Steve’s Software Trek

The BACnet committee has a dining tradition: when we go out to dinner, we get one bill and divide it equally among those present. The restuarant must serve beer, and have good food. The restuarant or brewery also needs to have some vegetarian fare since Carl is a vegetarian. If we have a meeting that is during the weekdays, the group dinner is held on a Tuesday. If we have a meeting that is during the weekend, the group dinner is held on Saturday.

I heard the legend of the “J” stamp a couple of weeks ago while at dinner during the BACnet meetings held in Atlanta, Georgia. Apparently, when the BACnet standard was first published for public review, many comments flowed into the committee from some unfriendly commenters. Some comments were either impossible or nit-picky. As the legend goes, a rubber stamp with the letter “J” was purchased and was used on some of these comments to notate “Jerk.”

During the last transition of committee chair in Nashville, Tennessee, Steve Bushby handed a wooden gavel to Bill Swan. Bill now uses the gavel to signify the beginning or end of the meeting, or a call to order. It makes quite a few of us jump!

Sometimes during a meeting something is said that requires a definition. I keep a notebook of some of these words, and what they mean in the context of a BACnet meeting. Here are some examples:

• “lame device” or “weak device” – a BACnet device implements only the minimal required functionality.
• “data sucking” – learning the system using Who-Is and reading the object list.
• “raising the bar” – changing the benchmark. Usually means that more than minimal functionality is now required.
• “tickling” – a method to get a client device to do something.
• “barf” – what a device does when it messes up. Similar to BSOD.
• MIRV – multiple independent reentry vehicle. Used to describe the Mulitplexer object.
• “lurking” – what an MS/TP device does before joining in on the token passing.

Sometimes during a meeting something memorable gets said. Here are some examples:

• “The first question to ask in response is: ‘Well, who the hell are you?’” – David Robin
• “Choice is the enemy of interoperability.” – John Hartman
• “Size Matters.” – David Ritter

Is there a BACnet legend, tradition, or phrase that I missed?

At work, I have spent the past month, and in particular, the past two weeks, preparing for the BACnet Interoperabilty Workshop. What does that mean? Well, we were going to take two beta products to the plugfest, as it is affectionately called. My original plan was to take our existing product, convert and upgrade the BACnet stack in it, and add routing capability. It probably requires about a month of work to do that. However, many other things came up at work, and I didn’t even come close getting a good start on the project. Two weeks ago I talked to our specification manager who is going to be at the plugfest with me, and she asked for a completely different product to have BACnet capability added.

So two weeks ago on a Monday afternoon, I pulled the product source code from our version control system (I wish it was Subversion). I setup the MPLAB IDE for that project, and connected my ICD2 to the PIC18F6720 based lighting controller. The product was developed by my friend and colleague Mark. We use a software coding standard at work that I created. Coding standards make it easy to work on each others code and it still looks somewhat familiar.

On Tuesday I grabbed the latest version of the BACnet Stack at Sourceforge from subversion, and set about integrating BACnet into the product. I cut out some code to make space for the BACnet stack – just the self test code and the RS-232 terminal. By Wednesday I had MS/TP communications and a Device Object responding to WhoIs service requests with I-Am service responses.

On Wednesday and Thursday I added support for the rest of the objects (8 Binary Inputs, 8 Binary Values, and 1 Analog Input) and DeviceCommunicationControl, ReadProperty, TimeSync, ReinitializeDevice, WhoIs, and I-Am services.

On Friday I fine tuned the MS/TP state machine. I started to use the Bitscope attached to my PC, but it didn’t have the deep capture and user friendly features of my favorite scope, the HP54645D 100MHz 2+16 channel mixed signal oscilloscope. The HP54645D is great for protocol analysis. For example, I capture 5 seconds of communication. Then I use the MegaZoom feature.  I simply twist the timebase knob and zoom in on the trace.  I use the horizonal scroll to move the trace in time to the place of interest. I can then use the cursors to measure the time or frequency or voltage. This works great, for example, when checking to see how long it takes from receiving an MS/TP token to actually using the token.  It also works well to spot a collision and determine which packet caused it.

This week, on Monday, I added the WriteProperty service, and began validating that all the services were integrated properly. I added hooks to write to various BACnet properties since I didn’t take the time to modify the keypad menus for setting things like the BACnet Device Object Instance Number and the MS/TP baud rate.

So, it took about five working days to go from zero to BACnet. Wow! I still have some work to do in order to design the networking interaction of this device with other devices of its kind. In other words, I have to figure out how to make it do simple control networking out of the box. I also have to add menus to configure the BACnet parameters. However, I am able to read and write to the 8 relays. I am able to read the values of the inputs. I am able to read the time and date, and set the time and date. I am able to control the device communications, and able to reset the device. All using the BACnet protocol.

I guess we will see how robust the unit is based on how it behaves at the plugfest. Stay tuned!

In April, 2004, the BACnet committee toured a NIST facility. Part of the tour covered the Building and Fire Research Laboratory, and we glimpsed some of the twisted metal beams from the World Trade Center after it collapsed on September 11, 2001, in New York. It was a very solemn moment for all of us.

The BACnet committee and working groups met in Quebec City for the summer meeting. The meetings were mostly productive, especially for my Lighting Applications working group, which didn’t meet. However, our proposal for the Lighting Output object received discussion time during the SSPC-135 meeting on Saturday, Dave Fisher (of Polarsoft fame) and I edited it on Sunday afternoon, and it was discussed again on Monday during the SSPC-135 meeting. It will most likely be voted out in an addendum for public review at our November meeting.

Quebec City was delightful! I had a little time to walk into Old Quebec, and also toured The Battlefields Park. I checked out Wikipedia to see if there were any articles on The Battlefields Park, and there were none. So I created one, and took pictures of it for the article. I also walked around Old Quebec with Dave and Coleman, two driving forces at Polarsoft, and did a little shopping as well. Dave and Coleman are delightful and intelligent and generally fun to hang out with. We often met other members of the BACnet committee while out and about in Old Quebec.

We were also in Quebec for the celebration on the eve of Saint Jean Baptiste Day. I walked up onto the Plains of Abraham in The Battlefields Park with Steve Bushby and Mike Newman, and we watched some of the free concert, waiting patiently for the fireworks. One of the Canadians on our committee, Roland Laird, stopped by and listened to the concert with us. We saw the bonfire, but decided to go back to the hotel shortly after 11pm. Apparently the celebration and concert lasted until 6am! I saw fireworks from my hotel room window around 12:30 am and also 1:30am (or was it 2:30am?). I added some photos of the event to the Wikipedia entry for Saint Jean Baptiste Day.

I participate in an international standard called BACnet. We exchange ideas and proposals via email and teleconference, and meet face to face at least 4 times a year in various parts of the U.S. and Canada. We handle a lot of the details by utilizing working groups. The working groups are usually formed to attack a certain problem area, and generate proposals to change the BACnet standard. The BACnet committee (AKA ASHRAE SSPC-135) then reviews the proposals, and if they are okay, vote them out for publication. At that point the proposals are open game for the world to assault, and sometimes they get public review comments. All the comments must be addressed before the proposal becomes a standard.

I lead a working group call the BACnet Lighting Applications working group. We have created a number of proposals over the past 4 years, and they are just now getting formal review by the BACnet committee. It usually takes several years for a proposal to become a standard. I hope I can be patient!

In my spare time, I work on a BACnet Protocol Stack for embedded systems. I figure that since I know a little about decoding the ASN.1 in the standard that I should put it to good public use. I also am committed to promoting BACnet, and think that having a free protocol stack allows anyone to give it a try.

I wrote the initial entry for BACnet on WikiPedia based on some history shared with me by the BACnet elders.