Metatrader known bugs ... - page 19

 

With a new build 625 of mt, if you try to compile EAs that are no explicitly testing the results of Order... functions, compiler will issue a warning. Freely leave it as is : the EA will continue working as before

 

So we can ignore the : "return value of 'OrderSelect' (or similar order functions) should be checked" warnings?

 
checkin:
So we can ignore the : "return value of 'OrderSelect' (or similar order functions) should be checked" warnings?

Yes

From tests I have done, no change in the way how orders are handled is you simply ignore that "warning"

 

One thing to be careful with : metaeditor allows that the same file gets loaded more than once in some cases. If you make changes in on of the copies and you compile, you do not know what version is saved and compiled. That is a very nasty bug from new metatrader 4 since this way you actually do not know what are you working at

 

debuger always works in strict mode - useless if you do not want to make your code strict

 

Stupid error from new metatrader compiler : for this line

Draw(t3, i, b << 16 + r);

It warns "check operator precedence for possible error; use parentheses to clarify precedence". "<<" is not a logical operation, it is a shift left operation

 
tampa:
Stupid error from new metatrader compiler : for this line

Draw(t3, i, b << 16 + r);

It warns "check operator precedence for possible error; use parentheses to clarify precedence". "<<" is not a logical operation, it is a shift left operation

:):)

If you enter it as

Draw(t3, i, b < 16 + r);

which is nonsense in this case (since it is not at all what it should do), it does not complain

Well spotted - the obviously forgot binary operations - and judging from the way they do, it will stay like that for a long time

 
Ovo:
... The last change by which I was affected most, was the empty structure padding. I. e. empty structure adds 1 byte of garbage since the editor 910...

Good news (for me). They removed the structure padding in Metaeditor 915 (packed with terminal 628). Still the empty structures have a single byte padding (which is no problem), but all inherited structures got it removed since now. I can move from 904 to 915.

 

hi mladen,

i've found another problem with TVI_v3 mtf nmc_strict in the tester of the 625 build: setting PopupAlerts=true or using different information predefined functions, like, Print(...) or Comment(...) nothing happens on the screen. All these functions work only in real trading or after starting a debugger (but in this case one cannot test on historical data). With the 610 build everything was correct.

Moreover, applying these functions to another indicators i got the same problem.

Do you also have the same thing while backtesting or one should do something additional to see values using these information functions? Unfortunately, i didn't find information about this sort of the problem.

 
qwex:
hi mladen,

i've found another problem with TVI_v3 mtf nmc_strict in the tester of the 625 build: setting PopupAlerts=true or using different information predefined functions, like, Print(...) or Comment(...) nothing happens on the screen. All these functions work only in real trading or after starting a debugger (but in this case one cannot test on historical data). With the 610 build everything was correct.

Moreover, applying these functions to another indicators i got the same problem.

Do you also have the same thing while backtesting or one should do something additional to see values using these information functions? Unfortunately, i didn't find information about this sort of the problem.

Sorry, that has nothing to do with the code of that indicator (all those are build in procedures and we do not have any control of how they work internaly)

If it works in normal mode and debug mode and does not work in a back test mode, then obviously, something is different in back test mode

_____________________

PS moveed these posts here since they belong here and not to multi time frame thread

Reason: