Errors, bugs, questions - page 205

[Sergey Gritsay]

Olegts:
Maybe they do, I haven't checked. I'm new to MT5 and having seen a similar picture of conversion of one timeframe to another, decided it was some bug (that's why I wrote in this thread), convinced that all these years are made up of minute bars, as it turned out it was not so, now I will check other servers, including metaquotes...

if i am not mistaken, if there are no minutes on the server before the minute bar records data from the day bar, i think it was mentioned somewhere by the developers

[Oleg Tsarkov]

IgorM sergey1294

Thank you

[Andrey Dik]

It would be better to swap out the bid and ask.

This, of course, does not affect anything, but visually it hurts the eye, it takes five seconds to move the checkboxes in the next build.

[Oleg Tsarkov]

connected to the metacvot server and felt the difference,

people, check the contents of the history folder before testing on other dc's.

it shouldn't be like this:

and this is fine:

be sure to check file sizes, annual history size should be between 25-30mb, otherwise minutes are replaced by hours or days, which would reduce the traffic for the initial download of history, economy - in a word...

If they are saving on traffic, something to think about...

[Andrey Sharov]

Olegts:

connected to the metacvot server and felt the difference,

people, check the contents of the history folder before testing on other dc's.

this should not be the case:

and this is fine:

be sure to check file sizes, annual history size should be between 25-30mb, otherwise minutes are replaced by hours or days, which would reduce the traffic for the initial download of history, economy - in a word...

If they are saving on traffic, something to think about...

I think there is something for developers to think about: if each DC has a hundred instruments with full depth of history for each instrument, you get gigabytes of wasted traffic if the user needs one instrument and one year. As far as I understand it costs a few lines of code...

[Renat Fatkhullin]

Ashes:
I think the developers have something to think about: if each brokerage company has hundreds of instruments with full depth of history for each instrument, you get gigabytes of wasted traffic, if the user needs one instrument and one year. As far as I understand, it costs several lines of code...

The history is downloaded exclusively on demand. If we do not open charts, the history will not be downloaded.

Also, only the required period is uploaded. If you are working with 2010 schedule, only 2010 data is uploaded.

[Oleg Tsarkov]

Ashes:
I think there is something for developers to think about: if each DC has a hundred instruments with full depth of history for each instrument, you get gigabytes of wasted traffic if the user needs one instrument and one year. As far as I understand it costs a few lines of code...

This brokerage company has a lot of tools, so they seem to have decided to protect themselves from a fool who will open all charts with history up to the hilt. Perhaps this will not happen on the real market. Of course mini-weeks are good as a base, but who knew that trading companies would start to change them into daily bars without warning anyone. I think they would have used a limit, say a month, in number of downloads of history data from this IP, then people would think before downloading them; now everything is primitive and sad. Guess for yourself...

[Renat Fatkhullin]

MetaTrader 5 servers have no load problems (neither CPU nor Network) from downloading huge amounts of data.

The whole system is designed to distribute any data - both trading history and chart history - as efficiently and cost-effectively as possible. In fact, all restrictions on data depth are removed.

The cluster architecture of the MetaTrader 5 platform allows for a linear increase in capacity as the load increases.

[Andrey Sharov]

Renat:

The history is downloaded exclusively on demand. If you do not open any charts, the history will not be downloaded.

In addition, only the required period will be downloaded. If you are working with the 2010 graph, only the data from 2010 will be downloaded.

My bad, I meant Tester. It could use this behaviour.

[Oleg Tsarkov]

Renat:

MetaTrader 5 servers have no load problems (neither CPU nor Network) from downloading huge amounts of data.

Why then brokerage companies compress historical data to 1 hour and then to daily bars starting from August 2010? This is an anti-advertisement for MetaTrader 5. After you download the history and decide to check your strategy in the tester and it shows one thing before August 2010, and after that timeframe does not change anything. I don't know why I decided to trade on your MetaTrader. Worth thinking about...