Table of Contents

• Using Data Acquired by Unidata
• Kudos to NSFnet (and Its Replacement)
• McIDAS Update
• McIDAS Documentation Update
• WXP Moves to Community Support
• Last 5 Months for Unidata/Wisconsin Satellite Broadcast
• IDD Update
• What's New on the Unidata Web Server
• Unidata Users Committee: Growing to Serve You Better
• Possible Rule Changes for Unidata Participation
• Peggy SOO?
• Hello, Jo

Using Data Acquired by Unidata

General Guidelines

1. The reliability of information gained through Unidata is intended to be satisfactory only in an education or research context; in those contexts, it is assumed that outages and errors are expected and dealt with by the users of the data1. These data are not suited for operational or decision-making uses, and it would be dangerous to use them in those contexts.

2. Unidata is funded by the National Science Foundation (and, for facilitating access to NEXRAD Information Dissemination Service [NIDS] data, by the National Weather Service [NWS]) to support education and research. Therefore, the terms and conditions under which Unidata acquires data reflect that objective.

Specific Rules

• As long as the data remain within a single organization and physical campus (typically a university department), all education- and research-oriented usage generally is permitted. In some cases, such as for lightning data from Global Atmospherics, Inc., the individual data access agreements require that research results be shared with the data provider and that attribution be included in all publications where the data (or derived values) appear.

• Great caution must be exercised in placing data on servers or otherwise permitting access from outside the organization, such as through the World Wide Web. Such uses are permitted for data acquired from the Unidata/Wisconsin service or the NWS Family of Services (FOS), as long as the purposes for making them available (such as to support K-12 school projects) are educational or research oriented. However, other types of data, such as lightning or NIDS data, may not be published on the Web or otherwise shared outside the receiving organization except with the explicit approval of the data provider.

(In the NIDS case, data over 48 hours old may be distributed freely, but earlier off-campus uses must be approved by WSI.)

Rationale

It may be illustrative to examine in greater detail three distinct cases--lightning data, NIDS data, and FOS--from among the various data streams that can be acquired with Unidata systems. For lightning data, all steps in the process of taking measurements, assembling the data, and creating products are performed by a private firm (Global Atmospherics, Inc.), so there is no question about who owns the data; in this case it is perfectly clear that access to these data by Unidata universities (especially on the current no-cost basis) represents a significant opportunity regardless of any usage constraints that may be imposed.

In contrast, the FOS data are collected and assembled into products at government expense, and anyone who acquires these feeds directly from the NWS (at a government-defined price) can use these products essentially without constraints. For reasons of economy, Unidata decided not to acquire these data directly from the NWS, instead choosing (after an open competition) to make arrangements with Alden, Inc., for injecting the FOS into Unidata's Internet Data Distribution (IDD) system. These arrangements limit usage to education and research, but that definition is consistent with Unidata's mission.

The NIDS system represents an interesting combination of public and private efforts. While the NWS operates the observation platforms, the system relies entirely on the private sector to operate the networks by which data are assembled on a national basis. Also, the only nationwide composite images of NEXRAD data that exist are prepared in the private sector. Due to the joint nature of this public-private effort, a small number of commercial firms (the four NIDS vendors) have proprietary interests in the data produced. This situation makes it impractical or impossible for Unidata to offer universities these data with the same flexibility as for the FOS.

Therefore, Unidata also held an open competition for the provision of NIDS data, in which usage flexibility was evaluated along with other factors such as price. Unidata's present agreement with WSI, Inc.--an agreement that has broken new ground and yielded promising results--grew out of this competition. In particular, we are very appreciative of WSI's trust in the IDD method for delivering these proprietary data. We want to encourage all Unidata users to help us warrant this trust by understanding and adhering strictly to all usage constraints that pertain. More detail can be found in the spring 1994 issue of the Unidata Newsletter, which was a special edition on data delivery issues.

Kudos to NSFnet (and Its Replacement)

by Sally Bates, Newsletter Editor and Information Manager at the Unidata Program Center

April 30, 1995 marked the end of a successful experiment. On that date, the National Science Foundation (NSF) pulled the plug and the NSFnet backbone disappeared. The event had been planned for years, and as with any major change, it was awaited with feelings of dread and predictions of apocalypse. Instead, thanks to the diligence of the Computer and Information Science and Engineering Directorate at NSF, most users probably didn't notice the demise of the backbone--it was smoothly replaced by a multitude of commercially based networks.

The Unidata Program has been dependent in many ways on the implementation and growth of NSFnet. NSF began building NSFnet in 1986, just as Unidata was preparing to implement its program of software support. The Unidata Program Center (UPC) was worrying how, with limited resources and a small staff (which in 1986 grew from 3 to 5), we would communicate with and support a widespread user base. In fact, one of the early ideas for the program was to build a communications system for the atmospheric sciences. NSF's decision to implement NSFnet allowed Unidata to concentrate instead on other community needs.

NSFnet was operational by the time Unidata held its first training workshop in early 1988. At that workshop, the staff established the now well-known pattern of asking participants to rely on e-mail for support ("Please, don't call us; send us e-mail and we'll get back to you as soon as we can!"). But many participants were from campuses without connection to the Internet.

NSFnet's model of building a distributed network helped alter this picture. Although the initial backbone connected only the NSF supercomputing centers, the foundation it established encouraged and subsidized the development of regional networks and facilitated their connecting to the backbone. NSF also provided grants to encourage institutions--including colleges and universities--to connect to the Internet. Unidata participants were urged by the UPC staff to prod their campuses to become connected. Indeed, the opportunity to fully participate in the Unidata Program provided the impetus for at least one campus to join the Internet--it was the first clear benefit the administrators could see for the expenditure.

Once established, the NSFnet grew at an astounding rate--the perfect tool at the perfect time. Starting with 56-kb lines in 1987, the network moved to a T-1 backbone in 1988 and a T-3 backbone in 1991. And the demand grew as the bandwidth increased.

Uses for the network expanded almost as rapidly. FTP became the workhorse tool, at its peak in April 1992 accounting for 30 percent of all packets sent over the backbone. Unidata's first electronic server, still in existence, was an FTP server from which users could obtain software, bug fixes, and some documentation. As the network grew, so did the number of FTP servers and finding information on the Internet became difficult. In 1990, Canadian Peter Deutsch and a team at McGill University developed Archie, the first Internet searching tool. Archie periodically searches through all the FTP servers available on the Internet and creates a database of the files found. Users can query this database using keywords to guide them toward needed information. Archie became so popular that the link between NSFnet and CA*net in Montreal became overwhelmed and other sites had to be found to mirror the Archie database.

In 1992, Mark P. McCahill and his team of computer scientists at the University of Minnesota developed a tool called Gopher. Like FTP, Gopher included information servers, but was much easier to set up and use. Furthermore, the Gopher client tool was much more efficient--the client obtains a directory listing of files from the server and users browse for the files they want on their own machines, not the server machine. (See the fall 1992 issue of the Unidata Newsletter [not yet available on the Web server, sorry]). Gopher was soon ported to DOS, Microsoft Windows, and Macintosh operating systems and quickly became widely used.

At about the same time, Thinking Machines, Dow Jones, and Apple Computer joined forces to develop a sophisticated database searching tool called WAIS (for "wide-area information server"). Unidata quickly set up a Gopher server containing a WAIS database of support e-mail so that users could search through previously asked and answered questions. The database, still in use, has been popular and has significantly eased the load of support queries to the UPC.

While Gopher and WAIS were making file exchange easier, physicists at CERN in Switzerland were busy designing a system--dubbed the World Wide Web (WWW)--for linking files across the Internet. The Web technology was primarily a tool of the computer-literate until the fall of 1993, when the National Center for Supercomputing Applications released its now-famous Internet browser--Mosaic. With Mosaic, a nontechnical user could access FTP, Archie, Gopher, WAIS, and the World Wide Web without having to learn about the underlying software or to deal with arcane UNIX commands.

Again, the Unidata Program benefited from new Internet technology. Our Web server grows daily (see "What's New...") and we have been able to efficiently enhance our support services. In fact, the explosive growth in both development and use of WWW tools is leading us to think about entirely new mechanisms for providing education and research materials, case studies, scientific data, and other information vital to the Unidata community. Somehow, it is fitting that during the last month of NSFnet operation, WWW protocols finally overtook the old workhorse FTP as the leader in packets transmitted.

So, thank you, NSF. You brought the Internet from infancy to adulthood, and by happy coincidence, this allowed the Unidata Program to mature as well. May your next network endeavor be as successful.

McIDAS Update--New Data, New Releases

New images

At the same time, we added increased coverage of surface and upper-air stations in the broadcast MD files. GOES-8 water-vapor images were also slated to be added to the datastream at this time. This proved unfeasible for the satellite-broadcast version of the Unidata/Wisconsin channel due to bandwidth constraints, so the images were added only to the IDD.

After the addition of the VIS and IR images, the Unidata/Wisconsin channel is essentially full. The added stations in the MD files are not the problem, rather it is that the pixel-to-pixel variability of the visible product is so high that it limits the compression attainable with the run-length encoding scheme used in Unidata/Wisconsin channel preparation. It can take up to 16 minutes to transmit the VIS product, so we have changed the broadcast schedule within the hour for the products. Bandwidth is less of a constraint with the IDD, allowing us to distribute many more products via IDD than the satellite broadcast will allow.

The new versions of McIDAS (see below) contain enhanced function-key menus which allow push-button access to the new images. Because the new menus take advantage of functions found only in the new releases, you will need to upgrade to the new versions if you want this functionality.

New releases

At the beginning of June, new versions of McIDAS-X (2.0) and McIDAS-OS2 (6.0) were made available to the Unidata McIDAS community. These new versions contain many improvements over the previous releases of Unidata McIDAS. Some of the significant features common to both the OS/2 and X releases are

• An enhanced function-key menu--including support for GOES-8 and NIDS imagery--that continues to make access and display of data easier for novices and experts alike

• A newly reworked color bar utility useful in annotating imagery loaded with stretch tables

• A reworked front-drawing program that includes spline fits to frontal locations from ASUS1 reports and barbs along fronts

• 4-panel displays of imagery

• A reworked TXT2MD command used to import non-McIDAS, point-source data sets into McIDAS

• An enhanced version of LA supporting the new AUX-block information in AREA files

McIDAS-OS2 version 6.0

The new version of McIDAS-OS2 now uses the freely available EMX/GCC and F2C compilers for code development. You can obtain these compilers from the Unidata FTP server.

The most significant enhancement in this release is support for higher resolution displays of satellite imagery through the use of the OS/2 Presentation Manager (PM); the PM allows 48 colors, versus only 13 in VGA mode. McIDAS-OS2 PM runs on the desktop with other OS/2 applications instead of as a full-screen session like VGA mode. In PM mode, text and commands appear in one OS/2 window, image frames display in another window, and there is an additional briefing window (see illustration ). The briefing frames are separate viewports into the image frames defined in the system. Each briefing frame can display a separate animation or static display, allowing you to have multiple loops running at the same time in different windows.

McIDAS-OS2 PM also supports separate image and graphics planes like McIDAS-X. This allows you to change image enhancements without affecting the 8 overlay colors. You can specify the image-frame size during setup and can select from two operating modes to optimize loop speed or memory use. If you optimize loop speed, you use twice as much memory for each frame, but looping is smoother. If you optimize for memory use, the wiping effect is more noticeable when stepping between frames in a loop. In addition, when memory use is optimized, window information is automatically rescaled if you change the window size.

In order to run in the Presentation Mode, you must have a system that supports Super VGA and you should have at least a 66 Mhz 486 processor and a minimum of 16 MB RAM for smooth looping. The rewards are worth the cost of the extra system requirements. A 100 Mhz Pentium with 32 MB RAM can display five animations simultaneously without any noticeable degradation in system performance. Of course, you can still run in the lower resolution VGA mode on your present systems as long as you are running one of the supported versions of OS/2.

Other significant enhancements include a routine (SVGIF) in PM mode for saving frames directly to GIF images, an enhanced SETUP routine that handles IDD connection setup automatically (no more manual editing of files), and an image-handling McBASI script for the F-key menu that makes it easier for the user to zoom in on satellite and NIDS imagery.

McIDAS-X version 2.0

The briefing-frame concept contained in McIDAS-OS2 6.0 is slated to be added to McIDAS-X for the winter 1995-96 release. In addition, we hope to release a port to DEC OSF/1 later this summer.

You can find out more about the details (including installation instructions) of these releases from the new McIDAS-OS2 and McIDAS-X homepages. There's also a page discussing what's upcoming in McIDAS. All of these are linked on the McIDAS homepage.

McIDAS Documentation Update

Manuals available via FTP

Currently, the following manuals are available via FTP from the docs directory in the passworded McIDAS account:

• McIDAS-OS2 Users Guide--SSEC portion only

• McIDAS-X Users Guide--SSEC portion only

• McIDAS-X Learning Guide

• McIDAS Applications Programming Manual

We will be updating the Unidata portions of the manuals (the Site Manager's Guide and the Guide to Unidata Testcode) this summer and will make those portions available when they are ready. One of the challenges we face with these two manuals is that the SSEC and Unidata portions for the command references have been produced separately. We are working with SSEC to find a way to merge these parts to make it easier for users to find the information they need. Our ultimate goal is to provide these command references on-line in a format that can take advantage of the World Wide Web (WWW) capabilities.

McIDAS tutorial now available

• manipulation of satellite imagery, gridded data, and surface and upper-air data

• frame and loop controls

• string-table manipulation

• graphic and image enhancements

The Learning Guide is designed to work with a data set from the March 1993 "Storm of the Century" over the East Coast of the United States. This data set includes high-resolution satellite imagery, gridded data sets, hourly surface observations, ship/buoy reports, and synoptic observations. The data set is available in a 29 MB compressed tar file or a 30 MB Zip file in the docs/tutorial directory of the passworded McIDAS FTP account. The data set is about 90 MB when uncompressed, so be sure you have plenty of disk space available.

McIDAS WWW documentation

Web documentation is also available on how to use the IDD to receive McIDAS data. You'll find complete instructions for downloading, installing, and configuring LDM-McIDAS and instructions for OS/2-only sites on how to receive the Unidata/Wisconsin datastream via IDD (see "Last Five Months..."). The McIDAS homepage also includes links to the McIDAS e-mail archive which allows quick access to McIDAS questions and answers. There are links to sites that use McIDAS to generate products for their WWW servers, including the new Unidata McIDAS Demonstration Machine, which uses McIDAS-OS2 to generate products for an OS/2 WWW server. McIDAS batch file and command examples used to generate the products are available on this server. Comments in the batch files provide tips on how to generate your own products.

Our goal is to make as much of the McIDAS documentation as possible available on line. If you have any suggestions on improvements to the McIDAS documentation or on new features you would like to see added to McIDAS, please send a note to support@unidata.ucar.edu.

WXP Moves to Community Support

by Dave Fulker, Unidata Program Director

The UPC is planning changes in respect to future Unidata support for WXP. These changes have been prompted by the departure last winter of a valued staff member, Mike Wright, which left us shorthanded in respect to WXP expertise. And the effect is exaggerated because Wright--as a half-time research assistant--provided a level of support well beyond what he was expected to perform.

The timing of this loss is unfortunate because Dan Vietor is planning a major new release of WXP that he says will differ structurally from the present version in significant ways. Wright played a major role in the release engineering for new versions of WXP. He created makefiles and tested and debugged all revisions on the hardware/software platforms (e.g., ULTRIX, AIX, and SunOS) common among Unidata users. He also packaged WXP for FTP retrieval from Unidata and revised the tutorials and various automated scripts to match the changes and to take advantage of new features. We anticipate that the new version of WXP will require significant release-engineering effort because of its new structure.

One may ask why we do not simply replace Wright, but next year's budget picture is so uncertain that it would be unwise to do so. In any case, filling his position (as a research assistant) with someone having the requisite WXP and C-programming experience would be unlikely. So we have been considering alternatives and seeking the guidance of the Unidata Users and Policy Committees. Our conclusion is to move toward a "community support" approach to maintaining WXP as a viable Unidata offering, and we ask for your cooperation and support in gaining this economy. The Unidata Users and Policy Committees have endorsed this move.

As we now envision them, the steps leading to community support for WXP will be as follows:

1. We will engage--on a temporary or contract basis--a qualified software engineer to release-engineer Vietor's newest version of WXP, adapted for use on Unidata-supported platforms. (We can afford, for a few months, the level of expertise required.)

2. After the new release has stabilized, WXP-specific questions and problems received by the Unidata support staff will be circulated on the WXP mailing list, with the trust that WXP community members will aid one another, as has occurred frequently in the past.

3. To maintain continuing viability for WXP, we will solicit community volunteers to play certain roles, such as compiling and testing WXP revisions on specific hardware/software platforms. With the help of these individuals, tested and up-to-date versions of WXP will be stored at Unidata (as source code) for FTP access by licensed users.

In developing this plan we considered and rejected the idea of providing a higher level of Unidata support for WXP by having one or more of our McIDAS or GEMPAK support people gain WXP expertise and split their efforts over two packages rather than focusing on one. Our conclusion was that such a move would result in inadequate support and impede our progress toward versions of McIDAS and GEMPAK that will be easier to use and meet a broader spectrum of needs. One such area of progress is our objective to increase the capabilities of OS/2-only sites in conjunction with their transition to Internet Data Distribution; this goal requires both McIDAS and LDM-related efforts. (I note that, with its Web server and other features, the OS/2 environment looks increasingly attractive for Unidata sites where running UNIX is problematic.) Stated simply, we think the nature of WXP makes it a reasonable candidate for community support, while concentrated support remains essential for the others.

We are still soliciting feedback on this plan. I would be especially pleased to hear from any of you who might be interested in becoming Unidata's WXP/xxx expert, where xxx is any computer platform of significant interest to universities.

With your help, I believe we can move successfully to "community support" for WXP, and that under this plan WXP will remain a strong player, satisfying those who prefer its design philosophy and its style of user interface and who wish to utilize new features as Vietor's energy permits him to add them.

Last 5 Months for Unidata/Wisconsin Satellite Broadcast

by Linda Miller, External Programs Coordinator at the Unidata Program Center

The end of an era is upon us, but things will only get better. The Unidata/Wisconsin (McIDAS) Satellite Broadcast will disappear on December 31, 1995. This broadcast has been the mainstay of Unidata from the beginning. Why end the broadcast? By moving to the Internet Data Distribution (IDD) exclusively, additional funding will be freed up to modify and add additional products to the Unidata/Wisconsin (McIDAS) datastream.

Currently, the slow speed of the satellite transmission (9600 bps) prohibits the addition of many new data sets. By moving to the IDD, the bandwidth contraints are minimized. In the short term, none of the nonsatellite products will be deleted from the datastream until solutions for incorporating data from the other circuits into McIDAS can be developed. Additional data sets available through the IDD include the full range of International and High Resolution Data Streams (HRS), which include the Rapid Update Cycle (RUC) and the Eta Forecast Model.

It is important that you begin your transition to Internet data reception now. The Unidata staff is always available to help you. We have been putting information on our World Wide Web server about how to proceed in receiving the Unidata/Wisconsin channel via the IDD.

If you don't have a web browser installed and don't know where to begin, or if you need general assistance, don't hesitate to send e-mail to support@unidata.ucar.edu.

IDD Update

by Linda Miller, External Programs Coordinator at the Unidata Program Center

If your site is not currently participating in the IDD system, we need to talk to you. We are trying to determine what steps remain to get each of you participating in the system and how we can help.

The IDD now includes 96 sites and more are preparing to join. There's still time to reap the benefits of participating. It's a great way to make connections with other sites and be a good neighbor.

You can access information on how to participate here. You can also send e-mail inquiries to support@unidata.ucar.edu. Or call me at (303) 497-8646.

Current directions

The more we work with the IDD, the more we realize it will never be a static project. One recent change is a new Topology Configuration Plan that is being created in collaboration with the IDD sites.

Statistics continue to play a key role in the IDD. Collecting, analyzing, and displaying them in a useful way has become a major responsibility. Some sites are not sharing their statistics with Unidata, but we encourage everyone to do so. The statistics help us determine where the data delivery is breaking down and work on solutions. See the current IDD statistics here.

The Local Data Manager (LDM) development continues with new capabilities being planned. The LDM 4.1 Site Manager's Guide is available to assist site representatives with installation and preparation for the IDD.

What's New on the Unidata Web Server

On the homepage:

• We added a new section devoted to UDUNITS (Unidata's software library that's used to convert and manipulate unit specifications such as temperature or distance).

• Under "Other News", we've added a summary of the questionnaire we conducted on displaying data in the classroom.

• There's a link to our searchable Gopher directory of community members.

In the "McIDAS" section:

• McIDAS-OS2 homepage

• McIDAS-X homepage

• McIDAS tips and tricks

In the "Data" section:

• We added a paper explaining the guidelines governing the usage of data acquired through Unidata (see the cover article).

  http://www.unidata.ucar.edu/data/data_usage.html

• We created a Unidata/Wisconsin Datastream page.

Elsewhere:

• In the area devoted to the Unidata Users Committee, we now make available the latest meeting summary and maintain an archive of past summaries. The committee last met in April 1995.

• In the area devoted to Other Unidata Committee Information, we have placed materials supporting the July Advanced Technical Advisory Committee (ATAC) meeting. We are experimenting with providing meeting materials on line instead of in hardcopy documents mailed to committee members. This allows committee members to receive up-to-date information.

• We added a link to the Cooperative Institute for Research in the Atmosphere "Introduction to GOES-8".

• We have also created an entirely new Web server, which we call the Unidata McIDAS Demonstration Machine. The server contains HTML documents on a "GoServe" server, a freely available Gopher/WWW server for OS/2. The products displayed on this server are generated automatically as the data are received from the Unidata/Wisconsin datastream via the Internet Data Distribution (IDD) system.

Unidata Users Committee: Growing to Serve You Better

by Mohan Ramamurthy, Unidata Users Committee Chair

The Unidata community continues to grow--there are now over 140 institutions participating in the Unidata Program (not counting multiple sites at several universities). In the early days, a majority of the sites were using Unidata systems primarily in synoptic meteorology instruction and research. But in recent years, participation in the Unidata Program has expanded to include other areas of the atmospheric sciences. In an effort to represent the varied views and needs of this large and diverse community, the Unidata Policy Committee has recently expanded the Unidata Users Committee from seven members to nine.

The Users Committee, whose members are appointed by the Policy Committee chair, is the primary mechanism for gaining scientific feedback on the effectiveness of the Unidata Program. Appointments to the committee reflect the range of large and small colleges and universities with undergraduate and graduate emphases where Unidata systems are in use.

The committee meets at least twice yearly and reports to the Policy Committee; members are appointed for three years. The committee is charged with three tasks:

1. determining the attitudes of the user community toward the Unidata Program,

2. soliciting suggestions for additions to the data stream and software products, and

3. facilitating the exchange of ideas among users on the use of Unidata systems.

The Users Committee is here to represent you and can achieve its mission only with input from the community at large. We hope you will contact any one of us when you have comments or concerns about the current program, or ideas for possible future activities, or just to say hello. There is also a Users Committee mailing list. To send a message to all of the members (and others as well) send your e-mail message to usercomm@unidata.ucar.edu (the Users Committee e-mail list is now a closed list).

Possible Rule Changes for Unidata Participation

by Dave Fulker, Unidata Program Director

The Unidata Program exists to serve the nationwide university community, as reflected in the mission statement1 and the participation policy. However, the boundaries of this community are not well defined, reflecting the broad and varied education and research roles that community members play. Consequently, occasional questions arise about precisely which organizations may access Unidata products and services. We are presently considering changes in the policies that pertain to these questions, and we welcome your feedback.

For several years, the Unidata Program Center (UPC) has employed a policy 2 that embraces two types of participation: core membership, for traditionally organized departments in U.S. universities, and associate participation, for organizations that offer potential synergies with the core community. The benefits that associate participants gain have been minimal but have included a) access to data if the associate has an appropriate arrangement (usually a paid subscription at a non-Unidata rate) with the provider of those data; b) attendance at Unidata training workshops on a space-available basis; and c) access to software developed by the UPC, including enhancements to GEMPAK and McIDAS if the associate has licensed these packages through the normal, non-Unidata channels.

Problems with this policy have included

• Imprecise definitions, especially for associate participation

• Confusion for core members about the Unidata status of their research partners

• Lack of interest after initial contact by most potential associates

• Failure to gain hoped-for synergies

In other words, if sponsored participants wanted to use data streams or software packages outside the limits imposed on Unidata by the providers, they would have to make their own arrangements directly with the providers. At present, a typical sponsored participant would have to work with Purdue regarding WXP, with the University of Wisconsin regarding McIDAS, with Cosmic regarding GEMPAK, and with WSI regarding NIDS data. If sponsored participation widens the scientific and educational benefits that derive from Unidata, we may negotiate broader permissions to distribute software and data to sponsored participants. Current core members of Unidata would remain in that category, and we would seek sponsors for each of the associate participants that now exist.

We think the approach has merit because the definition of a sponsored participant would be clearer than the current definition for an associate member, and the approach seems more likely to yield the kinds of external relations we have been seeking. This approach limits demands on the UPC for support and protects the rights of those who provide Unidata with software and data.

The Policy Committee will work to refine this policy before deciding whether to implement it. We welcome your comments.

Peggy SOO?

Q: What do you see as your most significant achievements at Unidata?

Bruehl: I think the most significant is the growth of the GEMPAK community over the past three years. (from less than a dozen active users to almost 100). While I don't pretend that it was all my doing, I do like to think that I was able to help foster that growth.

I think the things I enjoyed the most were working with individual users--and solving particularly sticky problems. I think that's what kept me going each day, the feeling of accomplishment. (Of course, I used to get a kick out of doing geometry proofs in 6th grade, too. Go figure...)

Q: Anything you had hoped to accomplish but weren't able to?

Bruehl: Well sure, there's always something else one can do. I would have liked to spend more time on ease-of-use issues, such as tutorials and user interfaces.

Q: What are the duties of your new position? How do they fit in with your interests and goals?

Bruehl: Well one of my goals is to bring the operational community closer to the academic community. One of the ways to do this is to make comparable software and computing resources available to the weather service offices in the field. That's the step the NWS training program is taking with the SOO (Scientific Operations Officer) community. The NWS is putting Hewlett Packard workstations into each of the 100 or so forecast offices across the country. In keeping with the government requirement that every project/program/idea have a three letter acronym, they've dubbed this the SAC (Scientific Applications Computer) program.

As the National SAC Coordinator, I am helping the SOOs and others in each forecast office learn to work in a UNIX environment and with the GEMPAK and N-AWIPS software. It's not a small task, since most offices are UNIX-phobic. I'm challenged by the role and already learning lots of new things.

Q: Is there anything you'd like to tell the GEMPAK community?

Bruehl: Send money...no, just kidding. I hope that the community knows that I really enjoyed working with them. I will miss corresponding with them.

Hello, Jo

When she isn't working, you'll find Jo skiing, hiking, or playing tennis. We hope you all get a chance to meet her as you come to our offices for meetings and workshops.