At a lack of Official MQL4 Profiler, MicroSecond count good practice to detect efficiency issues?

New comment
 

Profiler does not work on backtester (live charts profiler not good enough) and I need high quality code; should I be looking at other software? 

MQL4 Solutions only; 

MicroSecond count good practice to detect efficiency issues?
Testing trading strategies on real ticks
Testing trading strategies on real ticks
  • www.mql5.com
The article provides the results of testing a simple trading strategy in three modes: " 1 minute OHLC " using only Open, High, Low and Close prices of minute bars; detailed modeling in " Every tick " mode, as well as the most accurate " Every tick based on real ticks " mode applying actual historical data. Comparing the results allows us to...
 
nadiawicket: (live charts profiler not good enough) and I need high quality code; MicroSecond count good practice to detect efficiency issues?

  1. You are running with a network delay of 50-500 ms and processing [1 … few] tick(s) per second. Why isn't the profiler good enough?

  2. Speed has nothing to do with quality (high or otherwise.)

  3. What does microseconds have to do with milliseconds? You are worrying about noise.

    The microseconds the EA takes is irrelevant, compared to the 20-200 milliseconds the network takes, which is irrelevant compared to the minutes to do a trade because of the servers during news.

    OrderSend sends the request, network delivers, placed in the server queue. Once it reaches the top of the queue, you get filled or not, and the result delivered back to the terminal. Normally it is the 20-200 milliseconds the network takes to deliver is the limiting factor, but it can take minutes to do a trade because of the servers during news.

  4. Do not look at efficiency, until you actually have a problem.

 
William Roeder:

  1. You are running with a network delay of 50-500 ms and processing [1 … few] tick(s) per second. Why isn't the profiler good enough?

  2. Speed has nothing to do with quality (high or otherwise.)

  3. What does microseconds have to do with milliseconds? You are worrying about noise.

    The microseconds the EA takes is irrelevant, compared to the 20-200 milliseconds the network takes, which is irrelevant compared to the minutes to do a trade because of the servers during news.

    OrderSend sends the request, network delivers, placed in the server queue. Once it reaches the top of the queue, you get filled or not, and the result delivered back to the terminal. Normally it is the 20-200 milliseconds the network takes to deliver is the limiting factor, but it can take minutes to do a trade because of the servers during news.

  4. Do not look at efficiency, until you actually have a problem.

1. Profiler not for tester

2. Speed can have to do with quality

3. Pure philosophy if is achievable; why settle for lower level code if other is better or equal in tester? If just feels better, can be fun and enjoyable to go for quality. You claim its irrelevant bc its always going to delay at server?? I was not aware!! Great to know; if everything under 20 milliseconds just going to wait at the server; why worry? Is this the case? You are certain? Id still go for the achievable best. All these new exchanges so wonder if the rules will always be like this

4. Again thats very nice; I dont really have an issue however its a finished real product that I would like to be like a work of art

New comment