Wednesday, September 29, 2010

The Hazard of Change

To borrow a term I hear a lot in the Smart Grid world, it seems those of us on the BACnet Committee aren't doing enough to "socialize" changes, mostly additions, to the standard. This became clear today with an exchange today on the BACNET-L email list.

John Arwine (just a short ways down the road from here) noted that he was working witha new BACnet controller, one of the first BACnet devices from its manufacturer. The documentation for the device identified it as having Analog Variables, Digital Variables, and Integer Variables. He hadn't heard of the last and asked the list whether this was a defined BACnet object type.

I responded:

Yes. If you look at Addendum 135-2008w (http://www.ashrae.org/technology/page/20), you will find:
12.G Integer Value Object Type
12.H Positive Integer Value Object Type

This was published early this year.

And it was. It didn't take me long to dig out a copy of the addendum because I knew what I was looking for, even though I no longer keep all the BACnet addenda letters and their contents mentally correlated.

With 33 addenda, Addendum a currently through through ag, and many of them coompilations of miscellaneous unrelated changes, it gets difficult -- especially when I'm putting in a LOT more time these days on work on the Smart Grid than on BACnet.

But at least I knew where to turn and what to look for. The description of the contents of Addendum w is unusually terse, both in the addendum and on the web page, saying only "135-2008w-1. Add more primitive value objects, p. 2." This could be easy to miss if one were looking for "Integer Variables."

So we're making these additions, but not getting the information out to our users as well as we should. This is not a good thing.

---

Committee chair Dave Robin had a good response too (sparing me the necessityof a followup, thank you Dave!), noting that the terminology being used by the manufacturer was nonstandard. For many BACnet oldtimers the use of "unofficial" terminology provides a warning that something might be misinterpreted.

Quite possible in this case where a "Integer Variable" might be a new Integer Value or Positive Integer Value object, or it could be a Multi-state Value object.

So now John has a new task, to find out which of these object is intended -- if it's not a proprietary object!

No comments:

Post a Comment