Overview Bug

After watching what sanderling does in the overview for a long time I have come to the conclusion that this is not simple miss clicks when things go wrong it`s sanderling getting coordinates wrong .

What I have observed so far is that sometimes sanderling clicks the second column sometimes it clicks the 3rd column sometimes it clicks in between the 2 columns .
Sometimes it gets the rows wrong or even clicks the tabs at the top which is catastrophic if it clicks the tabs in the third column this basically breaks the bot because it puts distances in a random order .

Last time this happened to me I came back and my ship was approaching 1 asteroid then approaching another and going back and forth over and over I was lucky I wasn’t banned :smiley:.

So sanderling is definitely struggling to select the correct coordinates inside the overview everywhere else it works correctly it just the overview where it goes wrong.

What I don’t understand is sometimes it works ok and sometimes it’s terrible .
I now understand that sanderling process’s memory reading on your server and this worries me because if this is because of internet fluctuations sanderling is always going to have a chance of being unreliable .

I have seen a couple of other people mention overview bugs in threads here so anyone else that is seeing this bug please post here and maybe we can get it fixed.

I confirm. I’ve seen this bug a couple of times: bot clicked on the header of columns and this changed the sorting in the overview.
But It was a long time ago, Sanderling-16.06.23 If I am not mistaken. I do not use the Sanderling mining script for a long time, so do not know how things are now.

I do not know of an API for the whole column, the closest that comes to mind is the ColumnHeader on the ListView.
Ok, so when you want it to just click on the second column header, the following should work:

var secondColumnHeader = Sanderling.MemoryMeasurement.Value?.WindowOverview?.FirstOrDefault()?.ListView?.ColumnHeader?.ElementAtOrDefault(1);


Do you see any issue with this approach? Does this not click the second column header? From your description, it is not clear if that is what you used or if you tried some other approach.
Without knowing what code you are actually using to control Sanderling, it is hard to tell what is happening on your system.

As you can check with a very simple experiment, this is not the case: When you disconnect Sanderling from the Server, it will still deliver complete measurements.

The script I am still trying to make reliable is the default mining script and no I don’t want it hitting the column headers at all that’s the whole problem. When it hits the column header in the 3rd column it makes the distance column a random order which in certain situations makes the bot fail and get stuck.

If the processing is not done on the server why does it work perfectly some days then on another day it keeps going wrong? .As I have changed nothing else and there is nothing wrong with my PC (I fix computers for a living ) the 1st thing I can think of that it could be is something to do with the internet as that is the only thing left that is variable.

The only other thing I can think of is maybe the eve client starts differently sometimes which causes problems .I say this because a couple of times sanderling was not working correctly (e.g. not reading tool tips ) and restarting sanderling did not fix the issue but restarting eve did fix the issue .

I do also have an samsung evo ssd drive which is in turbo mode which uses 2gb of my 8gb memory I don’t know if that would affect memory reading.

As long as you are not sure that the problem is in the memory reading part, we should look at what part of the script causes the problem. So what line in this script is causing this?

This is by far the most interesting piece of information here. If the eve online client can have a state that results in Sanderling reading something wrong with every measurement, I could probably reproduce the problem given a process sample of that eve online client state. We already use this process when adding new features to memory reading.
In addition to the sample file, I’d need to know which part of the measurement was wrong when the sample was taken.

Ok I will continue testing and try and get you that process sample.

Edit: I just tried to go to the intructions on collecting samples following the link you gave me in another thread and the link is dead

nevermind I removed the “/1” at the end off link and it worked

1 Like