We are still working on our reference data APIs. We're hoping to have v2 of them out in the next month or so. However, the real-time and historic price data should be very solid. We're a very determined team of developers, so please reach out if you have questions/comments. Cheers!
Simultaneous meaning open connections or number of tickers you can stream at one time?
You can open 1 websocket connection at a time. But you can stream as many tickers in that connection as you like.
Does this answer your question?
1: The non-display fees do not apply to us as a company because we only touch the data in an effort to further redistribute the data to our users. It also does not apply to non-professional users, however professional / enterprise customers it does apply.
2: There are quite a few providers who...
The exchanges do not provide these APIs themselves for indirect feeds. For the SIPs ( CTA & UTP ) they don't even offer connectivity directly, you must use ICE SFTI and/or Nasdaq ( or another provider who gets their data from those interconnects ).
I guess it's also worth noting, if you get a...
We actually do stream real-time NOI data on our websocket service. It's not documented yet as it's pretty new. Just email support for more information about this.
We hope you give our platform a test drive. We offer 7day free trial to play around and see if it will fit your needs. If there is something we dont cover, we always appreciate feedback on how to make the product better. Take care!
Yes, the snapshot feature allows you to see all tickers bid/ask, last trade, previous day OHLC, and the last minutes OHLC. We do not limit API calls or ticker symbols, so you can poll this every few seconds throughout the day to stay in sync.
That was before we started getting data directly from the exchanges. It's now ~4x faster ( 20-30ms ) from the SIP timestamp to the time a client receives the data ( on including geographic latency, we are in NJ ).
@qlai
We consume the CTA and UTP feeds directly from ICE SFTI ( http://www.nyxdata.com/docs/connectivity ). We have 2x 10G cross connects, A networks and B networks for full redundancy.
We parse the UDP Multicast binary feeds from the exchanges ourselves.
Our average latency (broadcasted)...
The proximity is one advantage
Consider:
- Datacenters have much more reliable internet and power.
- A dedicated server will be a little more reliable with ECC memory, dual power supplies and RAID disks.
However:
- If you're just using the server for proxying/RDP'ing, then latency will be...
The latest Mac Pro towers can no longer upgrade to the latest Mac OS ( I believe they are stuck on the previous version currently ).
You can also use Docker on Mac if you need a linux kernel for some things.
This is a little late, but:
I believe if you're registered with the SEC, you're a "professional" even if you're not currently using the data in a professional capacity.
The volume of data can cause lags at all points leading up to your connection. Also, Internet and/or TCP can add quite a bit of overhead. Especially if there is packet loss. If you're looking for consistent, low latency - UDP Multicast with a cross connect is the best option - although, it's...