Is there any other way of reading buffers? - page 3

 
nicholi shen:

I think when you include the code with the EA the IND is calculated using the same thread as the EA. I make that assumption because the EA is way slower when you bake in indicator code. 

Yes 1 EA (all inside) - 1 thread.

1 EA + 1 iCustom call - 2 thread, one for the EA, one for the indicator.

Is this what you mean ?

 
Alain Verleyen:

Yes 1 EA (all inside) - 1 thread.

1 EA + 1 iCustom call - 2 thread, one for the EA, one for the indicator.

Is this what you mean ?

Yes. I was commenting to your point of "0 benefit". I'd also add that it's detrimental to performance, and it shouldn't be done unless you need to make something that easily distributable and want to hide its implementation details. 

 
nicholi shen:

Yes. I was commenting to your point of "0 benefit". I'd also add that it's detrimental to performance, and it shouldn't be done unless you need to make something that easily distributable and want to hide its implementation details. 

Yes and no. Yes 0 benefit (in general I insist, particular cases may always exist). No it's not even needed if you want to hide implementation details or be easily distributable as you can also embed the indicator as resource in the EA executable.