FOREXConnect market data responses missing milliseconds
Posted: Mon Jul 11, 2016 4:34 pmI'm new to this group so please pardon me if this first post seems awkward...
I'm using Java ForexConnect 1.4.1 (universal distribution for OS/X) running under OS/X 10.10.5. The problem I'm having is reproducible using the demo in
If I request tick data that is too close to the current time (seems like without about 8-10 minutes of the current time), the milliseconds field of the tick data timestamps are 0. Here are two examples:
It is currently 16:53. The output for a request for ticks from 16:30 to 16:49 is attached as zero-ms.txt. Note that the initial request produced ticks that have zero for their milliseconds field although the subsequent request (to obtain the rest of the data in the interval) produced "good" millisecond fields.
...a bit later, 16:59... The output for the same request is attached as better-ms.txt.
This appears to be a bug either in the API or data missing from the provider's ticks. It is problematic for developers counting on accurately timed ticked data though. Any thoughts?
Best,
David Shaffer
I'm using Java ForexConnect 1.4.1 (universal distribution for OS/X) running under OS/X 10.10.5. The problem I'm having is reproducible using the demo in
samples/java/NonTableManagerSamples/GetHistPrices/
If I request tick data that is too close to the current time (seems like without about 8-10 minutes of the current time), the milliseconds field of the tick data timestamps are 0. Here are two examples:
It is currently 16:53. The output for a request for ticks from 16:30 to 16:49 is attached as zero-ms.txt. Note that the initial request produced ticks that have zero for their milliseconds field although the subsequent request (to obtain the rest of the data in the interval) produced "good" millisecond fields.
...a bit later, 16:59... The output for the same request is attached as better-ms.txt.
This appears to be a bug either in the API or data missing from the provider's ticks. It is problematic for developers counting on accurately timed ticked data though. Any thoughts?
Best,
David Shaffer