What's Network?

I wrote last month that the Emission Inventory would be our very first nationally mandated report that was implemented exclusively on the Exchange Network. I’m not retracting the congratulations—and thanks to the emissions folks for working so well with us. But I wonder if we’ve been a little too strict with definitions. It could be that the Drinking Water folks got there way ahead of everyone.

As we started to build out the Exchange Network, we discovered that one size did not fit all (imagine that!). Someone realized that a desktop client can do all the reporting and data retrieval actions that some partners need. Nodes are important, and they can do lots of things that clients can’t (like publishing data, listening for incoming reports, and automating routine processes). Let’s be clear: Without a lot of nodes, there is no Exchange Network. However, when we start to count national reports, we decided long ago that using a client is right for some.

Enter the “web client”. CDX has created a client that is web-based. It uses Exchange Network methods and security, and implements all the network primitive methods. In it’s most “advanced” form, it has labels like solicit and submit and query, and you get to type in the URL of your target node, or look it up in ENDS. All very network-ish. Also totally inscrutable to an average user. So CDX is building customized clients that use the same back-end functionality, but hide "network-speak" from the user, and have controls like “QA EIS Data” in lieu of Submit, flowname=EIS, service=QA, …

Now to the point. The SDWIS State app builds a file that is clearly EN-compatible. Everyone uses a standard XML format. But most users are uploading reports via CDX Web. The user credentials, security, etc. all live in a different box. However, I’ve learned that behind the screen at CDX, CDX Web just hands the file off to the CDX Node processing mechanism—so it sure looks to the Office of Water like everything comes through the Network. Some folks have told us that it is still easier to check status and get error info through CDX web. It gets worse. The drinking water staffer can use CDX web without help. In many (most?) states that use nodes, he or she has to e-mail the file to the node administrator and ask them to submit it.

To distill the message: If you use our easy to use client and your credentials are stored in CDX Web, you are accomplishing the same thing, but no gold star. If you use NAAS credentials and a node or an inscrutable client that is less functional, then we’ll give you the gold star. This seems hard to explain to an end user. Maybe we shouldn’t try.

Here’s how we get to a more complete and robust network: We recognize that the benefit to EPA (for inbound reporting) is the ability to use a single channel. We go back to solid IT work--we build solutions for users that they can understand and use. It is our job to choose the right tool and reuse our investments. I think we agree that should include using the Network, but the users don't care, and shouldn't have to. If the client we gave the drinking water folks isn't network, replace it with one that is. It is OK if the user's can't tell the difference.

I’m not sure where this leads. Perhaps we check off the SDWIS federal reporting as “done”. If so, then here’s the next level for drinking water: Go beyond delivering mandated reports, and publish and share actual monitoring results. The health community will love that—they’ve been dying for seamless access to that information. By the way, the Association of State Drinking Water Authorities (ASDWA)is already working on something along those lines. More later.

ps—That will require nodes—maybe CDX, or maybe state nodes—probably both.

Give Chris the Hat!

Apparently the Wizard of the Wiki doesn’t blog! In my first-ever network blog, I challenged the EN community to find and make available the much-rumored Stylesheet that makes XML TRI files look like their paper counterpart. At the user meeting in Atlanta last week, Chris Staten (MS DEQ) got word of the challenge, and came through.

So it can be done. Now we just need to figure out how to share more easily. If it’s hard for me to find stuff—well, it is just too hard. If you have something that you think is worth sharing, I urge you to post it on the Wiki. Is that the ideal way to manage software distribution? No. Do we have anything else that easy for everyone to use today? No. I only wish we had so much stuff that managing it became an issue.

Speaking of Atlanta, It was great to see nearly all (both?) my regular readers there. If you weren’t there, where the heck were you? This may have been the best yet—the feedback received so far is great, but please respond to Kurt’s survey (anonymously). Tell us what was lame, or risk seeing it on the agenda again. For me, it was a case of too many good things. I missed a lot of things I wanted to see. If your presentation was one of them, I’m sorry, but it is bad form to wander off while moderating a session! There was nothing on the agenda that I didn’t want to see.

We introduced the meeting with a good news/bad news routine. The good news is that the network is behind the coolest application in decades. The bad news is that it only works in Georgia and North Dakota. We spent a fair bit of time talking in that session and throughout the week about the implementation matrix. I expect we’ll keep talking. This is a document that for the first time clearly lays out our progress in building a national network. The news isn’t all good, but there is plenty of room for hope. As soon as the matrix came out, people started explaining why things were better than it looks. A lot of the blank spots and inactive flows have already been fixed. If you saw what you think are errors, post comments on the Wiki. This tool could use some sharpening. Please remember that this is only counting national system flows. Not only are those one important line of network business, but they are the one we can measure.

Don’t get too hung up on the piece of paper, though. That matrix is so last year. Literally. Disregard the title at the top-the data was from CY 2008. Anything that didn’t happen in 2008 can’t ever happen in 2008. The Network governance has committed to fine-tuning this report and reproducing it regularly. I urge you to focus on the next edition—not the last. Use the Wiki to tell us what we might have gotten wrong, and to tell us what’s standing in your way. Implement or resurrect flows to change the colors in the boxes.

As usual, in Atlanta our tribal partners showed how they are redefining Exchange Network success to be what makes sense for them. I urge all partners to do the same. You get to define what success looks like for you, so we can help you get there. If you are "on the hook" for a flow that makes no sense to you, let us know that, too. The Network is about doing things smarter—not doing things that you don't need to do!

And to take it back to the top of the page, if you have a whole line of green boxes, look in them for any treasures you might be able to share. Your network is the network of the folks who are struggling. In the end, yours will share the same fate as theirs.