Hello John,
I appreciate your attention to this issue.
Real life scenario 1:
A developer makes a code change in development and validates that it is working.
When Approved, the developer makes a targeted deployment to production. To do this they copy their new code into the code already in production. Unfortunately, they added their new code 20 lines above or below where they had it in development. Everything works as expected. The developer thinks everything is fine. Technically it is fine.
6 months later, another change is made to the same package/procedure/function, except this time a scheduled deployment is done. During the deployment the functions are compared, but because the code blocks are in a different sequence, time is spent sorting through the differences.
The real issue is the ability to do a targeted deployment, currently unavoidable.
Attached are 2 functions demonstrating the issue.
Actually, Not bad if the code block was only 12 lines long.
Real life scenario 2:
Code to correct replication issues between 4 databases needs updated in all 4 databases.
This is a targeted fix every time in each database needed to synchronize the data.
The Procedures (6 in each database) range from 6000 to 9000+ lines each. They all deal with inserting updating or deleting records. Although the table names are different so each section should be separated, all of the tables contain one of the following 3 key fields: Person_NBR, ORG_NBR or ISN. This used to causes the match to split the comparison. When I tested it yesterday, the new matching algorithm seems to be doing a much better comparison than in the previous versions I used to use.
Real life scenario 3:
I have not tested this with the new matching algorithm.
I have a package with several procedures in it and I need to do a special 1 time 1off process.
So I copy a couple of procedures, in the same package. I give the new procedures a new name and make a couple of small changes in the code.
Then when I compare the package, it matches my new 1off procedure to the old original procedure.
Previously I would then link the two old original procedures together and verify that I didn’t accidentally update the original procedure.
Thank you for your assistance.
Sincerely,
Dan Peterson
Database_2_Test_2_Function.txt (1.84 KB)