On 08/07/2015 06:51 AM, Zoltán Balogh wrote:
So with Zeller Benjamin we have created a version of the mapplauncher
[4] what is doing the right thing. The next step is to profile and
document the performance impact of this solution on the application
startup times.
Next week you will hear more about that. If the results are promising
then we can discuss about how to proceed. If not :) then we can say that
we have properly investigated this possibility.
I would expect the results to be roughly the same that you already
provided. The introduction of the exec() call will likely slow down the
booster a bit, but it shouldn't have an effect on the application
startup time (because generally the booster is started ahead of time).