G
gus bjorklund
Guest
> On Oct 28, 2019, at 3:36 PM, dbeavon wrote: > > I might be going down an unnecessary rabbit hole here you are. when you run your application, it has the 4GL overhead as part of it. you cannot eliminate that. furthermore, the overhead added to your app by the 4GL is not just simple interpretation of r-code. it also has a lot of functionality in addition to the interpretation. but i digress. while (highly) interesting, not useful to app developers. when you measure the application performance, you should consider it in toto. the profiler tells you where the time goes with pretty good accuracy (though there are some things that are confusing and could be improved). it takes time to understand the implications of the data you get and what you can do to make improvements. the most effective improvements often are changing the algorithms used by your application’s code rather than micro-optimizations at the statement level (though those can be useful at times). to put it another way, many performance problems are caused by sh*t code. a long time back, i wrote a test harness that to measure the 4GL overhead. its purpose was to test performance improvements made to the 4GL runtime and to compare releases. i cannot share that with you since psc owns it.
Continue reading...
Continue reading...