[BLDG-SIM] hornets nest?

[Jeff Hirsch]

A couple comments on those of others ... fuel for the fire, or stir up the hornets

Quote: "First let me state that from all available evidence, the whole DOE (and here DOE is the name of the program - what a strange
choice for program name, BTW) effort is one of the most successful undertakings in a brief history of computer program development
and represent well spent government dollars by any standard."

Comment: First, all the contributors to DOE-2 thank you for your kind words.
The name was chosen, against developers advice, by USDOE, shortly after the
creation of the Department, to replace the programs existing name;  previously
the program was named CalERDA for the Calif. and ERDA (one org. moved
into the new DOE) who supplied the early funds.  DOE-1, DOE-2, DOE-2.0A,
DOE-2.1A, DOE-2.1B and DOE-2.1C (1977-1986) were mostly paid for by
DOE with small contributions from others.  DOE-2.1D (~1988) had smallish
percentage of private contribution.  DOE-2.1E (~1994) was about half private
contribution (mostly my company) and DOE-2.2 (current) is mostly private
contribution (mostly my company).  DOE-2 would not have come into existence
without the State and Federal government very large investment; surprisingly large
compared to similar private projects of the time; comparing public vs. private
contributions must be done using project output not $ input.

DOE-X only "took off" after private firms adopted and invested heavily: first
just PC ports like the defunct SI, Inc turned into Acrosoft, and ADM; then our
PC ports with many new HVAC/Economics/EndUseReporting 2.1E versions;
then interfaces Comply24, VisualDOE, EZ-DOE, etc.  This ivestment only happens
if firms can make $ from such efforts.  Such return on investment is not possible
unless licensing gives those investing the time and $ the ability to benefit from their
investment.

So, we owe DOE and CA/CEC thanks for making it possible that DOE-x exists, but
owe the thanks to the private sector companies that took the somewhat deflated
ball, added much air (hot at times) and ran with it.

Quote: "It seems to me that there is a quite a bit of confusion about PowerDOE, which is in my opinion, very good program.  PowerDOE
is nothing more than good pre and post processor, which can have DOE 2.2 or EnergyPlus as underlying engine."

Comment: Again, the PowerDOE development team thanks you for the kind words.
PowerDOE interface code was developed by my company with $ only from EPRI
and our own investment (85% vs. 15%). LBNL participated, with many other
organizations, in design meetings for the PowerDOE user interface. The PowerDOE
interface is not just a "pre and post processor" for DOE-2; it is a full interactive
implementation of DOE-2's BDL processor (which is somewhat of a table driven
table generator whose internal design is only applicable to just DOE-X, but any
application that requires extensive input organized into object tables.)  Editing
PowerDOE data screens is editing BDL memory in real time including the dynamic
recalculation of parameter defaults (like a very large spreadsheet with a formula in
every cell; thus the slow down - geometric- as problems get very large.) The
PowerDOE name was suggested by LBNL as they did not like our original winDOE
name; PowerDOE name was okayed by DOE, then accepted and trademarked by
EPRI (oops! - an poor choice of move motivated by political infighting amoungst the
EPRI/DOE/LBNL/JJH team).  I have licensed the use of the name and EPRI owned
interface code from EPRI for further development and commercial distribution.

Quote: "EnergyPlus is an open source program, and alpha version is just being released.  For measly $100 (I think) anybody can get
license agreement and build commercial strength pre and post-processor to it. "

Comment: EnergyPlus is NOT open source; but it should be, in my opinion, using UC's
(LBNL is managed by UC for DOE) own BSD (Unix) type license since it is developed
(almost?) entirely using public funds.  Why a fee; why not freeware like our basic DOE-2
versions? No published distribution licenses yet; and if these become available and have a
per copy royalty, there can be no freeware products that are EnergyPlus based unless LBNL
does them - in which case why would any private sector firm invest the time and $ to compete
against the government funds.  We could adapt both the PowerDOE and eQUEST (freeware)
interfaces to EnergyPlus in a straight forward manner fairly quickly; we cannot do this without
open source licensing and a working, supported (commercially viable) complete product - a
couple years away at best. I'm sure Eley can do the same with the VisualDOE interface and
EnergySoft with EnergyPro and so on;  DOE has no business funding work that so many
private firms are already doing and can do better; a similar argument can be made for
DOE's getting out of the engine business where the result is in "product" form rather than
NMF model library or "component toolkit" form.  The private sector can take the correct
next commercial product steps, better, cheaper and faster (and also provide the required
product marketing and support.)

Quote: "Sorry for the rant, but I'm also getting tired the people who think they can
purchase Power or Visual DOE and be a DOE2 expert in a week with no prior
experience -- simulation, HVAC design or otherwise."

Comment: we have many eQUEST users after quick, accurate, "first order" answers
doing just what you say is not possible.  But I agree that "specialists" will always be needed
to perform accurate, indepth analysis.  eQUEST 2.0 will be released within a couple weeks
(also freeware - and eQUEST 1.2 will be posted Friday); 2.0 will have a large fraction of
PowerDOE-like capabilities (2D/3D displays, HVAC diagrams, full DOE-2 details dialogs
and data editing, table/graphic report pages) and retains improved versions of the building
creation and EEM wizards.  We want to see thousands doing analysis not hundreds.

Quote: "PowerDoe hasn't been very useful for us.  Our buildings tend to be fairly
large with alot of detail, so it is easy to bring PowerDoe to a crawl on
both the P-pro 200 and PIII 300 machines we have."

Comment:  As I mentioned above, PowerDOE is like a very large spreadsheet with thousands
of formula; changing some items in a large building model causes tens of thousand of parameters
to be recalculated; PowerDOE 1 was developed as a 16-bit application - the memory sharing
between the 16-bit interface code and 32-bit interactive BDL suffers from very large windows
ALIAS-style memory sharing between processes overhead and the slowness of 32-bit numeric
calculation in a 16-bit interface environment.  eQUEST is fully a 32-bit implementation and thus
is much faster; eQUEST 2.0 has full BDL details editing.  Thus, we hope to greatly reduce your
frustrations in the near future.

Quote: "One more comment and question - the DOE is spending our money to create new code and have arguably abandoned the
multi-million dollar investment in previous DOE2.x developent. Is this the best use of taxpayer money?"

Comment: There is an important role, in my view, for taxpayer $ in this field; it is a good way
to get the basic models, algorithms, and lab/field detailed data collection and model validation
started (and sometimes the only source for this funding for some types of components.)  It is
a terrible way to get "products" done; DOE-2 became most successful because private
companies (and utilities) decided to invest heavily in taking the basic code to the next level
so as to be useful and useable by a wide range of "consumers" of the technology (designers,
researchers, code officials, etc.)  This required heavy investment in both the "raw" DOE-2
engine (mostly work done by my company) as well as many interfaces (GabelDodd
EnergySoft's EnergyPro, Eley Associates VisualDOE, Elite's EasyDOE, Item Systems
DOEPlus, EPRI/JJH's PowerDOE, JJH's eQUEST, DOE/PNL ComCheck-Plus, and many
others under development or private offering in specialized markets.)

Quote: "Perhaps my perceptions of success have been unduly tainted by the seemingly endless wait for PowerDOE and 2.2"

Comment: Lots could be said here, but not worth the space.  In summary, others decided to
announce the software development and schedule before the basic components were designed
and cost estimated or funds wre available to do the work.  My company advised against such
announcements and never made any prior to 1997 first versions being available.  No matter,
it is all now out there for you to use; more popwerful but probably new frustrations and
even longer "wish lists."

---
Jeff Hirsch
James J. Hirsch & Associates
Building Performance Analysis Software & Consulting
12185 Presilla Road
Camarillo, CA 93012-9243 USA
phone: (805) 553-9000
fax: (805) 532-2401
email: Jeff.Hirsch at DOE2.com
web: http://DOE2.com



======================================================
You received this e-mail because you are subscribed 
to the BLDG-SIM at GARD.COM mailing list.  To unsubscribe 
from this mailing list send a blank message to 
BLDG-SIM-UNSUBSCRIBE at GARD.COM
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.onebuilding.org/pipermail/bldg-sim-onebuilding.org/attachments/20000127/e4a7e846/attachment-0002.htm>