Tester in new build 604 doesn't work correct - page 3

 
cbalta:

hi, thanks for converter info. here M15 test of the same expert, in 604 and 509. i can provide the source code, it is a public developed one.

i did not expect this much difference. there are two possibilities:

a. data does not match.

Take the .hst files from the 509 build and import them in the 604 build while disconnected from the Broker, then set the Spread the same in both Terminals, set the EA settings the same in both Terminals. Even then it is possible to have a difference in performance but it should be minimised. To get the same result all variables MUST be the same . . .
 

********** I've got the bug ****************


Tester ticks are not complete / do not reach the low / high of the bar.

I wrote a Little EA to show the intern generated ticks

The flash-Crash in FDAX was on 2014.02.06 13:45 - 13:50 - bar.

18:12:49 2014.02.06 13:45 show_ticks Ger30Mar14,M5: Tick @ 2014.02.06 13:45:00: Bid= 9175
18:12:49 2014.02.06 13:45 show_ticks Ger30Mar14,M5: Tick @ 2014.02.06 13:45:02: Bid= 9157
18:12:49 2014.02.06 13:45 show_ticks Ger30Mar14,M5: Tick @ 2014.02.06 13:45:04: Bid= 9139.5
18:12:49 2014.02.06 13:45 show_ticks Ger30Mar14,M5: Tick @ 2014.02.06 13:45:12: Bid= 9135
18:12:50 2014.02.06 13:45 show_ticks Ger30Mar14,M5: Tick @ 2014.02.06 13:45:14: Bid= 9117.5
18:12:50 2014.02.06 13:45 show_ticks Ger30Mar14,M5: Tick @ 2014.02.06 13:45:16: Bid= 9099.5
18:12:50 2014.02.06 13:45 show_ticks Ger30Mar14,M5: Tick @ 2014.02.06 13:45:24: Bid= 9095
18:12:50 2014.02.06 13:45 show_ticks Ger30Mar14,M5: Tick @ 2014.02.06 13:45:26: Bid= 9073
18:12:50 2014.02.06 13:45 show_ticks Ger30Mar14,M5: Tick @ 2014.02.06 13:45:28: Bid= 9050.5 <--- this was the lowest generated tick
18:12:51 2014.02.06 13:48 show_ticks Ger30Mar14,M5: Tick @ 2014.02.06 13:48:13: Bid= 9174.5
18:12:51 2014.02.06 13:48 show_ticks Ger30Mar14,M5: Tick @ 2014.02.06 13:48:14: Bid= 9175.5
18:12:51 2014.02.06 13:48 show_ticks Ger30Mar14,M5: Tick @ 2014.02.06 13:48:15: Bid= 9176.5
18:12:51 2014.02.06 13:48 show_ticks Ger30Mar14,M5: Tick @ 2014.02.06 13:48:16: Bid= 9177.5
18:12:51 2014.02.06 13:48 show_ticks Ger30Mar14,M5: Tick @ 2014.02.06 13:48:25: Bid= 9178.5
18:12:52 2014.02.06 13:48 show_ticks Ger30Mar14,M5: Tick @ 2014.02.06 13:48:26: Bid= 9179.5
18:12:52 2014.02.06 13:48 show_ticks Ger30Mar14,M5: Tick @ 2014.02.06 13:48:27: Bid= 9180.5
18:12:52 2014.02.06 13:48 show_ticks Ger30Mar14,M5: Tick @ 2014.02.06 13:48:28: Bid= 9181.5
18:12:52 2014.02.06 13:49 show_ticks Ger30Mar14,M5: Tick @ 2014.02.06 13:49:28: Bid= 9182.5
18:12:52 2014.02.06 13:49 show_ticks Ger30Mar14,M5: Tick @ 2014.02.06 13:49:28: Bid= 9183
18:12:52 2014.02.06 13:49 show_ticks Ger30Mar14,M5: Tick @ 2014.02.06 13:49:28: Bid= 9183
18:12:53 2014.02.06 13:49 show_ticks Ger30Mar14,M5: Tick @ 2014.02.06 13:49:29: Bid= 9183.5
18:12:53 2014.02.06 13:49 show_ticks Ger30Mar14,M5: Tick @ 2014.02.06 13:49:29: Bid= 9183.5
18:12:53 2014.02.06 13:49 show_ticks Ger30Mar14,M5: Tick @ 2014.02.06 13:49:29: Bid= 9184
18:12:53 2014.02.06 13:49 show_ticks Ger30Mar14,M5: Tick @ 2014.02.06 13:49:29: Bid= 9184
18:12:53 2014.02.06 13:49 show_ticks Ger30Mar14,M5: Tick @ 2014.02.06 13:49:36: Bid= 9184.5
18:12:54 2014.02.06 13:49 show_ticks Ger30Mar14,M5: Tick @ 2014.02.06 13:49:36: Bid= 9184.5
18:12:54 2014.02.06 13:49 show_ticks Ger30Mar14,M5: Tick @ 2014.02.06 13:49:36: Bid= 9185
18:12:54 2014.02.06 13:49 show_ticks Ger30Mar14,M5: Tick @ 2014.02.06 13:49:37: Bid= 9185
18:12:54 2014.02.06 13:49 show_ticks Ger30Mar14,M5: Tick @ 2014.02.06 13:49:37: Bid= 9185.5
18:12:54 2014.02.06 13:49 show_ticks Ger30Mar14,M5: Tick @ 2014.02.06 13:49:37: Bid= 9186
18:12:55 2014.02.06 13:49 show_ticks Ger30Mar14,M5: Tick @ 2014.02.06 13:49:38: Bid= 9186.5
18:12:55 2014.02.06 13:49 show_ticks Ger30Mar14,M5: Tick @ 2014.02.06 13:49:38: Bid= 9186.5
18:12:55 2014.02.06 13:49 show_ticks Ger30Mar14,M5: Tick @ 2014.02.06 13:49:38: Bid= 9187
18:12:55 2014.02.06 13:49 show_ticks Ger30Mar14,M5: Tick @ 2014.02.06 13:49:38: Bid= 9187
18:12:55 2014.02.06 13:49 show_ticks Ger30Mar14,M5: Tick @ 2014.02.06 13:49:39: Bid= 9187.5
18:12:55 2014.02.06 13:49 show_ticks Ger30Mar14,M5: Tick @ 2014.02.06 13:49:39: Bid= 9187.5
18:12:56 2014.02.06 13:49 show_ticks Ger30Mar14,M5: Tick @ 2014.02.06 13:49:39: Bid= 9188
18:12:56 2014.02.06 13:49 show_ticks Ger30Mar14,M5: Tick @ 2014.02.06 13:49:39: Bid= 9188
18:12:56 2014.02.06 13:49 show_ticks Ger30Mar14,M5: Tick @ 2014.02.06 13:49:56: Bid= 9182.5


The Chart Shows These screenshots: the low of the bar (both M1/M5) is @ 9028.5 but the tester didn't reach the low. Thats why the stop didn't release in my first example!


M5


and in M1:

M1

 

test_604_hstminport509

done. thanks for tips. no change.

compiler know if variables are same or not. possibly, we need to know details of new compiler data types, compared to old mt4..

although no warnings given, they work different..

i need to test some other EA's, to compare with this test.

 
kla-mue:

********** I've got the bug ****************


Tester ticks are not complete / do not reach the low / high of the bar.

I wrote a Little EA to show the intern generated ticks

Do you have "mismatched chart errors" ?
 

509_2


604_2

another 604 and 509 test. possibly some precision problems with float-double numbers.. or with data-time variables.. without warning..

 

the strange thing is, both 509 built ex4 and 604 migrated/compiled code (w.o warning) shows the same poor behaviour under 604, for some EA's..

i need to know some more details bw. old and new compiler, to be able to catch the problems..

 
kla-mue:

********** I've got the bug ****************


Tester ticks are not complete / do not reach the low / high of the bar.

I wrote a Little EA to show the intern generated ticks

Do you have "mismatched chart errors" ?

====> NO, I took great care of a correct history (its a very young history, so its available from the broker)

Klaus

 
RaptorUK:
Take the .hst files from the 509 build and import them in the 604 build while disconnected from the Broker, then set the Spread the same in both Terminals, set the EA settings the same in both Terminals. Even then it is possible to have a difference in performance but it should be minimised. To get the same result all variables MUST be the same . . .


Hi, I have a problem with build 604, I can not download indicators or templates. Can anyone help Thanks
 
topcat275:

Hi, I have a problem with build 604, I can not download indicators or templates. Can anyone help Thanks
What does this have to do with "Tester in new build 604 doesn't work correct" ?
 
ubzen:
Sorry. This one is going to backfire on advanced users big-time. What I've realized is that when some software company tries to make things simple to become novice friendly. They just end up taking away Options. I have no interest in taking away my data-import just because importing data carries too many nuances.

Agree, but note that when I mean improve, I mean give automatic options. I'm not talking about lose control (as we lost in MQL5).