Interesting ... I just tested it too, and verified it does mis identify when 3 is the number. Very puzzling.

I don't see the custlink groups any larger than 10 per at the most out of a table total of 1500.

I'm afraid not always.. that would be too easy!

Thanks.. this works great, and is way simpler than I was trying to do.

Have a great thanksgiving!!

Thanks for the help. When I run this, I end up with a table with the rank field populated with "NULL" values, but otherwise matching the original table....

I'm not using the product "MYSQL" ... I just happened to name my server that. I'm using MSSQL 2000. Sorry for the misleading server name.

This worked great... thanks for sharing your expertise, and have a great weekend!

Thanks.. that worked great.   Is it possible to display the % sign in the calculated result and limit the decimal places to less than the 6 shown  by default?

Thanks

Ahhh.. so the "convert" does not mean it will convert the native data... .got it...

Thanks.. works great!!!

Thanks.... already looked there. but I don't (cannot!) convert the native data in the source table.. I just need to display and output it in the manner outlined.

Thanks... worked fine.   I enjoyed studying it and learned a lot from it... nice work...........

Absolutely

Yes... the number must be unique even across branches....

Thanks!

This creates 25 numbers, but doesn't find duplicates, and when run multiple times adds the same numbers... I don't see any way to get it to add the next group...

Jeff.. your code works great.. thanks.    One problem with it (the result of inadequate explanation on my part) is that when creating these new groups of numbers I periodiocially change branches, but...