I'm curious to know how other do their GPS on LARGE jobs.

I have jobs that last months and have GPS data that spans the entire period.

Often, new GPS data would look very nice, geometrically with old data. When this happens I add the new data and re-process all my data.

On a couple of occasions, this adding of data and adding of data and adding of data has seemed to cause some data to disappear( and reappear when I "delete" the latest data).

Also, since I'm adding data that I think helps the overall geometry, can I call a point on day one "control #1001" when it was only actually used as control(#1001) on the 4th session?

Imagine a dot plot of your project. Like the dot-to-dot projects we did in kindergarten. Each measured vector adds geometry to the picture. What order, what day, what month that they are measured is not nearly as important as the network adjustment that will take all those vectors and apply them to the geometric picture.

Get Practical GPS Surveying from the training folder on the ftp server. It is less than 20 pages of good advice.

I have still never been able to comprehend how to make data disappear from a project and then appear again but then there are computers that seem incomprehensible to me. Just because this has never happened to me does not mean it can't happen. It is always difficult to solve a problem that cannot be duplicated.

I always hated it when the tech support weenie on the phone would say, "Phil, you're the only surveyor in the world who has reported this problem." I try never to use that line but I must sometimes say, "That's a new one on me."

I have always known that each point is unique.

BUT, what if I have points 1001-1005 on day one, of which points 1001 and 1005 are conrolling points.

THEN on day 2, I now have 5 more points, of which ONE is point #1003(not a control point from day 1), which I used to now control my second day's work.

Let's imagine, just for the heck of it that I have downloaded both day's data after the second day.

NOW, if I name 1001 and 1005 as control, that makes full sense to me.

BUT, what actually happens if I then ALSO make point 1003 a "controlling" point because of the second days' data?

Will this action actually cause 1003 to control the first day's data too?

OR does A.S. somehow know that it's really not supposed to control the first day's data and sould ONLY control data after it's location has been computed?

Do I compute the both day's data and just let it fly, and THEN denote #1003 as control, then for day 2?

Here is another reminder to escape the MStar way of thinking.

A GPS vector is the delta x,y,z. We pick one end of it to hold fixed so we can use the delta values to compute positions relative to something.

During data processing I would not hold more than one point fixed on the Control Sites tab. Each vector will be computed relative to something but you will sometimes find in Ashtech Solutions that the vectors in a static project are not all processed from the control point. Nor are they always processed from known to unknown.

That is one reason why some of our friends on the message board keep reminding all of us to do that adjustment before we start looking at the coordinates. It is why the Ashtech Solutions manual agrees with them.

We do the minimally constrained adjustment first. Evaluate the control ties, create a site positions report, and then make a fully constrained adjustment. After some more analysis and perhaps a little more tweaking it will be time to export a coordinate file.

The manual actually states(set a control site) that one or more points of known coordinates should be held fixed. But let's not go there.

I'm under the impression that there's advantage(?) to actually only holding 1 point fixed, then AFTER processing to then actually entering the other control points or instituting the "fixed" feature of the control sites.

I guess I always felt funny about processing without "locking", and then relying on the "adjustment" to establish the relationships to the fixed locations.

Now, with this in mind . . . I get the distinct impression that A.S. will automatically establish point 1003(from my example above) as the only point on day 1 that got computed, THEN proceed to use THAT POINT(1003) as the basis for computations on day 2. Upon the adjustment of day 2's data, do I again NEED to state that point 1003 is the "fixing" agent for the second day's data, or is point 1003 NEVER really considered a "control point", even though it would, by default have to control points 1006-1010(the second day's occupations)?