HL Arledge

EnvisionConnect Batch Processing Faster than Envision

Customers and conference attendees have told us time and time again that financial batch processing in Envision was the fastest in the industry. EnvisionConnect, they also reported, is much easier to use, but not very fast compared to its predecessor. Decade's Development Team took that as a challenge, and early trials have everyone buzzing.

Smart Client Technology—the key to working on the web without giving up a powerful user interface—provides that all is processed client-side, including batches. Envision's advantage in the batch realm was its good old-fashioned client-server technology. Batches processed on the server run fast. The challenge in EnvisionConnect was to move the batch processing back to the server side without sacrificing the intuitive interface EnvisionConnect customers have grown accustomed to.

Through exhaustive research, careful prototyping, and much discussion with EnvisionConnect users, the Development Team provided a solution to a problem.

Perform Aging and Penalization and Perform Permit Billing are, by far, the most powerful of Decade's batch processes. From day one, EnvisionConnect was developed with a "hard things first" mentality.

Developers stayed the course with the batches, converting these two batches to server-side processing before all others, and in side by side trials—same data, same configuration, same hardware—these two EnvisionConnect batches now smoke the Envision batches, and they run in a fraction of the time it took the client-side batches to run.

In fact, Perform Permit Billing now processes half a million records in under one hour!

Want to know more? Join us at the Decade Software 2009 Training Conference, October 19th and 20th.

Customer design or Too Many Cooks Sink Ships

When designers attempt to please everyone, someone always ends up angry. The key to good design is bringing all interested parties together in order to balance interests and find solutions that work for everyone.

In the early days of EnvisionConnect, customer advocates on staff shouted:

“Our users know what they need better than anyone else. Let them design the software and force the developers to deliver exactly as they define.”

Who could argue with that philosophy? The customer knows best, right?

Let’s return to the mixed metaphor of “Too many cooks sink ships,” and let me tell you a story…

There once was a shipbuilder who hired the best engineers, but he sold ships to the government and was forced by contract to design the ship according to specs put together by a government committee.

The shipbuilder ordered the designers to build every gadget and widget the government committee could dream up, and the ship that evolved from the process was beautiful!

On this ship, you didn’t have to take the stairs to go from deck to deck. Everything in the ship was put on one easily accessible deck. And all of the gadgets, widgets, bells, and whistles glistened. It was the most spectacular-looking ship the government committee had ever seen.

When war broke, the government ordered the ship launched. It would sail to the other side of the globe and squash the evil doers.

Unfortunately, the ship was too heavy. It moved too slow. By the time the ship reached the enemy, the war was over. That’s not all that bad really. If every country followed this design model, the world would be a more peaceful place.

The moral of the story is that all interested parties have to be involved in design.

Luckily for us, the world of Agile software design doesn’t work like ship building.

We do insist that users of the software design by committee on the first pass (in industry jargon, the first iteration) but then it’s up to the engineers to weigh the needs of the many against the wants of the few and refine (industry jargon again, refactor) the product into that happy medium that makes everyone happy.

This year, our customers will notice that we’re changing the way we collect customer feedback—finding ways to extract information faster without wasting customer time, and our developers are defining “the how” and restricting customers to defining “the what”. And when some “what” is ultimately detrimental to a more important “what”, it’s our job to help customers understand the trade-off they are asking to make.

Now, you know why it’s taking a while to christen EnvisionConnect 4.0.

EnvisionConnect 4.0 as she stands today is the most beautiful and the most powerful ship in the fleet, but Envision outruns her. As goals go, our next BIG target is making EnvisionConnect faster, even if it means dropping a few bells and whistles overboard.

Watch this space to see how we do.

Another secret operation in progress

Last week, I was the guest speaker at one of our “global user group meetings”.

I’ve forgotten what the official title is these days, but basically it is a virtual meeting that allows customers to review and discuss new design proposals.

These little webinars were launched in the early days of EnvisionConnect, and they continue to serve us and our customers quite well.

You may recall that EnvisionConnect is the 21st Century reincarnation of Envision for Windows.

Over time, new wings and windows were added, until Envision grew into the most in depth tool available for regulating Environmental Health—but like all mansions that outgrow their original design, Envision became easy to get lost in.

Enter Envision Connect…

EnvisionConnect was designed from the ground-up to have all of the comforts of the old mansion, but with easier passages to move around in.

The new design was developed in those Thursday meetings.

So, why did I attend last week?

I’m building another kind of house to replace our web home page, and for all of the same reasons, I needed our customers to assist in the redesign.

What we came up with will amaze you…

…but you’ll have to wait until tomorrow to find out more.

Who's afraid of Orange County?

Orange County Environmental Health is one of Decade Software's largest customers—if you weigh large as the number of users. They have been Envision users for many years. In fact, their insights helped shape Envision 3.4 into the solid product that it is today.

Lately, OC has been contemplating a move to EnvisionConnect, and this week, they invited Client Services Manager John Jensen and I down south for a pow-wow.

In the days preceding our trip, the mumbling among Decade staff echoed down the halls.

"I pity JJ and HL. Those Orange folk are tough to deal with."

"Yea, they're really pushy. If there's a feature they want, they won't stop complaining until they get it."

"...and they're tough on defects, too. If Nick Reyes finds something wrong, you'd think the world was coming to an end."

I laugh anytime I hear such chatter.

I've worked with Nick Reyes, Karen Newe, Jeff Warren, Pearl Boelter, Ron Margheim and the rest of the OC crew since they came onboard, and I appreciate the fact that Orange is one of the many clients that helped make our products as solid as they are today.

We weren't always delivering products with zero defects

I promised earlier this week to provide the statistics I reviewed in my last public presentation. I will begin to keep that promise today. However, since you do not have the benefit of hearing the full presentation, you will not benefit from the full story of the many process improvements Decade Software has made.

Along the way, we have incrementally improved our processes, learning from the Rational Unified Process, Six Sigma, and others. Additionally, Kevin has invested more resources in the last year than any year prior.

I wanted to mention these little caveats to make it clear that I do not attribute our successes solely to Scrum or current leadership alone.

And with those disclaimers aside, let's look at the numbers...

During the life of Envision for Windows and it's related products, there were 8,596 defects reported, but—in spite of our best efforts—we were only able to close 2,221. Concerned with the defect count—and the fact that the 21st century web and smart-client technologies were improving usability standards worldwide—we scrapped Envision's user interface and began anew.

Two years later, EnvisionConnect had improved usability, but we seemed to be continuing our defect trend. A total of 472 defects had been opened, but only 190 had been closed.

We had to take measures more drastic.

That was about the time Kevin asked me to become a manager.

More tomorrow...

Joey, Patrick, Jessica Alba, and your socks

Ace Developer Patrick Glaspey stopped me in the hall last week.

"You lied to us, HL."

"I'm sorry, Patrick. I thought you knew I was joking about Joey and Jessica Alba."

"What are you talking about?" He asked.

"What are you talking about?" I responded.

"The last time you spoke to the Regional Jobs Initiative Software Development Cluster, you collected historical data on Envision and EnvisionConnect's development history, and you presented that data in your speech."

"I did. How do you see that as a lie?"

"You told us that you would share the same data in your blog, but it never happened."

"Indeed, but it was not a lie. It was—uh...—a test. Yes! It was a test to see if anyone really reads my blog." Of course, Patrick didn't buy it, so I apologized for being senile and vowed to correct the problem ASAP.

Watch this space this week for some stats that will knock your socks off.

Death to Envision? Long live EnvisionConnect!

I am asked all of the time, "Why have you stopped maintaining Envision, when so many are still using the product?"

A few years ago, our users demanded a more intuitive interface, a product that was less error-prone and easier to maintain, and a tool that leveraged the Internet for inspections.

They demanded, and we responded, but as more and more customers began to move to EnvisionConnect, it quickly became apparent that maintaining two products was stretching our programmers too thin.

Sure, we were hiring new developers, but why assign them to work on an older version of our product—something that would be shelved once everyone has upgraded to EnvisionConnect.

Delivering high-quality product requires focus.

Completing EnvisionConnect was our goal. In order to reach it in a timely manner, our developers could not take their eyes off that goal.

Discontinuing maintenance of Envision allowed us to do that.

As a result of this strategy, over 90% of all Envision-based agencies can now schedule upgrades to EnvisionConnect.

It seems to me that our strategy is working.

Why so much ado about San Bernardino County?

By now, you've heard that San Bernardino County has migrated from Envision to EnvisionConnect. You've likely heard that they have completed billing and have over 40 inspectors in the field carrying Toughbooks running EnvisionConnect Remote, but maybe now you're asking yourself:

• Hasn't Nevada County used EnvisionConnect for nearly a year?
• Haven't many other Environmental Health regulators across the country adopted EnvisionConnect since then?
• Hasn't Anchorage, Alaska successfully migrated from Envision to EnvisionConnect?
• Hasn't Sacramento and Orange County been beta testers for a couple of years now?

Of course, the answer to those questions is a resounding "yes".

That brings us to your next question: What's so special about San Bernardino County?

EnvisionConnect is almost ready—Are you?

Extensive Early Adopter studies have been done on the percentage of people who immediately adopt new technologies and those who do not. Before August 1, 2007, EnvisionConnect will have enough base functionality in place to allow approximately 90% of Envision's customers to migrate.

The question is: how many will do so in 2007, and how many will procrastinate?