Bug: Preference for Family Tree Starting Person
Since the New Person page became active, the preference setting in a user's account for Family Tree Starting Person (root) is not functioning correctly. The default should be the person's preferences, instead it is the last person the user was working on. This happens when one opens Family Tree - any tree setting, and also when one goes to "Person".
Comments
-
We are currently working on this issue. There is a new value (last visited PID) being stored and used when opening a pedigree pages or the old person page. This new value wasn't supposed to be used in all these places. It takes a little time to make sure we fixed it without making things worse.
2 -
Update (current behavior and sorry some more Ideas):
When setting Starting Person - the Tree now stores Home Person (yourself) and Starting Person at the top under Recents (Yay!).
Comment: Home Person button now takes the user to the currently set Starting Person - so either the definitions need to change or documentation about what Home Person is needs to change. If Home Person and Start Person are the same - the same terminology should be used. I always thought I was the Home Person - but now I'm just another Starting Person option.
Ideas: Increasing this settings storage (more than two) - could be helpful to researchers to restore Recent/Starting person currently working on - though it would be nice if wanting to set that there were a quicker way to do so from Tree rather than Settings (probably an option in Person card). The reason increasing this would be useful rather than relying on Recents is that one can quickly change Recents list until a person you were researching is no longer on the list. Making such a feature interact with Planner - would also be useful for researchers (a place to store current research - which would probably make Planner more useful).
Question (sorry this is now about Recents and Search features): Does the Recents Search capability only search that small Recents list? People have been posting for a Search Tree with Relationships feature/filter. If Recents is able to store recents then why not a Related able to store related - which is then searchable?
0 -
For your last question: input in the box at the top of Recents does one of two things. If the input is in the form of a PID, then it generates a URL for that PID's details page (https://www.familysearch.org/tree/person/details/NNN-NNNN) and takes you there. If the input is not in the form of a PID, then it searches your Recents list for that string of characters. Your Recents list consists of the primary name, dates, and PID of your 49 most-recently-visited profiles, plus your own profile.
A few things to note: one, inputting a PID makes no reference to your history, recent or otherwise. You can input the ID for a profile you've never before visited, and it'll generate the URL exactly the same as for any other ID. Two, the string-matching for the Recents "search" is fairly strict: it's case-insensitive (it'll find "Deceased" if you type "deceased"), but it makes no substitutions or equivalences for diacritics (Mária does not match Maria).
Your Recents list is basically a small cookie. It probably just stores the PIDs and looks them up to generate the actual list, which means that it's a file consisting of 450 characters including the spaces or other separators between IDs. That's less than half a kilobyte. In other words, generating the list of names and dates to go with that file and then running a basic string-match on it is a simple and finite task. The only updating necessary is to drop the oldest entry off the bottom of the list when you add the newest one to the top.
Your list of relatives, on the other hand, is (1) probably a lot more than 50 people, and (2) not stored anywhere -- and not storable, unless every single one of your relatives is read-only. The list would have to be newly generated for every session, by looking at your profile, listing all of the profiles it's connected to, then looking at each of those profiles and listing everyone they're connected to, and so forth and so on. It's not simple and not finite, even if you limit it to, say, four generations.
I am strongly reminded of the bird-photography xkcd panel (https://xkcd.com/1425/).
2 -
Thank you for putting in the hard work in answering the question. Yes, I am aware that Related would be a different search.
If I limit it to four generations starting with me, parents, grandparents, great-grandparents - then I have to decide if I include relations of those great-grandparents (parents,siblings) OR if I only include descendants. If I include those 2great-grandparents and descendants only - and account for a decreasing birth rate - I probably build out some hundreds to possibly a few thousands person tree... at least in my lines/tree... sorry I can't give a great estimate because I really don't know all my cousin lines - even within that few of generations (yes - your point).
...unless every single one of your relatives is read-only.
Thank you for proving my point that known relatives should be made read-only.
If the routine were to try to calculate all possible relatives - you are correct, it could be a per session possibly infinite loop... The point is that it could be done. Per session for limited generations - as the Tree apparently does already or as prototype demonstrated in another thread you've shadowed me on. The call to generate Relations - whether by Tree or some other app could be done at any point and within a certain amount of generations at the time of Search. Either way building that tree/relationships could be done in the Search instance/call (not trivial nor possibly finite but could be limited either by threshold cut-off or generations/descendants). And certainly if a large percentage of the profiles is read-only - then could be done much easier... again another great proof for why profiles should be on the path to read-only once/if proven sufficiently.
As far as searching relations - on another post I think I put the idea follow all your relatives up to the 4000 limit - then you can find them in your Following list.
0 -
In my family, counting just ancestors, their siblings, and the spouses of those siblings (not any of their descendants), I'm getting 95 people at four generations.
I'm too mathematically-minded to believe that anything in genealogy can ever truly be proven.
(I don't think my sister has posted anything here since the time limit on edits was instituted.)
0 -
@Julia Szent-Györgyi I thought math was about proving things... it's full of proofs. I also thought you said in another post 'genealogy is like sudoku'? Sudoku only has one right answer - so definitely proven. But I forgot - genealogical proof standard I think is more closely akin to a legal proof standard (beyond reasonable doubt) - sorry I ain't a lawyer neither.
I think counting aunts and uncles - no descendants - for 4 generations I'm up to ~ 250 (estimating somewhat ... I guess you are more mathematical).
0
