FCP Indicator and EA

[forexmoments]

I suppose the approval is connected with the ip address.

But I use Dynamic IP, once modem reset (Or lost energy) IP will be change.. And I am sure many guys have same problem as me.

Guys, come on - why do you think you're asked for your Empty4 account number and not your IP address?

I'll let you work that one out for yourselves.

[forexmoments]

I wondered whether it is possible and permissible to register two Empty4 accounts under two different email addresses. I would like to test both in an demo and live environment how the indicator works.

Sure if you want to test on both demo and live, that's cool.

[Shoddy]

Interresting, I've signed up

[luke1711]

Nice, count me in

[renexxxx]

Just a few questions to Ash, for the entry and exit conditions:

You say that the FCP indicator correctly predicts the direction of the next 1-Minute candle. So, does that mean that the EA would just open a trade in the direction of the arrow (if and when it appears, usually within 5 seconds of the previous candle close) and would keep the trade open until the end of the current 1-Minute candle? I.e. would the EA only ever have the trade open for roughly 55 seconds? Or is there more to it? And what is the exit condition for the 34% of the cases where the direction of the next 1-Minute candle is incorrect? Is there a stop loss?

Sorry but these questions were not answered in the PDF and/or first post.

[forexmoments]

You say that the FCP indicator correctly predicts the direction of the next 1-Minute candle.

Indeed.

So, does that mean that the EA would just open a trade in the direction of the arrow (if and when it appears, usually within 5 seconds of the previous candle close)

Exactly.

and would keep the trade open until the end of the current 1-Minute candle? I.e. would the EA only ever have the trade open for roughly 55 seconds? And what is the exit condition for the 34% of the cases where the direction of the next 1-Minute candle is incorrect? Is there a stop loss?

Currently I believe exiting at the close of the candle is probably the best way forward, so yeah, this is what the EA that's being developed would do. I'm not 100% on whether this is the best way forward on how to convert the knowledge that we have, (i.e. being able to correctly predict most candle directions), into a viable trading demeanour; I am of course open to suggestions. There is an emergency SL which the user can set themselves, just in case of any crap fundamentals out the blue.

[renexxxx]

I am a little concerned about the efficiency of the signal delivery. As I understand, a signal is "discovered" by Empty4 by sending an HTTP-request to your server. As the trade is only ever open for 55 seconds, it is crucial to send this HTTP-request within the first 5 seconds of the opening of a new candle, i.e. as early as possible. So, we could eg. send the request at every second, for 5 seconds after the open of a new 1-Minute candle. For 28 pairs that would result in 5*28 HTTP-requests, every minute. Or -- , the HTTP-request could be sent at 5 seconds past the 1-Minute candle open, resulting in only 28 HTTP-requests per minute, but then we would lose the advantage of knowing the signal a few seconds earlier. I am not sure about your server-infrastructure, but with say a 1000 subscribers would you be able to handle 140,000 HTTP-request at the start a new 1-Minute candle? I know that these requests have a very small payload and cost a very small processing time, but I am worried about the scalability of this approach. As for the user-side, if they run this EA day and night (as they should) this would add considerably to their data usage (unless they run it on a VPS with unlimited data)

May be my concerns are not warranted, but please enlighten me if they aren't.

[forexmoments]

I am a little concerned about the efficiency of the signal delivery. As I understand, a signal is "discovered" by Empty4 by sending an HTTP-request to your server. As the trade is only ever open for 55 seconds, it is crucial to send this HTTP-request within the first 5 seconds of the opening of a new candle, i.e. as early as possible. So, we could eg. send the request at every second, for 5 seconds after the open of a new 1-Minute candle. For 28 pairs that would result in 5*28 HTTP-requests, every minute. Or -- , the HTTP-request could be sent at 5 seconds past the 1-Minute candle open, resulting in only 28 HTTP-requests per minute, but then we would lose the advantage of knowing the signal a few seconds earlier. I am not sure about your server-infrastructure, but with say a 1000 subscribers would you be able to handle 140,000 HTTP-request at the start a new 1-Minute candle? I know that these requests have a very small payload and cost a very small processing time, but I am worried about the scalability of this approach. As for the user-side, if they run this EA day and night (as they should) this would add considerably to their data usage (unless they run it on a VPS with unlimited data)

May be my concerns are not warranted, but please enlighten me if they aren't.

It's not a problem, since a) as you mentioned the requests are tiny, and b) even if the user base grew to close to unsustainable levels, we'd introduce more servers to handle the numerous requests. It's not an issue.

The main issue is that there have been no trades this week, heh.

[SteveHopwood]

You would have to be mighty quick to trade this manually, so I have uploaded the EA to post 2 even though it took no trades last week. Give it a go and see if it works correctly.

[forexmoments]

Seeing as though the past two weeks have been a very boring affair, the indicator will be free for a number of weeks more.