[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: date error, alleged
In reply to Chris,
Well, this is what is bothering me. I have been looking at the stored
images and I have been trying to figure out how I could have make such a
consistent mistake in so many places. I write the date on the image disks
by hand the next day after the run. So I not only have to have set the
computers clocks wrong on three computers, but have to also read the date
wrong from a fourth, fifth and sixth computer (the linux processing
computers) in order for everything to agree. I also have to have continued
to make a date error when writing on the image disks for a number of days.
As everyone knows, the linux computers run forever without crashing, so it
is hard for somehow to have all the computers set wrong. Further, at the
end of the month they are correct again. So somehow I have to have changed
the clocks on three computers by a day a second time without noticing it
since they are all reading the correct date now. Well, one has since died.
But just before it died, it was reading the correct date.
OK, I might have done all the above for one computer, but doing it for nine
computers is a stretch. So I am really interested in hearing if it is just
a TOM2 problem or whether it is TOM1, TOM2 And TOM3.
Hmmmm! I would much prefer that I made a mistake in setting the clocks
than if it were something due to computers. Sigh! Then we would have to
try to figure out something in Bill's code.
Doug Welch has a good suggestion in a weather check. He is also right
about the reputation of the data. I am only slightly paranoid, but it
seems to me that there are those out there with deliberate intent to attack
the credibility of the tass data. This is one problem with running an
operation out in the open as tass does. Usually, these things take place
in the dark recesses of an experimental collaboration. One only sees a
final presentation. As the Kaiser said, it is not so good to let the
general public know what goes on in deliberations of the privy council, or
what actually goes into their sausages. Not quite the correct quotation,
but you get it. It is not so good for the general public to see what goes
on in data reduction. Sigh! But we are doing it in public, and we shall
have to live with the consequences. I would not have it any other way. In
the end, the tass data will live or die by it's utility. I can only do the
best I can.
Tom Droege
> [Original Message]
> From: Chris Albertson <chrisalbertson90278@yahoo.com>
> To: tass <tass@listserv.wwa.com>
> Date: 6/14/2004 11:47:56 AM
> Subject: Re: date error, alleged
>
> Is there a way to trace a data point in the database back to
> a FITS image file other then by reasoning it out like "If it
> has this time then it must have come from this image". In other
> words could someone do an SQL querry and ask for all objects from a
> specific XXX.FIT file? (If not then one more thing to fix.)
> But assuming there is tracability it might be worth looking at the
> time stamps on the FITS files and comparing to the date in the
> database. The error could have been introduced by software
> at any point along the way. In fact if I were betting I'd give
> even odds to a software bug and Tom entering the date wrong
>
>
>
> --- Michael Koppelman <lolife@bitstream.net> wrote:
> > I agree. We have some evidence that suggests this may have happened
> > but
> > there are other explanations as well.
> >
> > I know there are plans underway to synchronize via the network in the
> >
> > near future. This will be a very good step forward. Sub-second
> > accuracy
> > of your clock is very achievable and is the only way to go.
> >
> > Cheers,
> > Michael
> >
> > On Jun 12, 2004, at 6:18 PM, Thomas Droege wrote:
> >
> > > I wish you all would quite calling it a date error. I don't think
> > we
> > > know
> > > that yet.
> > >
> >
> >
> >
>
>
> =====
> Chris Albertson
> Home: 310-376-1029 chrisalbertson90278@yahoo.com
> Cell: 310-990-7550
> Office: 310-336-5189 Christopher.J.Albertson@aero.org
> KG6OMK
>
>
>
>
> __________________________________
> Do you Yahoo!?
> Friends. Fun. Try the all-new Yahoo! Messenger.
> http://messenger.yahoo.com/