D
dbeavon
Guest
@James I had seen that PCT started being installed as part of OE. But Progress didn't seem to be ready to offer direct support. I would be happy talk to Riverside about the performance problems in the context of schema operations, but I suspect they would redirect me back to Progress. I think I'll have more luck once there is a single point of contact that takes responsibility for the whole process (the build performance, along with the related schema operations) @danielb I think you are making the same case I am. If we want our automated builds to use inputs that consist of both code *and* schema from source control, then we will often be creating fresh OE databases from the related schema. Optimizing the creation of fresh OE databases is key to having good build performance. My point earlier was that if I compile against local schema, there is a price to pay; but if I compile 25,000 programs against remote schema there is an even higher price to pay. Ideally I'll find a way to optimize the creation of the local schema. Perhaps we'll need to start using our source-control tools (or nuget package control?) to check-in the schema along with a full zip of the corresponding OE database. Then the only CPU is cost at the time of build is the unzipping of the OE database. Its a bit ugly, but our build operations would be faster.
Continue reading...
Continue reading...