time discrepancy? "time" vs "rawtime"

[hypertrader]

The following code was used to check the relationship between present time delivered by "new Date()",
tickTime from "getValue("time",Symbol+",1T")*1" and
rawTime obtained through "getValue("rawtime",Symbol+",1T")*1"

I would have expected that tickTime equals rawTime*1000
(additionally a time zone dependent correction could be required; in my case 3600 seconds);
However, an additional correction of 60 or 61 seconds seems to be neccessary to bring rawtime in sync with time (tested with ES M4 for various intervalls).

I noticed this first when a log file routine showed that the time stamp of the ticks obtained through getValue("rawtime",Symbol+",1T")*1 was 60secs ahead of the eSignal synchronized system time.

Is this a new feature to look (one minute ahead) beyond present time, is it a bug or am I doing something wrong?

________________________________________

var Symbol = getSymbol();

function preMain()
{
}

function main()
{
if(getCurrentBarIndex() ==0)
{
var tickTime = getValue("time",Symbol+",1T")*1;
var tickRawTime = getValue("rawtime",Symbol+",1T")*1;
var now = new Date();
debugPrintln("tick time: "+eval((tickTime)));
debugPrintln("tick rawtime: "+eval((tickRawTime-3660)*1000));
debugPrintln("real time: "+eval(now.getTime()));
debugPrintln("price: "+close()+" volume: "+volume());
debugPrintln("");
}
}

[rmclean]

hypertrader,
This brings up a question I asked a few weeks ago and to which I never was able to locate the answer: I have read that rawtime does not work properly on 1 minute charts. However, those remarks were in a previous version of eSignal.

Does anyone know if that bug has been fixed?

Thanks, Bob

[hypertrader]

Hi Bob,

thanks for your comment.
I am still using 7.5 (Build 628) since I am somewhat reluctant to switch to a newer Version before it has been proven to be reliable.
However, the problem I mentioned is not dependent on the chart interval chosen.
The call I use to get tick raw time gets data updated on a per tick basis irrespective of the chart interval selected.

[gspiker]

HT,

I think I remember reading about this in another thread...and someone from eSignal said they had verified that there was a problem with rawtime and they were looking into a fix (which may have been fixed in the latest release).

This may just be old age making me mis-remember...anyone else recall this thread?

Garth

[gspiker]

Ahh...I found it. I was a bit off, but it sounds like what you are seeing has been seen before and reproduced:

http://forum.esignalcentral.com/show...awtime+vs+time

[hypertrader]

Hi Garth,

thanks for the reference to the old thread!

I presume that you are running already 7.6. Is there still this time offset to be observed?

This seems to be one of the (many) problems I was referring
to when I was requesting an up to date list of open items / known bugs, workaraounds and current status:

http://forum.esignalcentral.com/showthread.php?s=&threadid=8075

Wouldn´t it be nice to have ONE spot to check when coming across potential bugs?

[gspiker]

hypertrader,

I don't have an EFS written yet to test this. If you want to post one, I will check.

Garth

[hypertrader]

Hi Garth,

I installed the latest build of 7.6. The rawtime problem is still there.
You can check this using the function at the beginning of the thread. In this formula I have subtracted 3660 sec (61min) to get the correct(ed) - i.e. same - time out of "rawtime" as with "time".

I wonder if there is any reason not to use "time" instead of "rawtime".
Is there really a performance penalty for using this "time" compared to "rawtime" as mentioned in the eSignal Formula Script (EFS) Help Center and Library ?

[Guest]

FWIW, I ran a test at this link.

The time overhead is relative. If you are running only one instance in a single efs, it probably has no impact. If you are running several instances in each of your 10 efs's, and they are all running at the tick level, on actively traded stocks, contracts, etc, the impact can be significant.