Added support for this a while ago.
I am not fond of ellipsisā¦ I did add a separator bar though to delineate the crossover in form of options if that sort of makes senseā¦
Ok.
Itās far better to have menu feature that omits the ellipsis that one that falsely appends it.
Several Tools in Gramps mislead the user with an erroneous ellipsis. Conning users into thinking theyāre harmless to explore ā that there will be an safe way to bail out before the tool begins to independently change the Tree. But then they evilly mangle your data as you watch helplessly.
(Maybe it is just the Windoze version, but one of the features where Iāve NEVER found a working example in gramps is a Cancel/Abort for a long/progressbar operation. Every one of those feels like theyāre laughing at you, saying they are too important to abort. Iām almost surprised there isnāt a snarky āBut surely you were smart enough to do a backup, werenāt you? 
ā follow-up dialog.)
The new view looks great!!
My question is why you put the view under āPeopleā and not āRelationshipsā?
I see this view as similar to what @Nick-Hall is doing with the Combined View which he placed with āRelationshipsā.
Your new view does everything (?? I have not fully explored all features yet) that the Relationship view does and a few more. All three views takes the active person and aggregates and organizes events and people connected to that person.
3 Likes
Is this (relation filters and relation category filters) implemented? If that is the case, how do you use these options and what are their effect?
These Relation Filter switches do have an impact.
The options definitely add more Events to the Timeline where that extended family member holds a āPrimaryā (or optionally, a āFamilyā) role. (I havenāt checked how the options interact with other Role types.)
Obviously, there is going to be a processing penalty for including the Events of more people. (2 generation of ancestors & descendants could be a LOT of peopleā¦ particularly when in-laws & blended family are added to the mix.) And even more penalty for including more than the core (Birth, Marriage, Death, Burial) life Events.
Itās going to be necessary to determine how badly latency increases with a larger Tree to search. 3 generations in a Tree of 500 people might have fast refreshes. But those same settings in a tree of 50 thousand might become mostly unresponsive.
Also, some exploration is needed to see how the timeline upper & lower limits are determined. What happens for an Active Person with no dated Events? Without a known lifespan for setting limits, are ALL Events for ALL generations added to the timeline?
It is still very much in development, it does not do everything and I am certain lots of things I have not considered.
When I converted from Gramplet to view I used the Combined Nick as starting point at first and then refactored to not use the newer navigation class but the one in Gramps core original relation view.
It could end up under Relationships, it could end up under both. Minor detail for later on.
Throwing an unexpected error here. Crashes the program with some people, not all!
How do you prefer to handle the report? paste it here or somewhere else?
1 Like
Fringe cases like someone with no dated events probably broken, havenāt tested stuff like that. If no birth or death events found I am calling the gramps routine that attempts to estimate them and use that but that may not work in case of someone with no dated events, donāt know.
The timeline likely needs refining. For a person it is bound by birth event and then last dated event if a death event was identified otherwise estimated date of death. Relative events before and after discarded. Ancestors walks back but not forward, offspring only walks forward. With 50,000 people itās not looking at all of them, itās walking through the parent and child families. 3 generations seems a reasonable bound.
I had in mind other timeline views in other Gramps views possibly. Family would be all events for a family for example, from birth of parents (or marriage optionally) till death of last child. So that would bound it. But then Place might be all events at a given location within a given timeframe, or Event all events within a given timeframe. Those later two wouldnāt be in reference to a person or family so for those a 50,000 person tree would have an impact. I am unsure how useful those later two might be for people to have so they may never get done.
2 Likes
If you run Gramps from command line and can copy/paste or screenshot error and open an issue on github and provide what detail you can around the people and what you think about them is causing it that would be best.
1 Like
Yes, they should. The event category filters are not extensively tested.
If you pull latest code I committed there is now timeline coloring scheme option to color with Gramps person defaults, or based on relationship, or event category or confidence level for the citation supporting the event with the highest confidence score.
I also commit changes last night that should improve shared event handling and show role/participants.
2 Likes
Also suggestions for enhancements welcome in addition to bug reports. If those could also be submitted as issues on Github if no one suggested before would be appreciated.
Now back to my real day job. 
3 Likes
Just a reminder. @cdhorn has posted several more updates over the last 10 days and transcribed enhancement requests to his GitHub section. This conversation has moved to there.
2 Likes
Could testers be on alert for something seen but not yet isolated.
Iāve had a few instances where the Profile view modeās extensive Configuration dialog didnāt share its toys with Gramplets. Only the tabs related to the Profile were displayed. The Gramplet configuration tabs were not shown.
In my case, I had added the Quick Report Gramplet to the Bottombar and intended to select a different QuickReport via the Configuration. The Quick Report tab didnāt show as the Configuration was invoked repeatedly.
I began a bug report ā¦ then suddenly, I couldnāt reproduce the issue anymore. Please watch for any behavioral patterns.
Gramps minimum Python version supported is 3.3. So we should not rely on features of Python or its libraries from versions later than that.
Python 3.3 was released in 2012 reached End of life in 2017.
Python 3.5 which included type hints was released in 2015 and reached End of life 2020.
I guess itās already getting hard to remember what features were around back then and which not. 
The viewās config window needs to be brought down to size. I have to reduce my theme font size to get the config window inside of 22" 1680x1050 monitor. Maybe having the panel of the larger tabs scroll.
Even after reducing the font size, the config window still takes up most of the monitor.
Ah, I did not realize that, thank you for the heads up.
I donāt know if you or any of the other core Developers have had time to look at it recently from usability or code standpoint. Iād be interested in feedback if you or anyone else see better ways to do things I did not think of, or things I may have gone about incorrectly. I think the overall framework is sort of there now and most of what is left is refinement and details, or at least I hope so.
1 Like
Yeah that needs more work though a bit unsure how to best approach it. My main goal last pass through that stuff was to change the structure of the option namespace and consolidate the dialogs. But I agree the layout needs to be more accomodative somehow.
Just made a commit that adjusts things a bit. I moved the tabs to left side and the notebook pages are now scrollable and pulled in some. Let me know if you think that is a bit better or have any other suggestions to improve it. Thanks!
1 Like