Computer Policy Advisory Committee (CPAC)
Minutes August 19-20, 1997

Executive Summary


Table of Contents


CPAC Minutes August 19-20, 1997 Attendees:

Brian Mrazik
Alan Lumb
David Schultz
Jon Scott
Janice Fulford
Ken Mullen
Colleen Babcock (8/19 only)
Bill Carswell
Verne Schneider
Tom Wood
Jerad Bales
Richard Hollway
Isabelle Halley
Graig McHendrie
Scott McEwen
Kevin Oberg (8/20 only)

Absent:

Sharon Goss
Greg Allord
Steve Brady
Linda Pratt
Bob Davis

Tuesday, August 19

Opening Remarks

Brian gave an overview of district operations followed by a description of computing resources by David Parker.

Discussion of NexWIS

Alan gave a briefing on what has been done to date. A contractor has been hired to develop a plan. The document is being developed as a partnership of the contractor and NWIS staff. Two review meeting have taken place, one with Verne Schneider and Bob Hirsch and one with Merritt Blalock and Bill Carswell. Funding expenditures for FY97 were requested. These expenditures would be highly dependent on the fractions of time allocated by 7 NWIS staff members and this information is not documented but estimates could be made.

AI: Verne. In order to compute return on investment (ROI), construct FY97 expenditures on NexWIS by November 3, 1997.
Response to AI by Alan Lumb. ROI requires estimates of benefits and until this planning effort completes its logical steps that information cannot be provided. The FY97 costs are referred to, in engineering economic terms, as sunk costs. They are no longer useful for decisions for the future. They impact future decision only in that they reduce the additional planning costs that will be required. Senior management has asked us to do this planning, so that decision has been made. The FY97 costs may be useful for a retrospective analysis at some future date.

Ken Mullins described the step-by-step analytical approach using some of the figures in the draft planning report.

A requirements analysis tool was purchased ($1200) to begin to gather requirements for documentation, analysis, assignment of attributes for priorities, test status, and time frames.

Several questions were raised which will be taken into account as the planning process continues. Jon pointed out the biological and ecological data and what is stored and the relationship of the elements is vastly different today than what is stated in the content of the `91 SRS document and therefore, these new requirements must be added to the target. Dave asked about the details behind the process to control requirements' priorities and how conflicts are resolved. Isabelle asked about who is scoping the requirements and who is assigning the weights. It was noted that the Configuration Management program needs to be expanded and that there wan't a clear distinction between the conversion, November enhancements, '98 expansions, and the new data base.

Can the fundamental data base design be changed gradually? No. There is a parallel design team examining the NEW data base system.

Examination of other application packages from other agencies, countries, etc. will continue. For example, STORET-X from EPA has a prototype available now and promises a production version next year.

Jon was concerned about the concept of the NWQ lab treatment as a separate input and output. The lab must be treated as a distinct system until the lab and the QW community can develop a common set of requirements.

AI: Jon and the Phoenix Project will define the lab data issue and Jon will report to CPAC via e-mail prior to the next meeting.

AI: Janice, Verne, and Bill to write policy on who decides in the last bullet of section 10.1 and who decides what are `limited used data types' in the 15th bullet of the same section.

These bullets are:
  • [7.3] NWIS reserves the right to require the storage and documentation o f certain data types used in WRD that are known to be of limited value to anyone other than participants in a given study. Unless specified otherwise, there is no current requirement for these limited use data types to be stored in NWIS.
  • [9.0] Decisions to enter data into NWIS from other agencies shall be based o n direct knowledge by the USGS of the methods and quality assurance used and inc lude documentation of the knowledge as part of metadata.
AI: All. Write comments on sections 6, 7, 8, and 9 of the NexWIS Strategic Plan to CPAC via e-mail by September 5, 1997.

LUNCH

NWIS Configuration Management Plan

Ken gave an overview of some concepts for configuration management. There was some discussion on the chair for the board and a discussion on the importance of software design and documentation of the design. Several members felt there was not enough emphasis on design, particularly given past events.

Review of AI's

Graig reviewed both complete AI's and those still outstanding.

AI: Greg Allord and Verne. Develop a Water Information policy statement by September 2, 1997 that states web-page development should be at HTML 3.2 after January 1, 1998.

AI: Janice. Develop technical guidelines to quality assure provisional data on the web by November 1, 1997.

Alan, Ruth Thornberg, Kate Flynn, and BJ Bour are working on Software Library package development. Working through RIB(?) on Sun.

AI: Alan to announce HASS web-page availability via e-mail by September 2, 1997.

AI: Graig will contact Bob Wakelee about LAN/WAN monitoring and analysis training and report the status on the November 1997 meeting.

CPAC agreed to table the Total Cost of Ownership until the November meeting. Richard passed out copies of Steve's data for comment.

Tom and Verne gave an update on the status of the Annual Data Report Production group. A subset of the group will visit the Maryland District next week. Verne to hand out Reinitz e-mail on status.Cookbook for Preparing State Annual Water-Data Report

Equipment Inventory and Futures

Graig distributed copies of tables of inventory information collected for the WR. Similar tables are being developed for the other regions. This information can be used by FutCom committees to follow and observe trends, by DIS for procurements and training issues, by regions for District Program Reviews, by Sr. Staff to note assets, and by districts to locate other sites with similar equipment.

GIS Directions

Verne provided a briefing on ESRI's future directions. From the ESRI conference, ESRI announced that the Windows NT will be the new development platform, DG/UX Arc/Info will be static at revision 7.1.2 (which has yet to be delivered), and Unix functionality may be different in the future even though other non-DG/UX Unix will continue to be supported by ESRI.

AI: Tom will review and distributed Dangermond's letter to CPAC.

AI: Tom will review new features of ARC 8 and determine their applicability to WRD production.

FutCom3

CPAC agreed to charge a new subgroup, FutCom3, to update the year-old futures report.

New elements include:

Verne agree to pay for travel and per diem for non-technical support participants. The target meeting time is November 1997.

The new charge:

Update the FutCom2 report to describe the computing environment for WRD over the next three years [1998-2000]. The update should specifically address at least these items:

AI: Verne will make the WRD strategic plan available to committee members prior to FutCom3 meeting.
AI: Graig will firm FutCom3 membership by September 19, 1997 and e-mail to CPAC.

JAVA/Sedcalc

Alan gave a presentation on ConCalc which had been re-written entirely in JAVA. Although the Sed-Calc conversion to java was a useful exercise and will produce a useful product, the language itself, and particularly the currently available object libraries, are not yet ready for major applications in our environment. Interest was shown in the program so Alan agreed to have Tom McKallip complete some enhancements so it could be sent to the field as a test version for review and comment. Jerad Bales requested a copy.

The committee was concerned about how ConCalc differs from SedCalc and that the user requirements that went into ConCalc be documented.

AI: Alan will clarify the ConCalc requirements issue for the committee and will make sure that OSW has the requirements.

Emergency Data Information

Graig is a member of a Bureau-wide group that is investigating the dissemination of data during natural disasters. The committee has almost developed their first draft of the report but it has stalled in recent months.

AI: Janice. Get discipline offices to develop a policy statement on who needs data in an emergency.

AI: Janice. Get discipline offices to take the lead in forming a group to develop technical requirements for delivery of data in an emergency. Completion date for group formation is November 1, 1997.

Wednesday, August 20

DIS Analysis

Jerad presented background of DIS analysis method. DIS has been asked to do too much and there isn't a critical need to do some of their current functions.

Brian added that we are moving towards a more diverse environment so can we continue business as usual. Can we continue using a centralized support program? Our solution was to lay out the priorities then lay out the costs so we can determine the bottom line.

Verne noted that the importance figures are nearly all under 2.0 and therefore HQ should be applying more resources into DIS.

Isabelle asked whether there was significant overlap between HASS and DIS (example: JAVA).

Jerad answered that the scope of this report remains within the DIS program office but the overlap is an issue that was not dealt with by this report.

Alan was concerned that the direction of support was for Suns and SGIs and wasn't certain as to where this came from.

Both the FUTCOM2 report and purchases in FY97 both indicate an investment in Suns and SGIs and that centralized support is more efficient than using local support. Jerad mentioned that with 40 or so Suns are deployed in WRD and it is likely to be cost efficient to support SUNs and SGIs through DIS in FY98. FUTCOM3 has been charged to re-examine WRD's technical directions for planning for the NEXT three years.

Verne and Tom requested a chance to review the numbers in the tables and will send comments back to the committee.

Discussion continued as to where to draw the line in Table 1.

A suggestion was made to create an additional table to add detail to training costs.

Also, Software Technology Research and Procurement should be moved to between Security and GIS. NCTM `98 is a different type of item and should be removed from the table or incorporated in the SA training item because this is a nearly firm item.

Discussion continued as to the background of the Annual Costs listed in Table 2.

Jerad added that the listing in Table 2 are in alphabetical order not priority order.

Scott noted that we can't wait until the next CPAC meeting because of the implications to FY98's budget.

Jon stated that the process of setting priorities in Table 2 is non-trivial. The support issues included packaging, installation, user assistance, enhancement, platforms, etc. and at what level and costs. How do we design a process to make priority order and selection? Is commercial support available and at what cost?

Break

In the HW & OS Support section, drop ending clause of third bullet. In the second bullet add clarity for support of W95 and NT and not others. In the final section, it isn't clear how to correct the 84% ratio of salaries to total budget. Can the intent be clarified? Jon added that if the pass-through costs are added in (over $1M in contracts) the ratio and the meaning behind this section changes. Can a statement be added that recommends an increase in the amount of discretionary funds?

AI: A new draft with updated figures will be announced by the subcommittee by August 28 and comments from CPAC must be returned by COB September 2.

AI: Once approved, the preliminary document will be e-mailed to Verne from Graig. A cover memo will be attached by Graig and distributed to Sr. Staff.

AI: Tom will investigate BPA for leasing computers.

AI: RCS'rs will provide equipment inventory tables to FutCom3 by November 1, 1997.

LUNCH

NexWIS Architectural Concepts

Alan and Ken described some initial thinking on architecture options for future version of NWIS. With the several constraints and assumptions that are in the document, there is not a lot of flexibility. One of the assumptions is that DGs and Ingres will be used for at least the next 18 months.

Objectives of NWIS include to considerably improve the utility of NWIS for internal users by the end of FY2000 and to enormously improve the utility of NWIS for external users by the end of the year 2000.

Architectural options that are available include the server and DBMS for a central data warehouse, development standards, and tools for web enabling applications, and the common desktop environment. One element of the vision statement is to use an object-oriented data base and development environment National retrievals are a very important issue to be addressed. It was noted that the current NWIS data base does not include some of the NAWQA data.

AI: Alan will deliver to CPAC a technical and cost model analysis describing and justifying the requirements and solution of dissemination of national retrievals for USGS and public plus responses to emergency by September 19, 1997.

AI: Janice will provide CPAC with a list of Unix-based hydrologic applications by September 2, 1997.

Some initial Recommendations for NWIS development by Century Tech.

Current recommendations:

  1. Establish web development project
    Establish web development environment

  2. Establish national retrieval project
    Data Warehousing Solution

  3. Procure Windows-based workstations

BREAK

Web Designer's Workshop and Web Activities in WRD

The objective is to develop a Division-wide approach to serving NWIS data on the web. Focus on web-related applications for NWIS. Identify framework for doing web development (team, roles, environment, standards, tools) to obtain common ground. Identify methodology for web development. Identify design approach, testing plan, etc. Also, developments to date and have developers present their technical wares. Finally, develop an overall plan for disseminating data on the web. The first workshop should be a minimum of three days. The tentative target date is September 1997.

A leadership role needs to be established plus the workers need to be adequately funded.

Jon suggested scoping the objectives to meet the invited attendees. The compromise is to clarify that this first meeting will focus on surface water data.

Jerad added that there is a lack of good communication between the WRD Web development within WRD and district chiefs. The goals, objective, delivery dates, funding sources, structure, management, etc., should be reported to DC's via WRD memo.

AI: Verne and Kevin will report back with the results of the Web Developers Workshop at the November CPAC meeting.

Other business

AI: Bill will follow-up on AO replacement CPAC member for Sharon Goss.

Summary of Action Items

  1. AI: Verne. In order to compute return on investment (ROI), construct FY97 expenditures on NexWIS by November 3, 1997.
  2. AI: Jon and the Phoenix Project will define the lab data issue and Jon will report to CPAC via e-mail prior to the next meeting.
  3. AI: Janice, Verne, and Bill to write policy on who decides in the last bullet of section 10.1 and who decides what are `limited used data types' in the 15th bullet of the same section.
  4. AI: All. Write comments on sections 6, 7, 8, and 9 of the NexWIS Strategic Plan to CPAC via e-mail by September 5, 1997.
  5. AI: Greg Allord and Verne. Develop a Water Information policy statement by September 2, 1997 that states web-page development should be at HTML 3.2 after January 1, 1998.
  6. AI: Janice. Develop technical guidelines to quality assure provisional data on the web by November 1, 1997.
  7. AI: Alan to announce HASS web-page availability via e-mail by September 2, 1997.
  8. AI: Graig will contact Bob Wakelee about LAN/WAN monitoring and analysis training and report the status on the November 1997 meeting.
  9. AI: Tom will review and distributed Dangermond's letter to CPAC.
  10. AI: Tom will review new features of ARC 8 and determine their applicability to WRD production.
  11. AI: Verne will make the WRD strategic plan available to committee members prior to FutCom3 meeting.
  12. AI: Graig will firm FutCom3 membership by September 19, 1997 and e-mail to CPAC.
  13. AI: Alan will clarify the ConCalc requirements issue for the committee and will make sure that OSW has the requirements.

  14. AI: Janice. Get discipline offices to develop a policy statement on who needs data in an emergency.
  15. AI: Janice. Get discipline offices to take the lead in forming a group to develop technical requirements for delivery of data in an emergency. Completion date for group formation is November 1, 1997.
  16. AI: A new draft with updated figures will be announced by the subcommittee by August 28 and comments from CPAC must be returned by COB September 2.
  17. AI: Once approved, the preliminary document will be e-mailed to Verne from Graig. A cover memo will be attached by Graig and distributed to Sr. Staff.
  18. AI: Tom will investigate BPA for leasing computers.
  19. AI: RCS'rs will provide equipment inventory tables to FutCom3 by November 1, 1997.
  20. AI: Alan will deliver to CPAC a technical and cost model analysis describing and justifying the requirements and solution of dissemination of national retrievals for USGS and public plus responses to emergency by September 19, 1997.
  21. AI: Janice will provide CPAC with a list of Unix-based hydrologic applications by September 2, 1997.
  22. AI: Verne and Kevin will report back with the results of the Web Developers Workshop at the November CPAC meeting.
  23. AI: Bill will follow-up on AO replacement CPAC member for Sharon Goss.

Next Meeting

The next meeting is in Raleigh, NC, from 8:30am Tuesday, November 18, through noon on Thursday, November 20. Kevin Oberg will be the recorder.

February Meeting

The February meeting will be in Austin, TX, beginning at 8:30am on Tuesday, February 10, 1998, through noon on Thursday, February 12, 1998. Isabelle des Fontainnes will be the recorder.

 

Last modified: Mon Oct 20 15:34:30 PDT 1997