|
PureBytes Links
Trading Reference Links
|
Tuesday morning
Dear Ron & Groups,
Procedures ---
On a P233 using Win98, I am concurrently (and only) running:
A. Quote.com datafeed and Quote.com charts; and
B. DynaStoreLight receiving data from Quote.com and feeding it into TS
4.23, with a portfolio of about 10: several futures, several indeces, and
several stocks.
My monitor shows a 1-Tick INDU chart from each of the two programs.
Concurrently, I am also running a P100 with Win3.11 and TS 4.23 using BMI
satellite datafeed with a portfolio of approximately 20 futures and indeces and
charting a variety of issues, including a 1-Tick INDU chart.
Observations ---
Usually, the two P233 charts are indistinguishably different. I have not seen
ticks on the Quote.com chart that are not on the DS/TS chart. Further, the
ticks on the DS/TS chart appear more or less simultaneously in time with the
corresponding ticks on the Quote.com chart; if they do lag, it is usually not
more than a fraction of a second, perhaps half a second at most.
This morning, at around 1412 Eastern, INDU dropped precipitously. The frequency
of ticks on the Quote.com screen rose dramatically; those on the DS/TS screen
increased but not as much. As I write this, the Quote.com screen shows a time
of 1450; the DS/TS screen shows a time of 1427 --- in other words, a lag of 23
minutes. I have scrolled the Quote.com screen back to a time of approximately
1427, and there it shows the same INDU amounts and pattern as are now seen on
the DS/TS screen, which shows a time of 1427. (From which I conclude that the
difference between the screens is not merely due to both screens showing the
same data but with a discrepancy in the time stamping of the data --- the
difference appears real, in other words, that one screen is now showing the same
data that the other screen showed many minutes ago.)
Later: The Quote.com chart now shows 1457, and the DS/TS chart shows 1430: a 27
minute difference.
Later still: The times are now 1503 and 1432, respectively. At this time I
stopped collecting data with the Dynastore program and immediately restarted
it. The DS/TS chart now began showing ticks at a time of 1503 (producing a gap
in time between 1432 and 1503, with no ticks showing for any time during that
missing period of time). Approximately 10 minutes have now passed, and the
Quote.com chart shows 1513, while the DS/TS chart shows 1510. Visually, the
DS/TS chart looks perhaps 20 ticks behind the Quote.com chart.
Later: The trend continues: the times are now 1531 and 1525, respectively.
Later still: The times are now 1544 and 1434, respectively.
Throughout the period of my observations, the 1-Tick INDU chart from the BMI
satellite datafeed has looked identical to the Quote.com chart, although with a
consistent lag of approximately 2-4 seconds.
Conclusions ---
At an unusually high rate of ticks, the Dynastore/TradeStation-combined programs
running in my P233 computer did not chart the ticks concurrently with their
reception, resulting in an increasing delay between the times when the ticks
were received and charted. Stopping and restarting the Dynastore program
allowed the DS/TS program to begin fresh, without any initial lag, although at a
cost of loss of data corresponding to the period of time within the lag. The
lag, however, also begun fresh (at "zero") and continued to increase.
Furthermore, in two occurrences, once a lag was established, the DS/TS programs
were unable to decrease it, so as to return to the state where they charted the
data concurrently with its reception.
With the BMI satellite datafeed, the TS charting program consistently lagged the
Quote.com charting program by several seconds, although that lag stayed uniform
during all my observations. This several second lag appears to be inherent with
BMI satellite data.
***************************
Did anyone else make similar observations?
Does anyone else have alternative conclusions?
How can this problem be overcome???
Sincerely,
Richard
|