|
PureBytes Links
Trading Reference Links
|
Dick,
Thank you for your answer.
At one time I tried DTN. I noticed on a couple of occassions that some of
their "end of day data" (Indexes)
did not display the same prices as printed in the paper or on Yahoo. Have
you noticed that?
I currently use BMI cable feed at 38.4kb.
You mentioned that Signal transmits at 56kb through cable. Do they use the
same compression technology as BMI?
If they do, why not switch from BMI to signal? If they don't, are they
transmitting faster than BMI?
Is there data that BMI transmits that Signal does not?
David
.
Dick Crotinger wrote:
> David wrote:
>
> > I meant Isn't signal slower than BMI?
> > Does that mean the indexes will actually be slower?
> >
> Well, yes and no. The original Signal black box (back when it was run
> by Lotus) was 9.6Kb... the next upgrade of this product was double the
> speed, at 19.2Kb. At that point, BMI was faster than anything Signal
> offered, at 38.4Kb (where it remains today).
>
> Since those days of yesteryear, Signal has come out with a cable feed at
> 56KB, so that is the faster alternative. They bought BMI to obtain the
> fast satellite capability (and to eliminate competition, of course).
> For sheer bandwidth (via satellite) this is the fastest medium currently
> available at DBC.
>
> I'm currently running DTN side-by-side with BMI... DTN runs at 115KB, so
> is triple BMI's speed. My impression is that DTN pumps the quotes a
> good bit faster for most markets (the same quote on stocks arrives up to
> a minute quicker on DTN)... BUT it's really apples and oranges, since
> they both use different compression technologies in the datafeed, BMI's
> being the more efficient. This means that BMI would assume the quicker
> delivery in burst (fast market) situations. It's still not clear to me
> which I like better. I'd hoped the options side of DTN would be
> superior, but so far, I'm not convinced.
>
> Whether or not this expedites the delivery of indexes would depend on
> whether the index is computed by DBC or by the data source (i.e., the
> exchange)... my guess is that BMI used to pass through the index quotes,
> versus Signal computing it in-house, but I'm not sure of this. Now that
> BMI and Signal feeds are borne of the same womb, index delivery speed
> would be predicated on the speed of the medium alone.
>
> Dick Crotinger
|