|
PureBytes Links
Trading Reference Links
|
Those counts below are pretty much what three other traders and myself got
last week on DTN. The results of one test were surprising. It was done
with a split LNB signal into two DTN D8000 receivers. The output of one
went to a dual cpu machine with a TE920 serial port card(64 byte buffer) set
to 400K baud. The output of the other D8000 went to a single cpu machine
and the standard 16 byte buffer UART set to 115K baud. The tick counts on
both machines were within twenty ticks of each other. Perhaps under
extremely heavy volume days the tick difference might be higher. Another of
the traders and myself were running at 215K on TE920's. I on a dual 600 with
W2000 and PS2ki and the other W98(with Ensign) and the tick counts were
within 1% of each other. The fourth trader has been comparing lag time
between DTN and BMI. His statement to me on Friday was that DTN lagged BMI
in the first 4 minutes and then DTN gradually increased its tick count lead
over BMI the rest of the day. He says that is typical. Partway through the
morning using 72 tick bar charts on TS2ki/DTN and TS4/BMI, BMI lagged by 14
bars or 14X72 =1008 ticks if I recall correctly.
Now, if DTN would only listen to its new and old subscribers and add the
individual breadth symbols for up/down issues and volume for at least the
NYA, then I would be a happy camper with them. Saying that this information
can be plotted in Ensign is not satisfactory for the TS2ki and Ps2ki users
switching over from BMI. DTN has promised this breadth symbol capability
for over a year now. Last spring it was promised sometime this summer. How
much effort does it take to create a "calculated" symbol in the DTN feed?
They've already got the data in the @ISN and @VSN composite symbols. It
seems like neither DTN nor Omega are listening to the requests for such a
capability. Omega could solve this by adding another Symbol collection
template and I suppose DTN would have to add some more P codes. The
simplest solution though is for DTN to create the individual symbols. I
can't imagine that task taking longer than a coffee break. What's the
holdup?
BobR
----- Original Message -----
From: <ALVIS2JS@xxxxxxx>
To: <omega-list@xxxxxxxxxx>
Sent: Saturday, July 15, 2000 2:53 PM
Subject: Re: ESU0 daysession tickcount request
> TS 2000i - Win2000 - DTN satellite - Digi serial card with 128 buffer
>
> Thurs 7/13 22,319
> Fri 7/14 20,804
>
> Jim Alvis
>
>
>
|