reloadEFS() ???

[mkm]

Has anybody out there written a succesful script with
the reloadEFS() command? What ever I try, it crashes, I use the newest version!

Peter

[gspiker]

Yes, I have. As I stated in another thread that mentioned this command, it is not one to use lighly. I have never had a crash due to it, but one can imagine that it would be possible in a efs that wasn't taking reload's into account (esp if using arrays, or ). I have (while debugging an efs) locked up eSignal using this command to the point where I had to kill eSignal and restart.

In theory this command wasn't to be let into the publics eye because of the potential problems that could crop up if one wasn't careful. Now that it is out in the public, I will reiterate my warnings - don't use it unless you are sure there is no other solution to your problem, and the, only if you are a good at progamming.

Garth

[eSignal_JasonK]

Hello Peter,

Please post your formula here and tell us more about what you are trying to accomplish and why you nee to use reloadEFS().

Thanks,

[mkm]

Answer to Jason:
I need this, because it seems that it is the only one to solve the
"moving" problem

My problem is already listed in this thread

http://forum.esignalcentral.com/show...&threadid=1204

I found no other solution then reloading the efs

Answer to Garth:

Unfortunately I´m not the professionell in programming, I
´ve learned a lot but at the moment my programming is a combinations of some examples to reach my targets.
So I try to grow up my knowledge but it takes time I understand
more and more, but for ex. to divide what is a Array and understanding why this would work or not is quite difficult for me, so if you try explain it very very easily it would be great

To both:
Anyway I´m well convinced in this quite easy strategy! And hope that I will find a solution.

As attachement my formula "Automatic.efs" used on a
Renko Chart 0.5 BoxSize Source HL/2 for ES-Mini
Look at my post above!

[gspiker]

Does this run correctly is setComputeOnClose(true) is set in preMain()?


I think I see what may be the problem, but if the above is set, it should work.

[mkm]

You mean the "reloadEFS()" works when I include "setComputeOnClose(true)" ?

or the "moving" problem will be solved ?

Peter

[gspiker]

I mean that the moving problem might be solved with that command placed in premain...

[mkm]

Unfortunately not! It seems that the changes are not so fast but already here.

Look at the two pictures!

Backt to the reloadEFS(). What should I change in my formula that the command work´s ? Seems still the only solution

[mkm]

This on Monday Morning

[gspiker]

I still don't think reloadEFS() is the only (or even the correct) solution for this.

Can you describe to me what I am seeing in these two pictures?
Also, was this left up all weekend, or did you shutdow on friday and start up again on Monday?

Garth

[mkm]

I agree that the reload command would not be the best solution,
may be we find a better one.

I shutdown on Friday and start again on Monday(res. Sunday Night), the left is from the some bars after market starts in the early morning.

I realized today that if I include the "computeonclose" in preamain the left is only if more then one or two bars appear in the same moment, if only one appear it is ok, but if more it is again false.

You see in the first picture the blue line is my "automatic" and the red one is the normal SMA5,3, the normal SMA left one bar wrong, the blue bar is on this picture correct.

On the second picture above it is the situation on monday, the blue line has "moved" after some bars appear, the arrow show
a easy to compare situation (the green and red in many red bars)

The red SMA has no function, only to compare


Peter

[gspiker]

Let me look at this in more detail this evening, and I will get back to you on what I find.

G

[mkm]

Hi Garth,

have you found something?

If not, I will try the reloadEFS() command as solution, so back to my question how to change my formula that it works,

yesterday we had again a good example how my system works and bring money :-)

[gspiker]

Hi,

I stripped out all of the dll code, since I don't have those modules and it shouldn't affect the results displayed, and I also had to modify the times that the charting would work.

After that I ran the code in replay mode...but there seems to be a bug with replay mode and renko charts, everything worked great when I used candlesticks, but with renko the MA lines wouldn't line up.

It's not clear at this point if the replay mode bug could be related to why the EFS has a problem running in RT. But right now I have two charts up, one with renko and automatic and one with candlestick and automatic. If the problem exists only in the renko chart we have a clue.

There is also some places in the code where I see potential problems, but wanted to find out the answer to this question first before I started hacking at the code.

G

[gspiker]

Well, after running it today I noticed that the renko chart drifted rather a lot, while the candle didn't.

I want to look at this further, but it sure looks like something odd is happening with renko charts and this indicator.

I'm attaching a modifed version of the EFS. Of course it has the changes I mentioned before (commented out dots .dll references, changed the times when the study will execute), but I also made it start up much faster and use less resources as well as made some changes (as of yet untested in RT) that might help it produce less drift.


I will look at this more on Monday.

Garth