Let me take a step back and review what info is/should be in the tooltip. Currently it includes:
Other possibles that have been suggested are:
Please let me know which you think would be useful to add. If we add these, do we no longer need the Comment field in the input file?
Thatās appropriate only in the case of a full match. Some sibling matches are half matches (maternal or paternal only, not both) as I described earlier. Please note that Iām not talking about full siblings vs. half siblings; full siblings do have half matches, as I described.
Having said that, I realize it makes it hard to distinguish a full match from an unknown match. Maybe unknown matches could include some dots or striping within the bar, in addition to the color? (I have no idea whatās possible technically.)
The start and stop values would be very useful, only because itās painful to have to go back to the various individual association notes while one is viewing the graph.
The relationship would be necessary if you want to support all possible types of sibling matches.
The common ancestor would be least useful.
I hope you get other opinions besides mine!
I would probably not use it unless it were displayed in the tooltip.
When you mentioned the Comment field, you mentioned Common Ancestor. Is that the only reason for the Comment field, so that if it were deduced and printed in the tooltip, you would not need it? Or do you have other uses for the Comment field over and above the fields proposed?
No, I had nothing else in mind for the comment, was thinking of it more as general purpose, but itās probably not worth it. Could always be a future enhancement if others have a particular need for it.
Hereās what would really be useful: let me right-click on a segment and have an an option to open the corresponding association note! 
(There was an earlier suggestion about clicking on a segment to switch to a different active person, but that is not what I am talking about here.)
Curious - what exactly would you edit in the Note? I cannot imagine changing the chr/start/stop/cM/SNP. And you dont want the comment any more. So is this to edit the M/P/U flag? I was still pondering whether to add this flag since it potentially breaks if the user uses SyncAssociation as the Note is linked to the new association, and the M/P/U flag is not necessarily the same for the reverse relationship. Kinda violates the principle of least astonishment.
Yes, that would be something I would do whenever I figured out which side an unknown match belongs on.
Good point, I hadnāt thought of that. Personally I wouldnāt create reverse associations for this. Since Iām the only match that any given person has (based on entering only my match data), I donāt see the need to create a reverse association for them just to point to me. And if I were to load all of another relativeās matches, those would be separate Associations and Notes and I would enter the flag accordingly.
I think youāre right about both the not hotlinking the Note and the M/P/U flag problem.
It would be immediately useful if you had an option to open the Edit Person dialog (say from double-clicking the name in the chart title) directly to the āAssociationsā tab for the Person being charted. From what Iāve heard, this might not be possible. You can certainly open the Edit Person dialog with the 1st tab (āEventsā) but not target a different tab.
But that would allow re-ordering the Legend and temporarily changing the Association Type (to something other than āDNAā) so that match will be removed from the chart.
I also tried placing variants of centromere markers on chromosomes 13-16. Youāre right there too. Even on a moderately populated chart, it is just too busy to notice them as anything but clutter.
Yes, Iām still trying to figure out how many matches I want to add to this, not so much because of the data entry but because the colors quickly become too similar (which is another reason why the tooltip is so important).
Iām currently analyzing my matches (outside of Gramps) to try to figure out the minimum number of them that I would need to include in order to cover as much as my genome as possible. This is because Iām more interested in knowing which ancestors contributed which segments than knowing how many and which cousins share those segments, but of cource other users may have different priorities.
Could there be a configuration option to do the coloring based on ancestral couple rather than matching relative?
That would result in a more limited set of colors, and make it easy to see at a glance all of the segments inherited in a given line of descent. The tooltip would still reveal the identity of the matching relative for any particular segment.
Again, this is just based on my personal bias of being more interested in where the DNA came from rather than who else shares it.
Below is a replacement file for dnasegmentmap.py that addresses a lot of the issues. Can you try this out and let me know your thoughts. It DOES add the optional M/P/U flag as well as the right-click on the segment to bring up the Note Editor.
Please provide feedback.
Update DNA Segment Map for bug 0012564 and 0012712
Addresses issues :
Mostly very good, except:
I donāt see that change yet.
With the M/P override, I find that I donāt need a āBā option for the segments where I have a full match with a sibling; rather, I can just enter two lines for the same location, one with M and one with P, and that looks better too. The only problem is that the common ancestors each of the segments should be our maternal and paternal grandparents, respectively, rather than our parents.
Oops, I forgot that change. Just added the Start/Stop cMs to the tooltip. Grab the file again and try.
Common Ancestor: Are you saying that the Utilities ā Relationship Calculator gives a different Common Ancestor than the tooltip?
I donāt know that common ancestor is viable. A distant cousin may not even have parents when you first add them to Tree and the Gramplet generates the chart. They are just floating until you work out their connection. Even if possible, wouldnāt it have to be a tooltip over the Legend rather than the graph to ensure uniqueness too?
If there is no common ancestor in the tree, then that is not included.
I put it on the segment since I am looking at neighboring segments to see if there is a commonality, for instance. Putting on the legend seemed much less interesting - and not really a driver for more research or review.
Currently if you hover on the legend, you get a tooltip reminding you what action happens if you right/left click. I would have to remove that if I put the common ancestor there. I didnt think it was obvious that there was an action without the existing legend tooltip.
No, that is working as expected. What I am saying is that, in the case where I have supplied M or P because I know (through triangulation) that the match is only on the maternal or paternal side, and the segment then appears correctly on that side only, it no longer makes sense for the other parent to be mentioned as a common ancestor.
Itās probably not worth doing anything about this unless you want to provide robust support for sibling matches, and the user has checked GEDmatch to figure out which segments are full- vs. half-matches, and has entered the data accordingly. (In other words, am I the only one who cares about this? )
I still donāt see it. Maybe I donāt know how to find the updated version. Sorry if Iām being dense.
I agree, and it would also be helpful to mention in the other tooltip that users can right-click to edit the association note.
Please verify the version number. I had a an old 1.0.7 copy of dna.py in the same folder with a 1.1.4 version of the dnasegmentmap.py Gramplet. And I was launching the wrong one.
The way it hides all the other peopleās segments when you right-click a person in the Legend to open their Edit Person dialog.
The Edit Note on right-click of a segment fells empowering too. I guess it doesnāt matter if thereās something ādoableā after that. The drill down to raw data āfeelsā right.
Any way to push the legend color swatches a couple pixels downward?
