you’re right, there shouldn’t be fixed solutions
for a tractor it will be useful, for my jobs not
OK, so the next question is this:
If the correction stream is lost; and age of differential increases beyond n seconds (5,10,30, etc.); and then the software breaks the fix and goes into single mode, what will you use the single mode coordinates for?
2 Likes
When floating, I can capture data and then do the postprocess with RTKLIB, and I make sure the fixed solution
Maybe what is needed is a different indicator, like “FIX+” which has some conditions like we have in the point collection mode. Where PDOP must be under a certain value, and age of differential is under a certain value, etc. etc.
That way for some users they can keep an extended normal fix, and other users can make sure that they only work with “FIX+”.
…just an idea.
3 Likes
EXACT, that would be good
1 Like
OK, I’ll put in the request for “FIX+” and see what happens. 
edit: Done. See the grey link at the bottom of this post 
2 Likes
I don’t think “Fix+” is a good semantic. More like “Fix Degraded.”
2 Likes
I hear you. How you term it is all a matter of perception though. Playing devil’s advocate now: If the integer ambiguities are fixed, then it is fixed, and if it is better than single mode, then it isn’t really degraded, is it? 
Some of us are looking for the best fixed solution, some are looking for any fixed solution. I don’t think the second group wants to be called ‘degraded’ 
1 Like
If the integer ambiguities are fixed, then it is fixed, and if it is better than single mode, then it isn’t really degraded, is it?
No, unfortunately that wouldn’t be accurate. It sounds to me like it is using the last correction signal, and assuming nothing has changed. You can still use the old numbers and come to a “fixed” solution. At some point, they would be about as accurate as single. When we say “fixed” we are talking about math. It’s short for “fixed integer ambiguities.” All it means is that you’re rounding some numbers to a nice number that is making sense when you compare them. Being able to do clean mathematical operations on bad numbers may give us nice looking numbers, but it doesn’t mean it’s good measurements.
It’s very much like someone let go of the other end of a tape measure, and you’re still reading your end down to the millimeter like everything is great and calling it fixed, while the other end of the tape measure is swinging around in the wind.
Think about this: If I started moving your base station around slowly while you were surveying, you would still show a fix. The controller may say fixed, but that doesn’t mean it’s good data. In fact, I could move the base so that it’s lat/long reading remained exactly the same, chasing the signal like a drone. Eventually, I may have moved it 10 feet from where it started. And you’re still showing fixed. Well, that’s pretty close to what’s probably happening in fixes with ages more than a couple of minutes.
2 Likes
What I meant however, was Fix+ sounds like it’s better than a normal Fix. The perception would be exactly opposite of reality for most people reading that. Fix+ is more like a float at 30 seconds, and more like a single after a few minutes.
I fully support your words and very well explained
3 Likes
I like the analogies, and thanks for the explanation.
Sure. Again from the other side of the fence: The bottom line is that it is a fix if it is fixed. You could say, “evil, decietful fix”, “confused fix”, and “normal fix”, or just “extended fix” and “fix”. At the end of the I still like “fix” and “fix+”, but think the marketing side in me is winning today though. 
2 Likes
there is no other possibility, if it is fixed it must be fixed
2 Likes
He speaketh the truth 
1 Like
But yeah, instead of confusing the single, float, fix paradigm, maybe we just add an icon?
or
2 Likes
I agree that it would be good to add another mode for fixes beyond a certain age, and maybe some sort if expected accuracy at 95% confidence level. You could watch it expand as the time ticked up. Then it’s up to the user to determine when too high is too high.
3 Likes
OK , now we’re all agreed. Who’s going do code it? 
2 Likes
In the Promark 3 RTK when the latency was greater than 2, it automatically became floating. In Trimble I think they opted for a latency of 5 since with latency of 10 it gave erroneous measurements
1 Like