The main speed tip is to use Fast-forward or Run-to-Cursor-Line whenever possible. In the other modes, such as Go or Step-*, Serenji's support code on the server is being called after each command.
Run-to-line sets a one-time breakpoint on the line you point to, then runs 'fast' to that place, using the same 'fast' as in 'fast-forward mode'.
The difference between Fast-forward mode and Go mode is in the ability to interrupt execution. With Go mode you retain the ability to interrupt execution using 'Break'. Fast-forward mode allows your application to execute much faster between breakpoints, but you cannot interrupt execution. It will continue until a breakpoint is encountered or 'break-on-error' is triggered.
If you choose fast-forward but there are no breakpoints set and break-on-error is not selected, Serenji will revert to normal Go mode. This is a precaution to stop you getting into a state where you can no longer re-gain control. If you're certain you want fast-forward mode in this case, select fast-forward again.
When you start a routine from the Shell using 'Do', it uses Go mode. If you do not require this, you can speed up the execution by clicking on fast-forward.
When you are stepping through code, you can make Serenji run faster by closing the Variables window unless you really need it. This saves the overhead of fetching the symbol table at each step. The values of the variables can still be obtained by using the data tips in the Code window, or by writing them in the Immediate window.
Although Data Breakpoints are a powerful feature, they also incur a high cost. The expression you monitor is evaluated after each command is executed.
Last reviewed: Nov 15, 2004
Copyright 1997, 2005 George James Software.