E9.2 Level Break Headers Don't Fire Before First Do

[eydeak]

Hi,

My basic question is, Should all level break headers on a section always fire before the first main section Do Event fires?

I've been working on a custom UBE for about a week. The section has 3 level break header sections with code in their do events. Then I also have code in the Do Event for the main section. Up until sometime yesterday, all 3 level break do events would fire before the first main section do event. Then late yesterday only the first level break header code would fire before the main Do Event. I've verified via debug that it is hitting the code for the first level break then doesn't hit the code for the next 2 level breaks.

I can't fathom the ER changes I've made would suddenly make the level breaking stop working. Any ideas would be helpful.

Thanks,
Ellen

 

[eydeak]

And, of course, once I break down and post the question, I finally figured out what stupid developer trick I performed to cause the level breaks to stop working.
I had been using debug logs to look at the SQL statement generated and DB2 query analyzer to help optimize performance. The 2nd level break header field was an alias that was returned in both tables in the BSVW. When I originally added the LBH I picked the field from the second table instead of the first. To optimize index usage I needed to switch to using the column from the first table. I changed the sequence on the section and didn't stop to think that it would break the connection to the LBH on the field.

I actually used PatWel's Object Browser to finally see it. When looking at the properties on the LBH section in RDA, it showed that it was connected to column I changed the sequencing to use. But when I looked at the same section properties in Object Browser it showed that the LBH was on the field in the other table. Once I saw that in Object Browser, I remembered changing the sequence and realized the error of my ways.

 

[craig_welton]

And, of course, once I break down and post the question, I finally figured out what stupid developer trick I performed to cause the level breaks to stop working.
I had been using debug logs to look at the SQL statement generated and DB2 query analyzer to help optimize performance. The 2nd level break header field was an alias that was returned in both tables in the BSVW. When I originally added the LBH I picked the field from the second table instead of the first. To optimize index usage I needed to switch to using the column from the first table. I changed the sequence on the section and didn't stop to think that it would break the connection to the LBH on the field.

I actually used PatWel's Object Browser to finally see it. When looking at the properties on the LBH section in RDA, it showed that it was connected to column I changed the sequencing to use. But when I looked at the same section properties in Object Browser it showed that the LBH was on the field in the other table. Once I saw that in Object Browser, I remembered changing the sequence and realized the error of my ways.

Ellen,

You have always been a strong advocate for JDE Object Browser. Thanks for the shout out!

Craig

 

[DaveWagoner]

Could not do my job at anywhere near the efficiency I get without ObjectBrowser. Thanks as always!