|
PureBytes Links
Trading Reference Links
|
DBC's explanation to me was stock quotes were the problem. If you were
using BMI or Signal for futures quotes only it would be reliable. They said
the increased stock volume in the past couple of years was more than the
bandwidth could transport without losing tick data. I think they are
incorrect about futures data being reliable based on the tick reports I'm
seeing and hearing about now. Interesting report you made about Windows 98
and ports. So changing to Windows 2000 should relief the port problems I've
been experiencing. I'm going to look into that thanks.
Robert
At 12:48 PM 7/4/00 -0700, BobR wrote:
>Haven't got much new to add on the topic of tick counts except that Cynthia
>Kase contends that tick bar charts give you a 15% advantage in timing over
>time bar charts everything else equal. I stopped making counts when it
>became obvious that others had gone the same route and came to the same
>conclusions. Whether it was in Texas, Canada, Florida, Massachussets, the
>path was similar. Drop BMI. Don't use ProSuite with the BMI Data Manager.
>Dedicate a computer for data acquisition. Make that one with a fast
>cpu(faster the better, a dual 600EBMhz is equivalent to a 1.5gighz), plenty
>of memory(256 or 512meg) so disk swaps are not necessary, or if they are use
>a fixed swap file size to keep windows from fragmenting the rest of the
>drive and locate it to the outer edge of the drive, use scuzzies or udma66,
>go to windows2000, partition it with system and applications in different
>partitions, use a larger buffered serial port such as TE920, or a USB port.
>One surprising result was that esignal via dsl modem gave similar emini
>counts as dtn satellite(115MHz to 400MHz), that was on a dual cpu with
>scuzzies. A recent test using W2000 and a TE920 port(64byte buffer) gave
>same results as W2000 on a standard serial port(16 byte buffer) both running
>ProSuite. That adds to what DTN Chameleon manual said about W98 not
>servicing the serial ports as well as W95. Looks like w2000 corrected that.
>Wonder how Millenium Edition will compare. I had initially focussed on the
>emini as a test for tick counts until I discovered that stocks like CSCO and
>MSFT had orders of magnitudes more ticks and greater percentage losses on
>BMI and I suspect DBC satellite also but have no proof of it for DBC sat.
>That was when the decision to switch rather than fight was easy to make.
>Once you start getting all or nearly all the ticks the buy/sell signals
>become clearer and your confidence in trading should improve.
>
>BobR
>
>
>----- Original Message -----
>From: Fred <srqblue@xxxxxxxx>
>To: ted stampeen <tedco@xxxxxxxxxxxxxxxx>
>Cc: <omega-list@xxxxxxxxxx>
>Sent: Tuesday, July 04, 2000 10:35 AM
>Subject: Re: TICK COUNT
>
>
> > On Tue, 4 Jul 2000 17:59:37 +0200, you wrote:
> >
> > >HI all,
> > > I'm currently using DBC-sat-feed real time, I know they
> > >are slow, but I'm wondering just by how much.
> > >
> > > I'm trading the e-mini, so I was wondering if someone here, using
>another
> > >feed , wouldn't mind telling me on average how many ticks
> > >they receive for the day session e-mini (sp)and bonds.
> > >
> > > thanks in advance for your time.
> > >
> > > Ted
> >
> >
> > <bobrabcd@xxxxxxxxxxxxx> has got to be the world's expert on this.
> > Try emailing him and asking for a re-post of some of his extensive
> > work results.
> >
> > Fred
> >
> >
> > "Success is the only test of genius" -R.Adm Daniel Gallery 1901-1977
> >
> >
|