The ANSI SQL issue was debated amongst the team. We knew it would "break" some legacy code but felt it was important to do it right from here on out. So we put in the SYS function for those that didn't want to be ANSI compliant :-)

And there are some performance gains in VFP 8 as well. There's a huge gain in creating and releasing objects under some circumstances. Calvin showed this at the last DevCon. His changes to the source in regards to managing objects have as dramatic impact on objects as Rushmore had on data. I kept looking for a catchy name for it (lol).


>>Glad to hear it went well! I know that SQL stuff was a pain but I guess it's all resolved now. Have you noticed any differences in operation?
>
>Yes, at first I wasn't aware of the ANSI SQL issue. Then, I had a choice to make. I decided to make it compliant for future expansion and to remain as up to date as possible with the VFP default.
>
>Yes, I've seen some increase of performance at several places. Some of them have been related to the fact that I now use an approach to get the primary keys first and then do another SQL to get the other fields. But, for those SQL that remained untouched, I have seen some increase of performance at several locations.