|
PureBytes Links
Trading Reference Links
|
Gary,
Probably a long shot that this could be related to the problem you had, but
since you collect data from QFeed, it might have some relevance.
I collect data every 120 seconds via QFeed and feed it into my own
software. I was gone for most of the Market today, but when I got back I
attempted a basic update of my data files from the QF data that had been
collected. Immediately, the update program started spitting-up and
couldn't finish the update.
I had to do a little brain surgery to find the problem, but it turned out
that the QC Sunnyvale-4 server had delivered a bunch of data with holes in
it while I was away. I switched to a different server (Sny-02),
re-initialized the files, and everything updated perfectly.
Just my $.002
_____________________________________________
At 12:06 PM 03/11/2002 -0700, you wrote:
> > purely a guess--sort of like checking to see if your "broken" VCR
> > is plugged in--check to be sure you're using the same expiration
> > month.
>
>Yes, it's the same symbol.
>
>(But why is it flashing "12:00" ?? :-)
>
>I realized that the mystery value of 1533.7 *was* the low for the
>bar, until 2 minutes before the bar closed. In the 11:58 1min bar it
>dipped below that, and on the 11:59 bar it got to 1132.1, which was
>the low for the 30min bar.
>
>His PC clock is running fast. I suspect something was screwy in the
>way TS4 handled the close of the bar. Even though his chart shows
>1132.1 as the low for the 30min bar, and that **SHOULD** have been
>the Low that the system code saw, I think in realtime execution it
>actually closed out the bar a few minutes early and used the wrong
>value of Low. I.e. the chart was built using the Quote.com timestamp
>and the system execution bar was built using the PC timestamp, or
>something like that. Screwy.
>
>And in fact we just wached the close of the 2:00 ET bar, which popped
>a new high in the last few seconds of the bar, and exactly the same
>thing happened, as I expected now that I understood what was
>happening.
>
>So that explains the second problem. I don't think the same thing
>could explain the first problem (where no signal got generated at
>all), but I'll see if I can explain it that way.
>
>Gary
|