Errors, bugs, questions - page 2901

[Aleksey Vyazmikin]

Why would a code like this

enum ENUM_type_NAME
{
   a,
   b,
   b
};
input ENUM_type_NAME type=ENUM_type_NAME(0);

the styler turns into such:

enum ENUM_type_NAME
{
   a,
   b,
   b
};
//+------------------------------------------------------------------+
//|                                                                  |
//+------------------------------------------------------------------+
input ENUM_type_NAME type=ENUM_type_NAME(0);
//+------------------------------------------------------------------+

How can I teach the styler not to draw caps with this kind of enumeration?

[Igor Makanu]

Aleksey Vyazmikin:

How can we teach the styler not to draw caps for such enumerations?

it's also annoying to stretch down the source code because of inserting a comment block

I remove the top 2 lines of this styler comment and leave only this:

enum ENUM_type_NAME
{
   a,
   b,
   c
};
//+------------------------------------------------------------------+
input ENUM_type_NAME type = ENUM_type_NAME(0);

This way the styler will no longer add the"Function Header" comment in this place in the pi subsequent application of the styler

[Aleksey Vyazmikin]

Igor Makanu:

also straining to stretch down the source code due to the insertion of a comment block

I remove the top 2 lines of this styler comment and leave only this:

That way the styler will no longer add the"Function Header" comment.

I just do not understand where he saw the function there - if you put in front of

enum ENUM_type_NAME

I just don't understand where he saw the function there - if he put it in front of "function", it would be ok, but even your suggestion strains me.

[Valeriy Yastremskiy]

Aleksey Vyazmikin:

I just don't understand where he saw the function there - if he put it in front of

I just don't understand where he sees the function there - if he puts it in front of a function, it's tolerable, but even the variant you suggested bothers me.

This is at the end of the enumeration (counts as a function) apparently puts, as if at the beginning of the next. Komment interline put right away. There's no such thing in 4, but code stretching strains me. It would be nice if they made styles for the styler different)

[fxsaber]

How does the language make friends with these two ChartRedraw overloads?

[Igor_Gagarin]

mt5 build 2667 (same on 2660)

One test agent stopped testing, percentage of completion rises above 100, execution/end time increases. The others have finished their part.

Sometimes this happens with 2-3 agents.

CPU usage stays at the same level.

Memory consumption sometimes stays the same, sometimes decreases.

Memory consumption is 800MB per agent, during normal operation.

Nothing unusual in the agents logs, nothing unusual in the terminal logs either.

If the agent service is restarted, it will run a couple of passes and write in a "ready" state, with less than the number of jobs completed than it is standing in the queue.

A second restart does nothing.

If I use local, not network agents, the situation repeats.

Perhaps the bug is related to the number of test parameters (73 in this case)

upd. stood like this for a few hours, then threw the idle agents 2-3 jobs each , one of them is now in the same state.

[Aleksey Mavrin]

5-ca build 2650. At some point, it gets strangely stuck: Indicators open subwindows and headers and nothing else happens. The CPU is loading decently.

I closed all possible indices and windows, only restarting the terminal helps.

[Aleksey Mavrin]

It still doesn't work. I additionally run the indicator in the debugger, it works fine, I can see the buffer values, but the chart and data window are empty.

At the same time it was working fine before this glitch, which does not go away after restarting the terminal. The first thing I did was to check for an endless loop, but the debugger shows that everything is fine, just nothing is displayed on the chart,

I've got a good impression that this happens even with built-in indulators - most likely it's in the terminal control system itself.

Has it all gone down that badly? Maybe I should switch to the previous build? I can't finish the order, the customer will have to wait(

[fxsaber]

Are these files encrypted?

Bases\ServerName\trades\AccountNumber\*.dat

[Aleksey Mavrin]

Aleksey Mavrin:

It still doesn't work. I additionally run the indicator in the debugger, it works fine, I can see the buffer values, but the chart and data window are empty.

At the same time it was working fine before this glitch, which does not go away after restarting the terminal. The first thing I did was to check for an endless loop, but the debugger shows that everything is fine, just nothing is displayed on the chart,

I've found the reason.

I found the reason for it because of the deadlocked loop in the indicator. Even if I delete the indication and close the chart, even reopening the same symbol leaves it frozen.

Shouldn't the forced closing of the indirect should solve the problem in case of an endless loop?