Create Mobile Map Package: Related Records

[MarkCederholm]

I'm quite impressed by the Create Mobile Map Package tool: the data model appears perfectly preserved, including the geometric network and relationship classes.  One thing I noticed, however:  when I limit the export to a display extent, all records are exported from related tables instead of just related records.  Are there any plans to add an "Export Related Records" option?

Also, are there plans to support annotation?  Our Gas division doesn't use it, but the Electric division uses it heavily.

[JayJohnson2]

How quickly does the mobile map package tool work for you?  On a state or regional level, it has taken five hours here.

[MarkCederholm]

For us it only takes a couple of minutes for a 450 MB mmpk*.  However, I've run into another bug where certain relationships cause the tool to try to create duplicate tables, which crashes it.  I have opened a case with tech support, but it's currently on hold until Pro 1.4 comes out, to see if it's fixed there.

It's kind of a moot point anyway right now, at least for Runtime users, because the first release of .NET v.100 didn't solve any of the mmpk rendering bugs that I'd reported for the beta.  We'll see what the next release is like.

*Created from a file GDB that's exported from ArcSDE on a nightly basis.

[DavidSpeight]

Did you ever get any news on the bug you described?  I am still having that same issue at Pro 2.0.

[MarkCederholm]

Yes, the problem I encountered is still there at 2.0, but I've been letting the issue slide for a while.  Since we're planning to migrate our gas data model to UPDM, I wanted to get some sample data migrated before I continued testing.

[DavidSpeight]

I did some additional MMPK troubleshooting last week with the help of an Esri support rep from the Navigator team.  I had stripped out all relationships classes in my source fgdb as well as the geometric network.  We isolated a specific polygon feature class as the problem layer and saw that it had some suspect geometries.....polygon slivers with an area less than a square foot.  I deleted the bad features out of the dataset and the MMPK built with no errors.  I've been able to consistently rebuild it with no errors ever since.  Now I'm gradually adding back relationship classes and rebuilding to see if I bump into the error again.

The odd thing was that I would get the "path already exists" error when the problem feature class was in a map with other layers, but would get a generic 999999 "function failed to execute" error when it was in a map by itself.