Reston, Virginia, March 26-28, 1996

COMPUTER POLICY ADVISORY COMMITTEE MEETING
RESTON, VIRGINIA
MARCH 26-28, 1996

Attendees:

Chair
Kathy Peter 

Committee Members:
Verne R. Schneider
Gloria Stiltner
Alan Lumb
Isabelle Halley des Fontaines
Peter Rogerson
David Shultz
Arlen Harbaugh 
Brian Mrazik
Jon Nowlin
Jon Scott
Sharon Goss
Jerad Bales
Bill Carswell
Kevin Oberg
Richard Hollway
Steve Brady
Graig McHendrie
Charles Tibbals

Facilitator
Betsy Daniel, Branch of Technical Development 
                and Quality Systems

Invited Speakers/Guests:
Ken Lanfear, Office of Water Information
Jo Ann Macy, Office of Water Information
John Briggs, NWIS Program
Susan Trapanese, DIS Program
Jim Fulton, DIS Program
Chris Barnes, DIS Program
Merritt Blalock, Northeastern Region, WRD
Ernie Brunson, National Mapping Division
Kathy Fitzgerald

Minutes 
Judy Griffin, Office of Technical Support

Tuesday, March 26
 
1.  INTRODUCTION/OPENING.  Kathy Peter opened the meeting at approximately
8:30 a.m.  Introduced Judy Griffin as recorder of minutes for this meeting,
and committee could decide at a later date how future minutes will be
provided.  There was a discussion as to whether a technical person should
record the minutes as opposed to a support staffer.  A discussion ensued as
to how technical the minutes should be in consideration that they will be 
presented to a wide audience of the Water Resources Division (WRD). 
Several committee members objected to a non-technical person taking the
minutes because of the technical nature and the complexities of this
particular meeting.  Jon Nowlin made a motion to have the minutes recorded
by Judy for this meeting.  The committee  agreed.  

2.  UPCOMING MEETING.  It was proposed to go to Orlando for the next
meeting because the subdistrict has many pc's integrated with their
DGs.  Charles Tibbals suggested either June or September.  The hotel
is within 2-3 blocks of the office.  It was decided to wait until the
end of the meeting to schedule the next meeting based on the
priorities derived from this meeting.

3.  STATUS/UPDATE.  The status of outstanding action items was updated
including strategic planning work groups.  How do we keep track of the
action items?  Do we need an accounting of the action items as to how they
are detailed and/or status?  Kathy Peter will  clarify the action items and
keep everyone aware by email as the items are completed.   Kathy will send
the list of action items to the committee later this week.  Completed
actions will appear in these minutes and will not be mentioned in further
minutes; status will appear in further minutes.  

Action Items from Previous Meetings:

Action: Steve Brady should be provided more details and he will look into
it.  DONE
 
Action: Jim Fulton needs to get details on G2 problems in Missouri and make
sure they are addressed by the new release.  DONE
 
Action: Terry Reinitz needs to check on the Frame template issues and make
sure they are covered by releases of new templates expected mid-summer. 
DONE
 
Q: what is being done about standardizing on AMLs, marker and symbol sets
for meeting publication standards?

A: A lot has been done locally but little has been captured and
standardized.  This is an issue that CPAC should address.

Action: Gloria Stiltner will have Doug Nebert follow up on this item.  DONE
 
Action: Kathy - Before the next town meeting, we should suggest that the
users get together in small groups and discuss their issues.  We could
provide some pump-priming questions if needed.  A CPAC member must be
assigned to coordinate with the users at the meeting site.  NEXT CHAIR
 
Action: Members should send comments on charter to Hal by end of the week. 
Hal will turn it into a fleshed-out document within a month. He will
circulate and give 2 weeks for comments back.  A completed draft will be
presented at the next meeting.  KATHY PETER TOOK THIS OVER.  DRAFT CHARTER
WILL BE AVAILABLE AT NOVEMBER MEETING.  DONE
 
Action: Tom Yorke will prepare a list of questions/issues that must be
addressed by DODEC in evaluating its merger into NWIS II.  ONGOING--ON HOLD
 
Action: Pete Rogerson will coordinate the DODEC evaluation effort.  The
evaluation should be prepared along the "Must Know" format presented above.
 ON HOLD
 
Action: Bill Carswell will bring up CDPs at the next Human Resources
Management Committee Meeting.  DONE
 
Action: Within 2-4 weeks, each Regional Computer Specialist must
contact each of their Districts to discuss and assess what additional
computer resources will be needed for NWIS-II.  ON HOLD
 
Action: Greg Allord, Gloria, and Verne continue their discussions with Jim
Biesecker on the need to better consolidate efforts of the DIS and
Publications groups and committees, perhaps even merging the existing two
committees into one.  DONE
 
Action: Isabelle should notify the email archiving subgroup that this is a
high-priority issue and request that group to supply us with a timeline and
work plan within a month.  DONE
 
Action: Gloria, Graig, Steve, Jerad will gather information for a
strategic plan synopsis by June 16.  DONE
 
Action: Jon Scott will take the lead on refining the "Must Know" list by
the next meeting.  ACCEPTED
 
Action: Members send their comments to Jon by June 1.  NO COMMENTS
 
Action: Susan Trapanese, Jeff Miller, Merritt Blalock, and Graig McHendrie
will continue to explore the idea of MS-Windows/DOS applications server for
the network that would serve the applications to an X device.  DONE
 
September 13, 1995 Update:
 
Western Region, Branch of Regional Research, Office of Hydrologic Research
has ordered an HP 500 to test the approach for serving MS Windows
applications to X devices.
 
See the Web page at <URL:http://wwwqvarsa.er.usgs.gov/~clbarnes/NT/ntpage.html
 
Action: a subgroup of HASS, NWIS, DIS, Discipline Offices and some National
Mapping Division (NMD) folks will evaluate GUIs.  Alan Lumb will chair. 
Will present their results in the "Must Know" form for the next meeting. 
MERGED INTO DIS "FUTURES" SESSION.
 
Action:   CPAC recommended that we send out Corporate Picture of software
issues instead of sending out a survey. The compilation should ask the WRD
users what functional needs are we missing in our assumptions.  Don't ask
specific questions about lab equipment.  It is DIS's position to support
the important packages. We will start with this list. It was suggested that
we support data exchange between packages (import/export).  DONE
   
Action: Verne Schneider and Tom Yorke-- Port of the Primes.  Full analysis
of state of NWIS-II report and full estimate and port plans of cost of the
port by November Meeting.  REPORT AT THIS MEETING.
 
Action Tom Yorke - CPAC recommended that more information from NWIS-II to
the users be disseminated.  DONE
 
Action: Verne Schneider and Greg Allord will look into status of
disseminating this information.  DONE
 
Action: CPAC recommended that the  Status of Strategic Planning for
Internet Dissemination, Exploration, and Retrieval (SPIDER) group be
disseminated.  Greg Allord will ask Ken Lanfear to do this.  DONE
 
Action: Ask Chris Barnes to add Greg Allord to NT server.  DONE

4.  REVIEW OF AGENDA.  Kathy Peter inquired whether CPAC wants to provide
information to Senior Staff or make specific recommendations on the World
Wide Web (WWW) to Senior Staff.  This will be discussed after Ken Lanfear
gives his presentation.  

Distributed Information System/National Water Information System (DIS/NWIS)
procurement plans and issues will be presented if enough time after the WWW
Issues.  NWIS will be presented virtually all day tomorrow, March 27. 
Ernie Brunson of the NMD Team will give briefing on the NMD review of NWIS.
 This team briefed Bob Hirsch, Chief Hydrologist, and Jack Fischer,
Associate Chief Hydrologist, on Monday, March 25.  No vote on
recommendations until Thursday, March 28.  It is a very difficult and
complicated issue.  Charter will also be discussed on March 28.  Dave
Shultz recommended that before leaving on March 27, some general idea of
proposed NWIS vote should be drafted   The group concurred.  In this way,
each member can talk to their constituents to get their reactions before
voting.  Kevin Oberg had been asked to provide an update to the WRD Data 
Chiefs before official votes are made.  This will have a
tremendous impact on the morale of the Water Resources Division (WRD).  Bob
Hirsch is concerned about how this information and decision gets out to the
rest of the Division and the U.S. Geological Survey (USGS).  Will there be
a formal release from this committee to the Division?  CPAC makes
recommendations to the Senior Staff; the Senior Staff makes the decision on
how this is presented to the WRD.  Thursday morning there will be a
discussion of NWIS information-dissemination recommendations to the WRD. 
Informal discussions can be made with each committee member's constituents.
 

5.  WWW ISSUES.  Ken Lanfear gave presentation using viewgraphs.  Three
major points.  Status of Sites,  Usage Statistics, and Guidelines.  

Handouts on WWW, including http://h2o.usgs.gov/usgs/www.html

Status of Sites:  How many have used the WWW?    Ken described WRD Home
Page.  This should be getting a lot of traffic today because of the Flood
at Lees Ferry or other gages on the Colorado River.  Ken has had a comment
from some people of the Division "I can't find anything." There has been
much attention to the public pages.  The Internal page of WRD Home Page was
revised during the furlough to provide Quick Links.  This includes things
you might need to know--WRD Calendar, AIS, NTC Training, etc.  Travel
Information--latest per diem rates; weather; WWW, help in creating pages. 
Ken is recommending that the local sites use a page like this or something
similar. The intranet is definitely picking up speed by use of people
within the Division.  

There are 22 States serving real-time data.  There are seven variations of
the software.  SPIDER has discussed using the different systems.  It was
decided that the Montana version would be used for now.  Most States are
now using Montana, version 1.  Eventually, we will be using the Illinois
software.  Montana 2 has much better graphics, but it does not contain
historical data.  The Illinois software integrates better with the historical data. 
Montana is funded by Office of Surface Water and Water Information. 
Question:  Is there any oversight group on this?  Yes, it is the SPIDER
group.  Briefing was given to Senior Staff on WWW issues late February. 
They are aware of the issues.  Memo being prepared (draft with Chief
Hydrologist) on this information for the field.  Data base is maintained on
Web server.  Jerad Bales is concerned about getting information into the
data base if Primes are changed.  

Historical Streamflow.  The system was completely redone by the Illinois
District.  It is available for all States; 37 are open to the public.  All
but Texas and Colorado use this system.  

This system is the easiest way to get streamflow data from USGS. 
Corrections and updates are a problem.  Solution may be letting District go
into their data base and report back to Headquarters.  Historical data on
the Web was compiled by Jeff Miller.  It was pulled from ADAPS onto a disk
in Reston, reformatted station by station.  Master control file tells what
stations each District has.  There are two basic elements of system.  In
pulling all data, it was shown to the Districts and asked what they thought
of the data and could it be released to the public.  There are problems in
some District data bases.  Discontinued records were found or gaps in
records.  Throughout the Districts, the cooperators like this data base. 
They do not need a password to get into the data base.  

Local sites--36 Districts have their own WWW Home Page sites.  Similarities
are USGS identification, professional content, USGS backbone navigation
path.  The differences are local needs and interest, and design skills. 
The goal is to have a training program for the Web similar to GIS Training.
 

Usage Statistics.  A couple of facts are H2O gets 400,000 hits from 28,000
users each month.  It is growing at about 16 percent a month.  About 95
percent of the users are outside the USGS.  (Many are cooperators but very
many are people that WRD has not had contact with previously.)  WWW is 1
percent of GEONET traffic.  H2O distributes 16,000 fact sheets per month. 
Almost every new fact sheet is online.  This a major public outreach
effort.  

Guidelines.  The SPIDER group was formed; this group listed things we could
put on WWW.  Some guidelines are in WRD Memorandum No. 95.04 which is on
internet.  Summary:  Simple, practical rules, local approval, room to
experiment.  Bureau guidelines are being developed.  It will give
consistency among the Divisions.  (See handout--Attachment 1.)

6.  PUBLICATIONS.  Jo Ann Macy gave presentation.  She mentioned that the
new Assistant Chief Hydrologist for Water Information (WI), Lewis Wade, is
now on board and very much wants to meet people in from the field.  Jo Ann
reported that there is no guidance as to what gets on WWW.  There are new
guidelines that were published about 6 months ago.  Abstract is
electronically sent to WI and then published on WWW.  Office of WI does not
have their own home page.  Greg Allord is going to take on this project. 
Hard copy reports (the traditional way) will be put on the WWW.  A
nonpaper, nondisk, report--this has not been decided as yet.  A clearly
defined process for reports being done electronically is not yet in place. 


Should all reports be on the WWW?  (See Attachment 2 which are the
viewgraphs of Jo Ann Macy's presentation on the subject of electronic
reports.)  The Division Library is trying to move forward.  We do not know
if the Library system is compatible with WRD's system.  Isabelle mentioned
that this is a bureau problem.  Gloria said that this is something that the
proposed Information Council addresses.  The structure of the system is
relational data base stated Ken Lanfear.  

The three divisions are meeting every week to talk about electronic
publishing.  A major meeting will take place week after next to come up
with a plan of action to be presented to the Information Council.  One of
the decisions Senior Staff made is that we will always have paper report
(hard copy) for the short-term not just an electronic version.  Greg Allord
will take on the project of electronic publishing.  

Are all of WRD customers ready for this?  Jo Ann believes that cooperators
are not ready for reports done electronically only.  We will lose a large
portion of our customers if we do not have a hard copy of a report.  Will
we save money by not printing hard copies?  What are the archiving
requirements?  Jo Ann believes this is a very important issue.  How long
does the report remain on WWW.  Does the library have a long-term support
role for archiving.  What reports do we want on the WWW?  If you go to 100
percent of electronic publication, this means the customer may have to pay
for a printout of a report.  We may lose the good will of a customer in
this way.  


7.  DISTRIBUTION OF SOFTWARE ON WWW.  Alan Lumb made presentation on
Distribution of Model Software.  (See Memo from Hydrologic Analysis Support
Section (HASS) Group Team (Steve Regan) on Approval of Hydrologic Software
for Distribution.)  (Attachment 3)  The software is ready to be linked to
Ken's page on WWW.  Alan drafted some configuration management proposals. 
The process is to get a Configuration Management Board and control change
on software.  This software will be distributed on tapes when approvals
come from Discipline Offices.  It can be pulled from WWW as well; however,
tape is more efficient.  Individuals encouraged to speak to site
administrator before getting copies of models on line.    

Alan distributed copies of the Applications Software home page to the
committee (see Attachment 4).  Interim fixes, bug fixes will go onto the
WWW page.  He is trying to evolve one-stop shopping for Division so
everybody knows where to go for software information.  Internal page
indicates what software is not there.  When Discipline Offices approve, it
will be put on public page.  The National Flood Frequency program is on
WWW.  It was a major struggle.  The important thing is that there is a
chapter for each State.  This is where we are currently.  There is a draft
configuration management plan.  

8.  WWW Issues Identified by CPAC:  

o  How far down the road do we go for online interpretation and analysis? 
Will we gain if we provide packaged data with applications allowing misuse?

o  Are we promising the public our data software--a "free lunch."  Can we
afford to do that?  What about our credibility?

o  What will cooperators think?  How will they react to free WWW data
software?  It is one thing to have data available, but by also having
application software available to general public, are we opening the door
to misrepresentation of USGS products?

o  What are our goals for WWW distribution?

o  What is the cost/benefit?

o  Should be require Districts to have a site?  If we required them to have
one, they would ask Headquarters for money.  

o  There is a difference between real-time data and historical data in this
discussion.  

o  Who does the updates of historical data?  You decide in the District
when you do the update.  If the update is sent to Reston,  the update will
be automatic.  

o  WATSTORE needs to be updated right now. (K. Oberg)

o  There is redundancy between data sets (triple redundancy).--WATSTORE,
WWW, ADAPS.

o  Discipline Offices are more concerned with NWIS getting up and running
rather than with getting more information on WWW.  

o  What is priority of WWW distribution vs. NWIS II?  

o  Who uses WATSTORE?

o  Can we eliminate historical streamflow data from WATSTORE if it is all
available on WWW?

o  H2O is the server of the future.  It has a direct connection to
internet.  When someone wants to put software on server, each one has a
separate partition.  (Ken Lanfear)

o  Central vs. distributed data base software WWW site maintenance.

o  When NWIS-II is working, the WWW is a very efficient mechanism on
answering fairly typical questions.  WWW cannot answer complex questions,
but for basic retrieval, it is very good.  WWW will be updated from
NWIS-II.  

o  How many groups are working on WWW mandated by Headquarters.  Gloria has
a record of nine groups.

Action Item:    Set up small group to formulate recommendations for CPAC on
WWW by June 1996.  The group consists of Kevin Oberg, Greg Allord, and
Richard Hollway. 

8.  SOFTWARE LIBRARY PROPOSAL.  Jon Scott discussed actions on this item so
far.  There was a problem in developing code; how do we get software in
hands of users?  Proposal was prepared by subgroup of Software Issues Team.
 It takes two people per year to maintain, discuss benefits, what to do to
code.  Some of those things are starting to happen.  A decision is needed
from CPAC on how we come up with implementation plan.  How much resources? 
How do we get the Division ball rolling?

Alan Lumb discussed the issues of implementation plan for a software
library.  Progress included in NWIS report that should be ready this
afternoon.  We have an evaluation copy.  HASS will have a gatekeeper for
all HASS software and 150 different programs in the system.  There is a
process, and it will continue.  There is a draft of Configuration
Management Board with Discipline Office representatives.  Resources are
allocated although limited.  How much does the Division want to allocate to
this?  We need to write up actual procedures; this is a high priority. 
What do we want to do with our models?--this was a task from the Chief
Hydrologist.  We will be meeting to start work on that.  

The current HASS staff became the librarians for the software, and Kate
Flynn is the gatekeeper.  What would the steps be that an author would go
through?  It is reviewed by HASS; there is a form to fill out that goes on
home page.  There is a test data set.  Widespread software dissemination
would have approval before being distributed to Division.  
Handout is available on: http://hassrvares.er.usgs.gov/ 
 Action Item: 
How many FTEs would it take to make a software library operational and what
are the timelines?  Alan to come up with a draft of his recommendations to
CPAC on this by the June meeting.  (Alan Lumb)  

9. CAREER DEVELOPMENT PROFILE (CDP)--Report on committee recommendations to
Senior Staff.   David Shultz reported on evaluation of current CDP system,
core needs, and outline of basic concept of CDP system replacement.  There
is an 8-page report.  The evaluation covered who was using the current
system.  The only user is the Human Resources Management Support group.  It
is important that a replacement system be easy to use with clear
instructions.  The recommendation was that people be able to take resumes
and separate data fields; send them by GEOMAIL to a server at Headquarters,
read data between delimiters, and send back documentation showing that it
had been accepted.  The data could be output in several different ways and
to meet needs of Human Resources Management Committee (HRMC).  Using
information most people have available, it would encourage use of the CDP
system.  Feedback from Cathy Hill, Assistant Chief Hydrologist for
Operations (ACH/O), is that there will be a discussion of this issue at the
April meeting.  Cathy Hill said that there would be a computer specialist
as part of the HRMC.  Sharon Goss suggested using the system that is in
place, employment history is available through FPPS.  Every user would be
responsible for updating CDP information.  Some Districts are doing this
for the employees.   Recommendation:  CPAC will review proposed programming
recommendation for Senior Staff.  Efforts should be kept to a minimum.  
 
10.  DIS--GENERAL PROCUREMENT ISSUES, STRATEGIES, AND PLANS.  Gloria
Stiltner distributed a package to the committee. (Attachment 5)  Systems
costs--DG current investment is $40 million.  At the peak of the primes, we
paid $5 million in maintenance.  We are paying $600,000 for the primes that
we have now.  Telecommunications--160 sites on network.  We cannot afford
much more than 56 Kb on most lines.  DIS Office is concerned about the
health of the primes.  Quite a disruptive failure in Oklahoma.  Gloria
stated that Computervision was asked to assure us of the spare parts that they
have.  Prime failures that have been reported are on the graph "Prime
System Problems that Caused Downtime."  Other categories are
air-conditioner problems, board problems.  Right now there are no plans to
put out an RFP.  The recommendation would be to do time and materials. 
Because operating system is frozen, there is a problem when we reach the
year 2000.  

There will be no higher CPU boards available for DGs.  There are plans
to freeze the operating system on the DGs; it will not cause problems
especially with ARC.  There are CBD announcements of the maintenance
contract to be awarded.  It cannot be awarded this September.  There
cannot be an alternate maintenance without a 30-day notice to DG.  DIS
is concerned about the lack of software support from DG.  INGRES and
FRAMEMAKER (Adobe) have been asked about cost of software support.
INGRES has agreed to do software maintenance at a cost which is very
reasonable.  Adobe is probably going to make us an offer to move us to
other products.  One of DIS proposals is to have an analysis of
software maintenance costs.  Until there is some knowledge of where
NWIS-II is going, DIS is on hold on some of these decisions.  We
basically have a UNIX system that is frozen.  Vendors are refusing to
port to DG UNIX.

11.  Z-MAIL.  Jim Fulton gave presentation on Z-Mail.  (See Attachment
6.)  Procurement was awarded in September 1995 for Z-Mail.  This will
be for bureau-wide usage--it can be used for PCs and MACs as well as
UNIX.  It requires much more memory for MACs than the Eudora Program
that MAC users currently have.  It also includes a calendaring and
scheduling program.  The contract is with a vendor called Spectrum.
The handout presents a comparison between Exmh and Z-Mail.  Z-Mail is
much easier than exmh to read and send attachments.  Z-Mail has its
own filtering system.  It seems to crash a lot on the UNIX platform.
If you mistype an address, it will save an address to allow you to
re-edit it.

12.  WINSIG.  Chris Barnes gave presentation on the work of the Windows
Integration Group (WINSIG) which met in February 1996..  Chris distributed
a handout on the WINSIG Meeting (Attachment 7).  His presentation gave the
status of the window-server testing and the suite of software that has been
recommended for testing.  There is an interoperability problem and WINSIG
wants to solve it using current hardware and software.  Charter statement
is being defined as to what WINSIG is doing and the timeframe for
completion.  This will be presented to CPAC at least by the end of the
fiscal year.  Two solutions that would work well in WRD is NCD's WinCenter
Pro and Hewlett Packard HP500.  These are the two solutions that the group
plans to test.  Software applications programs (see attachment) were listed
that will be tested.  (There is one MAC server that Doreen Lucas has been
testing.)  Software licensing requires one copy of license for CPU or one
user per license.  This will be investigated and reported upon.  Security
implications will be investigated for the operating systems.  Chris also
distributed a second handout (Attachment 8) on WINCenter Pro for WRD and
also the brochure on WinCenter.  Chris distributed another handout (see
Attachment 9).  Chris then gave a demonstration of the WinCenter Pro. 
Gloria Stiltner indicated that a proposal to consider is that we wait until
the NWIS-II solution.  If a new expenditure is made, it should be a PC
Pentium, discourage any improvements in UNIX DG for offices having
x-terminals, and purchase the windows server to service mainstream micro
software.   


Wednesday, March 27

13.  PURPOSE AND DESIRED OUTCOME.  Bob Hirsch would like CPAC to have a
memorandum to all employees of WRD on the outcome and recommendations from
this meeting.  Verne Schneider stated that in a conversation with Jack
Fischer last night, Jack indicated that Bob Hirsch will sign a note to WRD
on the status of NWIS-II and what we are doing.


14.  NWIS.  Verne Schneider gave an update on status of Division
discussions and views of the review of NWIS-II of the National Mapping
Division (NMD) and discussions with the Chief Hydrologist.  He gave an
overview of what will be presented.  Verne Schneider believes that we
should learn from the mistakes of the past in the NWIS-II Program and that
we should move forward in a positive direction.  What does it really cost
to make NWIS-II work?  CPAC will be instrumental in making decisions
regarding the future of NWIS-II.  The work groups in place are very
accurate in identifying  the key issues.  

Alan Lumb presented his views and said he would be presenting background
information on what his office has been doing.  We are working on a
solution for the Division to get the NWIS-II working.  There are a great
many frustrations, but he praised his staff in being able to work through
these problems.  Verne indicated that a well-formulated plan by CPAC will
influence the Senior Staff's opinion of what will happen with NWIS-II.  

Alan suggested going around the table to get the opinions of the committee
so he can know what your issues are and what needs to be resolved.  We will
go back through this list and see what has been resolved.  This way they
can report back to their constituents and also see what the gaps are and
what has not been resolved.  

Committee Issues:

oTiming is critical--Charles Tibbals

oRecommendations vs. decisions.  Vision that CPAC is going to make the
final decision of where NWIS-II is going.--Brian Mrazik

oWhat is the status of NWIS-II--Bill Carswell

oThe Districts need to know what we are doing so they can make resource
decisions.--Jon Nowlin

oWhat functions do we lose if we port vs. going to NWIS-II?  What about
contracting vs. in-house.  Will WOTSC increase?  How does NWIS-II changes
fit with other hardware/software.  How much have we already spent?--Jerad
Bales

oConcerned that we won't move forward with decisions.  We need to make sure
that our investments and hardware will have some value with our existing
equipment --Steve Brady

oProposal to stay on prime for indefinite period of time--Merritt Blalock

oWhat were critical mistakes of past so that they won't be repeated--Dave Shultz

oWhere do we go with NWIS?--Arlen Harbaugh

oTrue performance issues with the current NWIS and performance solutions. 
Are there fundamental design flaws in the data-base structure.  Can a
combination of porting and finishing portions of NWIS-II work?  Does WRD
have the expertise and technical skill to finish the NWIS project?--Graig
McHendrie

oThis is a mission-critical system.  The hardware failed miserably.  We
have to put our mission-critical data base on hardware that can be
repaired.  If not, it will only get worse.  --Jon Scott

oI am concerned with the impact of NWIS-II on WRD staff.  I would like to
have a better understanding of why it took so long to get to the place we
are today.--Richard Hollway

oWe, as CPAC, need to define the scope and complexity of what this system
needs to do.  Scope and complexity are separate.  We, as the management
oversight group, need to define past failures.  If we do not define, we
most likely will do them again.  We need to implement corrective action. 
We need to be able to define the interface between NWQL and the NWIS-II
system that has not come up as a issue.  Hardware is a derivative.  It is
not an input.  We must define the system and its requirement and the
hardware--Peter Rogerson

oWhen and what critical information do Districts need to know to be able to
allocate resources?--Sharon Goss

oScope and complexity.  What makes us think that we will succeed now after
the fourth attempt?  I don't see any indication that this time will be
different.  Reengineering--I would like to see that defined.--Isabelle
Halley des Fontaines.

oMy concern is the constraints to deal with and the resources.  I would
like an assurance that our ability to provide data to customers is going to
continue uninterrupted.  I would like an assurance from the Division that
we will have the resources.--Alan Lumb

oDoes CPAC have the authority to make this decision.  I want to hear the
bottom line from Alan Lumb.  Hirsch's memo must have decision--define
process later.--Gloria Stiltner

oLogical justification for decision.--Kevin Oberg

oSet reasonable expectations for the system.--Bill Carswell

15.  VISION STATEMENT.  Alan presented the vision statement of the NWIS
staff.   "The NWIS staff vision statement.  The NWIS staff, management and
customers through teamwork, open communication, commitment, ownership, and
accountability will provide a user-oriented, user-supported hydrologic data
system.  The NWIS vision is to provide a data repository for hydrologic and
environmental data and to make available well documented intuitive,
cost-effective software that utilizes appropriate technology to meet the
data-processing needs of USGS and its customers."  

16.  NWIS STATUS.  About 6 months ago after accepting this job, Alan spent
2 weeks visiting staff in Arizona as well as at Headquarters.   He came up
with some key items of what needs to be done or addressed.  Six key items 
to be addressed are:

a.  reorganization
b.  critical requirements
c.  milestones
d.  configuration management
e.  data model review
f.  performance testing  

The reorganization has been established, and Alan explained his decisions
and choices.  

Most of the District support is coming from Arizona and Arkansas.  There
are some in North Dakota and Virginia working with NWIS staff.  This
represents about 25 personnel.  (Alan distributed a package of material to
the committee earlier.  Refer to this package.)

Requirements:
An 80-page table of requirements was rewritten by the consolidated user
group, (defined by disciplines); a check list was put together for the
February 26 test.  This had never been written down or clear before this
document was created.  Pete Rogerson said the NWQL has set up system
specifications.  He does not want the Lab left out.  It must be an integral
part of this system.  Alan said he was equally concerned with that.  Both
NWQL and NWIS has assigned contacts to be working together on the NWIS. 
Alan will present the user group to CPAC and ask the question what is
missing from this group.  

Milestones:
How to get to the alpha test.  There will be new people looking at the
system.  From the alpha test, we go to beta tests.  Then from the beta to
full implementation.  Susan Trapanese and Merritt Blalock were working on
these issues.  There are more reality checks to meet deadlines
appropriately.  The timelines previously set were ones that the staff knew
could not be met.  There were 52 installations which included 4 beta tests.
 Each District needs 8 weeks of data transfer and 2 weeks of on-site
training.  Six Districts can be done concurrently.  Six 8-week periods
equals 1 year.  Isabelle asserted that there is nothing wrong to set target
dates.  What is the minimum that we can accomplish by that target date? 
Realistic target dates can be set and constantly monitored so that we can
meet that target date and make mid-course adjustments.  The staff has never
had that kind of information.  Target dates were  not monitored.  

Configuration Management:
How big is NWIS?  There is a GUI tool--generated code.  There are 24,906
pages of developed code.  The NWIS code C, pearl, scripts--9 or 10
different things in the system.  There is a substantial amount of transfer
code written.  Reference lists include 13,026 pages.  System
documentation--there is very little out there.  User documentation is 75
percent complete.  We need a configuration management system.  There are
high volumes to configure; it is costly to create our own; very costly to
do nothing.  NWIS has reviewed packages and is evaluating the short list,
and a selection will be made in April 1996.  We have spent over $100,00
trying to do that ourselves.  Alan has combined efforts of HASS and NWIS to
do this.  

Data Model Review:
Changes were made with no formal process.  Now we do a data-model review
which documents NWIS-II data model, document business rules, enterprise
modeling has started.  We have purchased some tools on Windows NT server so
we have multiusers using this server, and we need a process to analyze the
data model.  

Performance Testing:
This had not been done.  This is now planned following the February 26
test, and it is one of our highest priorities.  Colleen Babcock is now
working on this.  There are hardware and software solutions which will be
described later.  

There has never been any review of anyone's code.  This will no longer
continue.  

Status of Critical Requirements:
There are 838 requirements.  This led up to the Arizona test.  The system
was not that far along.  We gave documents to NMD that showed 99 percent
complete.  The staff actually thought that we were that far along.  We had
to convince the NWIS staff and the Division that this was not the case and
that we needed to identify the critical requirements.  NMD was very helpful
with this so we could determine what the status of the system was.  The
result from the Arizona test--the bottom line is that 26 percent of the
requirements were accepted, 22 percent provisionally accepted; 12 percent
not accepted; 26 percent are not available, and 15 percent have not been
tested.  Alan showed two viewgraphs of what requirements have been met or
the status in ground water, surface- water quality, biological, and
water-use components.  Alan also showed another viewgraph of the sample of
the NWIS top 70 problems--navigation paths not clear to user; simplify
interface; system too slow; data base too complex; can't give users what
they want without being complex; output; optimize portions of data
base--identify 90 percent most used; more teamwork needed; need change of
attitude (no patching); realistic timeline, and resources.  



17.  NMD Review Team Briefing.  Ernie Brunson was introduced and gave a
briefing of the review done by NMD team.  The review team was from the
Systems Development and Integration Branch.  Jim Jancaitis is the Chief of
that Branch.  NMD became involved in this review by the request of the
Chief Hydrologist for an independent outside review.  The first meeting was
in August of 1995.  The charter from Bob was to work with NWIS and come up
with an independent evaluation.  What is the actual status and where do we
go from here?  Tim Miller and Verne Schneider were the interface members to
the review team.  Ernie stated that he has worked with Alan Lumb and
visited the Arizona and Arkansas NWIS teams.  On Monday of this week, Bob
Hirsch and Jack Fischer were briefed on the results of NMD's review.  

Evaluation Criteria:
This includes observations, analysis, conclusions, recommendations, and
salvage recommendations.  (See Attachment 10)  Does the system really do
what it was designed to do?  Is there good documentation?  How do you
really configure management processes to configure the code to ensure
everyone was working with the same version of the code?  If the system is
not friendly to the user, then it is not going to be accepted.  User
documentation is important.  

Summary comment:
NWIS is a nonfunctional system, a first prototype, and has grossly
unacceptable performance.  Many functions that worked in previous tests did
not work in this test.  What they believed to happen is that when one
function was fixed, another then became inoperable.  Many of the users and
testers said they did not have time to validate that they were getting
correct answers out of the system.  There should be known test sets with
known results.  NWIS staff was unfocused, and there was no clear lines of
authority.  In the interviews, questions were "who do you report to?", "who
do you get assignments from?"--there were no clear answers.  We found
inconsistent use of the tools that were available.  Personnel were
untrained in what they had to work with.  Users who were conducting the
test were not trained in how to conduct testing--not academic training or
subject matter.  Substandard software tools were noted.  NWIS team decided
to write their own code as opposed to buying commercial programs available.
 There was a poor data-base design.  The testers input data but could not
retrieve it.  There was no evidence of standards used for good coding. 
Asked programmers to bring up code for their most difficult module.  There
was no consistent header information--who the programmer is, what the
program does.  In many cases, testers could not follow the screen,
terminology was not that currently used in the Division.  There was no
system perspective.  Every designer was interested in whether it was their
own discipline.  There was no an overt effort to integrate all the
disciplines together.  Many people were concerned that they had the support
of their District Chief.  The entire system is designed for graphic
interface.  Given the observations, they concluded that the progress was
not very great, and major changes need to be made in order to be completed.
 They believe the system is only 10-20 percent completed, and it will be
very difficult to fix.   Many problems have not been discovered.  Most
prototypes are throw-aways.  Prototypes are done to know how to do a good
system.  The good news is that you have a team that has learned much; if
you reengineer the project, the team has a knowledge base that will know
the pitfalls of a large system development.  There may be some code that is
reusable.  

Primary Recommendation:
Stop working on the existing system and start reengineering and rebuild on
what you have done in the past.  NMD thinks it very important that senior
managers of WRD make it known that this is a very important program and
that it will be funded.  For a smooth and controlled environment, you need
a buffer between users and programmers.  Buy off-the-shelf products for
configuration management.  Dedicate some level of staff to develop the
system.  The current turnover of people is high.  Appoint a Technical
Director for oversight.  

Questions and Answers:
Could some work be contracted out?  Yes.

Do we have the expertise we need to write the statement of work that would
be needed to contract out the entire development?  Don't know the answer to
that.  We did not clearly define the requirements.  Verne stated that we
could even have a contractor assist in writing the statement of work.  

Could you comment on whether it is better to contract out fully or have a
blend of in-house as well as contract workers.  You need a clearly defined
list of requirements.  This would fail if you do not have good milestones.

Is a distributed system  best or a more centralized system?  It does not
matter with configuration management.  

Can you comment on salvage recommendations?  Out of what we have seen and
people we have talked with, the feeling is that the time-series portion of
NWIS works better than any of the others.  If you do not want to go the
"throw-away" route, you would have to do output, discrete, and activity
subsystems.  Some could be saved.  You would be reengineering about
three-fourths of the system.  

The NWIS staff has not been briefed by NMD as yet.  Many of the staff came
to the same conclusion as the NMD has concluded.  This will not be a shock;
they have not received an official report.  It was not appropriate for Alan
Lumb to distribute such a document before this was presented to the Senior
Staff.  This is no secret; it was just presented last Friday, March 22, to
the Chief Hydrologist.  NMD has not written a formal report.  Bill Carswell
stated that he believes this same briefing should be given to the NWIS
team.  Action Item:  NMD will brief NWIS staff on the results of their
review of NWIS-II.  (Alan Lumb)

In the past, the tests said one thing and the reports stated another
results.  Ernie Brunson said this is what he  heard from personnel at the
test.

18.  NWIS OPTIONS.  Alan invited Susan Trapanese, John Briggs, and Kathy
Fitzgerald to join the group for this presentation on options.  This was
also presented to Senior Staff.  From a Division standpoint, we need to
rethink our needs for NWIS-II.  We are talking about Configuration
Management Boards, etc.  The consolidated user group includes Nancy Barber,
Bill Bartless, Verne Sauer, Bill Kirby, Charlie Demas, Rene Garcia, Kathy
Fitzgerald, Linda Gieger, Keith McFadden, Ed Putsay, Terry Schertz, Al
Onions, Lorri Peltz-Lewis, Jim Slack, Greg Pope, Howard McCormack, and
representatives need to be selected from National Water Quality Assessment
(NAWQA) Program, NWQL, and DODEC.  Who will make management decisions
regarding resources (purchasing requirements).  

NWIS-I Options:
NWIS-I does not meet critical requirements.  If we cannot do that in the
near future, how are we going to continue?  How do we get by in the
meantime?  On NWIS-I options, do we maintain prime, port to DG, or port to
new platform?  

NWIS-II Options
NWIS-II options--redesign new platform, redesign DG data server, apps on
new platform, reengineer DG, redesign new platform, redesign DG data
server, apps on new platform, redesign new platform.  

Verne stated that if we can reclaim or salvage some part of NWIS-II, will
this be cost-effective.  We have to analyze the time involved, and what the
numbers are.  Alan showed a viewgraph (see Attachment 11).  This flowchart
outlines a series of tasks to determine improved performance on DG/INGRES. 
These tasks are a major data model review; add task to select platform and
data base, redesign all subsystems for new systems, limited additional data
model review, continue development on DG/INGRES, reengineer selected
subsystems DG/INGRES.  This is what is planned to do over the next 6
months.  The bottom line is that it is unacceptable now.  One
option--forget DG/INGRES.  Second option--give DG/INGRES a chance.  Alan
does not agree with NMD that only 10 percent of NWIS-II is completed.  Some
of the reengineering tasks are to refine critical requirements, review
reference lists, testing (unit to system acceptance), review designs and
code, tools, documentation and training.  Alan also defined the
reengineering tasks (subsystems).  We have 25 FTEs down from 45 FTEs in
1991.  There are a certain number of resources and a schedule to meet--
something has to be cut as far as requirements are concerned if there
cannot be some adjustment to either resources or the schedule.  

19.  PORT ISSUES.  John Briggs gave presentation which included work
underway by Gary Rogers.  (See Attachment 12).   Gary Rogers has done
limited testing, no performance testing, or evaluation.  Nothing here that
would preclude from doing a port.  There are additional questions such as
security, backup and recovery.  These issues need to be discussed and
decided up front.  Converting FORTRAN and CPLs is not a major time of the
entire port process.  A good bit of time would be in writing utility and
maintenance programs.  Updating and writing documentation needs to be done
which takes a good bit of time.  A certain amount of time would entail a
couple of training classes in Denver to train system administrators.  How
soon can this be done?  This depends on the number of people and resources.
 Given the right people and right testers, this could certainly be done
within a year, perhaps less.  Most of the NAWQA data is on INGRES

20.  PRIMES.  Susan Trapanese gave briefing on the prime.  The primes
should no longer be operating by the year 2000.  (See Attachment 13)  If we
want to be off the prime by the year 2000, we have to meet the dates
indicated on attachment 13.  Prime is costing $2.4 million for 1996 and
$2.4 million for 1997.  Cost goes up to $2.7 million for 1998 and also for
1999.  If the prime goes down, you lose your system.  If the DG goes down,
you don't lose the entire system.  


21.  Strategies to meet critical requirements

oIs performance the right issue on which to make a decision?

oDo we need a short-term solution; is there any way to use the $40M
investment we have in hardware and software.  

oThe bandaid solution is off the table for NWIS-II.  

oWhat is a good data-model review process?

oCan we make a determination on level of performance with NWIS-II in its
present form?

oWhere do critical requirements fit in?

oHardware and/or software?

oOne of the recommendations from NMD is to make NWIS-II platform
independent.  Do we need to become hardware independent?

oIn-house vs. contractor?

oIs there a data-base management system on the market better than INGRES?  

oShould we segment the system and do an analysis of what is critical in
order to simplify?

Decisions to be made:
Long term
Interim
Roles (Configuration Management Board)
Port
Reengineer/redesign/salvage
In-house vs. contract
Platform 
Communication to WRD
Constraints/criteria for decisions
How do we ensure that we don't end up here again
Cost and Benefits of what we decide
Keep the Prime?

Constraints--
$3-5 million per year; 
$90 K this year; 
50 percent of 25 FTEs this year
Lack of skills
DG
Prime contract cannot be extended beyond 1999
Difficulty in reallocating money
Restrictions on Hiring


Agenda for Decisions

(1)  Vote on the Prime

Prime Issues--
Consolidating primes
Software emulating primes

Move off the prime ASAP
  contract
  costs
  failures
  ability to provide continuous service to customers
  High probability of catastrophe
  Mission critical software on undependable hardware  
  Make a formal recommendation to the Senior Staff to move off 
  prime by 9/97

Reasons to stay on the prime

  PayPers still on prime
  (not an issue if PayPers is successfully moved off by scheduled
  date of 9/96)

(2)  Porting NWIS-I

Porting
All NWIS-I ported
Partial port of major modules
(no change in functionality)
(fix bugs)

Recommendation agreed upon:  Port all current functions of NWIS-I to DG
with no changes in functionality.
(Implementation completed by 9/97)

(3)  Process for developing NWIS-II

(4)  Housekeeping (Communication, Next Meeting, Charter,
     Work Group Meetings, Briefing on Enterprise Modeling)

Motion:  Extend membership of those who would move off the committee this
year to December 31, 1996.  New chair will be elected by CPAC in January
1997.  The committee agreed.  
Action Item:  Kathy Peter will revise changes to rotation schedule and send
draft to committee members for approval.

Thursday, March 28 

Voting took place on the four decision items from Wednesday, March 27, see
above for action items and decisions made today (March 28).

22.  PROCESS FOR DEVELOPING NEW NWIS.  Discussion took place.  Decisions
need to be made regarding the following items:

oDevelop new NWIS by starting from scratch or salvage, platform or platforms.

oAnalyze what went wrong the last time.

oCorrect review of the critical requirements.

oTechnical Director, Configuration Management Board, User Groups need to be
considered carefully.  

oResources, critical skills required, in-house or contract.

oSystem design.

oWhat is the importance of the data base to the Division.

oCredibility.

23.  Charge to NWIS Staff:  Direction to NWIS Staff is to prepare a plan
for redesign of NWIS-II and report on resources necessary.  Use the
recommendations from the NMD review.  Rebuild on existing capabilities.  

  oporting NWIS-I
  ocritical requirements
  oconfiguration management
  oConfiguration Management Board
  oresources (skills and staff)
  odata model review

Action Item:  Submit outline of proposed plan to committee by e-mail by
April 19, 1996.  (Alan Lumb)

Action Item:  NWIS Staff Plan for revised NWIS-II, as outlined above, to be
sent to committee by June 1, 1996.  Report back to CPAC at the next
meeting.  (Alan Lumb)

Action Item:  Prepare draft memorandum (to all employees from Chief
Hydrologist) describing process to redesign NWIS-II.   (Kathy Peter, Kevin
Oberg, Jon Nowlin, and Isabelle Halley des Fontaines--any comments due to
Kathy by 11 a.m. tomorrow, March 29.)  

Action Item:  Subgroup to craft justification for Prime recommendation. 
(Alan Lumb)

Action Item:  NWIS staff to complete software maturity evaluation by June
1996.  (Alan Lumb)

Action Item:  For field, prepare description of implications of porting
decision (hardware) by NCTM Meeting (May 20, 1996).  (Gloria Stiltner)

24.  Followup Issues that have not yet been addressed or completed during
this meeting:  (See Issues from Monday.)

oDIS-III procurement and planning.

oWhat were critical mistakes--need to know so we don't repeat them again.  

oProvide Chief Hydrologist with a summary of CPAC decisions after each
meeting.  (Chairperson of CPAC)

oAgenda items for next meeting--lessons learned and actions taken.

oProgress is communicated in terms of meeting milestones.  

oCharter/Membership change issues to be discussed at September meeting.  

25.  FUTURE MEETINGS.  The next meeting will be held on June 12-14, 1996,
in Altamonte Springs, Florida.  There will be a town meeting with CPAC on
June 12 from 8 a.m. to noon with CPAC continuing in the afternoon 1:00 to 5
p.m.; June 13  from 8 a.m. to 5 p.m., and on June 14 from 8 a.m. to 4 p.m. 
Allow 1.5 hours to get to an airport on Friday afternoon. 

The September meeting will be held in Lawrence, Kansas, on September 4-6,
1996.  The town meeting will be held September 4, 1996, at 9 a.m. followed
by the CPAC meeting until 5 p.m.  September 5, CPAC will meet from 8 a.m.
to  5 p.m., and on September 6, CPAC will meet from  8 a.m. until noon.  

26.  ADJOURNMENT.  The meeting adjourned at 11:15 a.m.