Join Action Block returning strange characters

[Former Member]

Forum,

I am using the Join action block in 11.5.3 b66 which attemps to join 3 columns in two results sets (DateTime, TagName and Value1 with DateTime, TagName, Value2). The join will break out any differences between the two sets when using the Full Outer Join.

When I run this trans the join returns all matches and differences, but for 14 of the 240 tags I get the following set of characters � in either the value1 or Value2 loction. From Google I see that they are Unicode quoting characters.

When I view the two sources before the join these � are not present.

Example:

DateTime TagName Value1 Source1 Value2 Source2

10/13/2008 0:00 FI-3715 3060.570068 VAX � ---

10/13/2008 0:00 FI-3715 � --- 3060.570068 VMS

As you can tell the values are correct and match fine. There are some differences within the 14 tags that have this problem but they are displayed the same way.

Any ideas out there?

Roger DeWeese

[Former Member]

Hi Mike,

The strange characters are Excel's interpretation of the xml "NAN". I was attempting to do a side-by-side comparison in Excel to ensure the logic was correct and these were throwing me a little.

Everything appears to be working as designed.

Thanks for the response.

Roger

[agentry_src]

From discussions with some of my coworkers, those symbols come up once in a while in various situations. With your input in mind, I wonder if most of them occur in MS controlled software.

Thanks,

Mike

[agentry_src]

Roger,

It looks like you have the Tag Query set to display 6 digits.

Can you look at each of the data sources directly and see the values that are being rounded to the six digits? In other words, can you check the raw data from MII to see if the values after the 6 decimal places are the same? You can display the tag query results as xml to view the entire number (not limited to the displayed value).

From your posting, I would guess that the join is based upon all three columns and I suspect that your two datasources are giving slightly different results in the decimal place noise. Just a guess, however. Let me know what you see.

Thanks,

Mike