).
> So, if HOSEF doesn't shuffle hardware, then the next best thing is
> InstallFests?

Not really, no.   Education and Advocacy aren't "installfests".

>
>> I'd like to sit on a board of equals, one that isn't constrained by
>> or   subject to the force of one personality.  I'm not fond of cults,
>> and don't wish to spend my time or money on one. To be clear, I don't
>> want to be the cult leader, either.  I view the board seat I hold as
>> being of service to the organization and its membership.
>
> Members? HOSEF has members? what does that mean?

you pays your dues, you is a member.

>
> One doesn't have to be a member to subscribe to the mailing list,  
> right?
nope
>
> So what does membership bring beyond the mailing list? Benefits?

voting in the general meeting, for one.

> Obligations?

None, really.

>
>> Jim
>
> Thank you Jim, for taking the initiative to start-up some new energies
> for moving HOSEF in new directions. But, is it necessary to shut down
> old directions at the same time? Legally? Financially? Operationally?
> Is it time to share why and how the board came to its' conclusions?

The HOSEF board hasn't come to any such conclusion on the subject.

Why do you assume it has?

It does recognize the challenges that lay along the present course,  
and is considering a change of course.

The more pressing matter is Scott's recent behavior.

Jim