Support Board
Date/Time: Tue, 01 Jul 2025 15:06:14 +0000
[Programming Help] - Can I develop using C++ or do I have to use the acsil framework?
View Count: 1043
| [2022-03-11 01:09:32] |
| User32960 - Posts: 1 |
|
Its not 100% clear if i can just build my own cpp against and api or not. Thanks. |
| [2022-03-11 02:31:13] |
| 1+1=10 - Posts: 270 |
|
ACSIL is a C++ library used for creating SC Custom Studies which are integrated into the platform. However, your ACSIL dll can include any other C++ libraries, provided you compile the dll yourself. (So you can more or less do anything with ACSIL that you can do with C++.) ACSIL itself has functions to make HTTP Get/Post calls and even run external programs by using SC’s built in alert functionality. SC also provides the DTC Protocol which allows sending binary/JSON messages to get marker data and send orders to the SC program. Here you can use any programming language and have a completely separate process. The two downsides are: 1. implementing the DTC protocol is a significant task. 2. you can’t get live CME Group data through DTC! There’s an example of creating an SC DTC Protocol client a user posted on the forum as well as quite a few GitHub repos I could point to if it would help. |
| [2025-06-21 07:48:13] |
| User373245 - Posts: 57 |
|
I got NTFY to work and can send test alerts from the send test notification as well as from the . SC chat ...how do you tie the simple alert to it all?
|
| [2025-06-21 14:02:42] |
| ForgivingComputers.com - Posts: 1072 |
|
There’s an example of creating an SC DTC Protocol client a user posted on the forum as well as quite a few GitHub repos I could point to if it would help.
I would be interested in looking at a working example. Learning the DTC Protocol seems like a daunting task, and understanding the hoops I have to jump through will help me decide if it is worth it. |
| [2025-06-22 22:02:03] |
| User373245 - Posts: 57 |
|
Those messages are incongruent. I got NTFY to work from the SC platform in General. Not ACSIL to DTC. DTC is my last ditch effort for that exact reason. I know getting it to work will be a battle! Honestly I am trying to build it all ground up from direct from a different data feed but I use around 36 calculations on 4 time frames, for a single signal to the buy and sell side of my Algo. To study it I need 8 million rows of data to build it out. The live API isnt a big deal but converting sec aggregate data to real time and not end of bar in some really complicated engineering..it is really difficult. No out of box library works in the iteration for a calculation within any given second. So I have to build ground up, everything from the iteration to the function call. And doing that from 8 million rows of data per asset per calculation...is not an easy task at all! And I am not sure I trust myself as an engineer to get it right in the look ahead bias aspect. It's why that replay environment is boss with Machine learning. You know if you put a calculation on that chart and run through 8 years it's exactly as it would have been live! This is critical in ML but getting that piece of information over to the right broker in the right way! HOLY SMOKES! I have a rough outline of DTC script but have not committed to wrestling it yet..SC is talking about alternatives outside NTFY ???if I can get the webhook aspect of that to function....I will just export a base signal and combine it with my ML script running on my computer and than shoot that to the broker. But I dont like it all! It feels like the alley way! I certainly would not feel comfortable putting substantial capital at it....I have only been engineering for 3 years so I have a lot to learn still! But If I jump to DTC and get something working and compiling..I will let you konw. |
To post a message in this thread, you need to log in with your Sierra Chart account: