S6 A tree from another program

A family tree from another program (MyHeritage, Gramps, FamilySearch, Windows programs and others) comes into the app and into the research as a GEDCOM file. The chapter follows the whole path: what loads and what is lost, how it shows where a fact comes from, why the family tree's facts are only leads, how to go back to a fact from the family tree, what the card What next offers, how to go back in time, how to send the tree back and what a family tree loaded again does. Describes the Strom app 3.26.0 and the research 1.30.0. The details of reading the file are in A2, the export in A7, merging in A8 and the research taking over a family tree in R11.3.

A file written by Strom itself (the header 1 SOUR STROM) and the app's JSON file are not a family tree from another program: they have no origin mark, no card What next and no protection of the family tree's facts. Research that takes in such a file has its values as the user's family tree (S6.3).

S6.1 What loads from a family tree and what is lost

The app takes over the people, families, sources, notes, events and media it knows, and in addition the IDs the program gave to people and families. What it does not take over it counts and shows in the import summary; nothing is lost silently.

• Enforced by the code:
  • the IDs of a person and a family (_UID, UID, RIN, EXID, _FSFTID, AFN) and the original record identifier are kept and come back in the export (S6.6) and into the research;
  • the import summary has the line “Kept for the export “Back to {program}”: N tags (…)” when the file has the program's own tags.
• Limits: photos the file only references by a path or an address are not in the file: the summary offers Attach media files… (A2.1.1).
• Controls: Import → Import GEDCOM; the import summary (the window GEDCOM conversion).
• Since: app 3.26.0 (IDs, media records, own tags, the Windows-1250 encoding), research 1.30.0

S6.2 Reliability marks: where a fact comes from

Each fact of a person has a mark of how reliable it is: who confirmed or proved it, or that it came from another program's family tree and nobody has verified it yet. The steps differ by dots, a tick and a frame, never by colour alone.

Other facts show the research's degree of certainty (R3.1). All the marks, the mark on the card in the tree and verifying facts against the originals are described in S7.

• Enforced by the code:
  • the mark From family tree holds while the fact has the value from the file. An edit ends the mark; for a fact changed in part it says which part is from the family tree and which a person changed;
  • for a tree linked to the research (S1) the app takes the fact's origin from the research; otherwise from the app's record of the import;
  • a tree without a record of the import (an older import, Strom's file) does not show the mark From family tree: nothing is guessed;
  • in a research tree, a fact the research took in from a family tree from the Strom app and nobody has changed since has the mark of its degree of certainty, without a degree From family tree; never ✓ (S6.3);
  • a line at the person says where the facts come from: “All of this person's facts come from {program} (file {date}, loaded {date}).”, or “Facts marked “from family tree” come from …”.
• Limits: that a fact has no source the app does not yet show at the fact: there is no mark No source at a fact (S7.1).
• Controls: the person card (the mark at the fact, What the marks mean); Settings → Person card → Reliability at facts in the person card (on by default) and Show on card → Sources and reliability (on by default; the card shows the weakest of the facts name, birth, death, S7.2).
• From app 3.26.0: two full dots at Proven go to a fact for which two readers of the record agreed when the research verified it; the card in the tree has one mark instead of the earlier two settings.
• Since: app 3.26.0, research 1.30.0

S6.3 Facts from a family tree are leads: the agent proposes, a person decides

Another program's family tree holds someone else's conclusions. The research takes its facts as leads: it changes them only by a record it cites, otherwise it only proposes the change. What a person entered or changed, in the Strom app or in the research, the agent never changes: it only proposes.

• Enforced by the code:
  • each fact in the research has an origin: from the family tree, from a person, or from the research;
  • a value from a family tree from the Strom app (a file or a tree of the app that the research took in) that nobody has changed since it was taken in is the user's family tree's: the agent does not change it, it only proposes, as for a person's fact, but it does not count as confirmed by a person. An edit, a value put back, an accepted proposal and an edit sent from the app are a person's fact;
  • a fact from the family tree the agent changes only with a cited record; without one, and always when a person decides about hypotheses (R10.3.2), a proposal arises as a conflict for a person to decide;
  • a person's fact the agent never changes, not even through a file loaded via strom sync; always only a proposal;
  • a proposal whose proposed value already holds closes by itself. A proposal whose value has changed since it arose is stale: it cannot be accepted, only closed with the value kept;
  • the value from the family tree stays stored next to the changed one (S6.3.1).
• Result in the app: at a fact with a proposal, the line The research proposes changing a fact from the family tree (or The research proposes changing a fact entered by a person) and Conflict ›. In the conflict, Now · from the family tree (or Now · entered by a person) stands against Research proposal; the choices are Keep the family tree's and Accept the proposal. Once accepted, the fact is confirmed by a person and Originally from the family tree: … with Revert stays at the fact.
• Why: a family tree gives a direction for searching, not proof. What a person entered is their decision; the agent can only back it up with a proposal.
• Controls: app: Conflict › at the fact. Terminal: strom conflict list, strom conflict show X… (the agent's proposal and where today's value comes from), strom conflict resolve X… --take research|user.
• Since: research 1.30.0, app 3.26.0

S6.3.1 Originally from the family tree and Revert

A fact from the family tree that the research or a person changed keeps its original value. The app shows it at the fact as Originally from the family tree: with what it was changed by, and with the button Revert.

• Enforced by the code:
  • Revert writes the value from the family tree again (with a citation of the family tree). The reverted fact is then a person's fact and the research no longer changes it by itself; the record it was changed by stays with the source;
  • only a person may revert: in the app, or at their own terminal. strom refuses reverting to an agent;
  • the date, place or value of an event (each part separately) and a person's name or sex can be reverted.
• Limits:
  • Revert works only on the computer where the research runs (Reverting works while the research runs on this computer.); not on a phone;
  • only a tree linked to the research has the line Originally from the family tree and Revert; the shared facts of a couple (marriage, divorce) do not have it.
• Controls: app: the person card → Revert at the fact → Revert to the family tree's value?. Terminal: strom event revert E… [--part date|place|value], strom person revert P… --name N [--part given|surname], strom person revert P… --sex.
• Since: research 1.30.0, app 3.26.0

S6.4 The What next card

After loading a family tree from another program the app shows the card What next: where the tree comes from and what can be done with it next. It shows once after each import, whether the tree was created new, inserted or merged.

• Contents:
  • the origin: “Tree from {program} (file {date}) · N people” (for inserting and merging, into which tree and how many people);
  • the data check line: “Found N things worth a look. Nothing is corrected automatically.” and Review (A5), or “The data check found nothing to fix.”;
  • when the file holds several families with nothing connecting them: Separate disconnected parts…;
  • the research: without research on this computer Ancestor research with this tree and Ancestor research; with research Start research with this tree; a tree linked to the research Research with this tree (places without a location and people with facts from the family tree), Find places and, when the research can verify in bulk, Check facts against sources (opens Verify facts in the tree, S7.5);
  • the sentence “The research takes facts from the family tree as leads: without proof from a source it does not overwrite them, and changed ones can be reverted.”;
  • the links Back to {program}: export with sources… (S6.6) and the help A tree from another program.
• Limits: the card is only for another program's file. It replaces the earlier questions after an import (the check, separating families).
• Since: app 3.26.0

S6.5 Backups and going back in time

Before each import into a tree and before a merge the app saves a backup of the tree. The state before the import can thus be restored.

• Enforced by the code:
  • automatic backups of the tree arise daily and before an import or merge; Insert into tree takes a backup too;
  • backups live in this browser, not in the tree's file: they disappear when the browser's data are cleared and are not on another computer. A lasting copy comes from exporting the tree to a file;
  • with large backups the oldest are deleted sooner; the newest three always stay.
• Controls: app: the tree menu → Backups → the window Backup history: Create backup now, Restore, Download; the switch Automatic backups of this tree.
• Since: app 3.26.0 (the backup before Insert into tree)
• From app 3.27.0: the backup Before the data update of a new version is made before the app, on opening, adjusts a tree saved by an older version; today only to join sources that differed only in reliability (A2.6.1).

S6.6 Export back to the program, or to share

The GEDCOM export first asks What the file is for: Back to {program} (or Back to another program), or To share with others. By that it protects living people and returns program IDs.

• Back to {program}: “The whole tree with sources and citations, for one's own genealogy program.” Full data of everyone, living people included; sources and citations. People and families from that program's family tree get back their original record identifier where that is safe, and the lines of their IDs (_UID, RIN, EXID …); an identifier that cannot be used goes as REFN with TYPE {program} xref. The program's own tags (_XXX) from imports of the same program come back.
• To share with others: “Protects living people; the content can be adjusted.” (the default GEDCOM protection, A7.2.1). The deceased keep their sources; for living people the citations are left out and the window says so (“For N living people, K citations are left out. The deceased keep their sources.”).
• Enforced by the code:
  • the GEDCOM export keeps the sources and citations of deceased people in every option; the JSON, HTML and CSV exports do not change (A7.2);
  • the file has no photos and attachments: other programs do not read them from the file;
  • at Back to {program} the summary's row Sources and citations (N sources, M citations) counts the source records the file writes. The sources the import made from one source record for each page cited (A2.6) go back as that one record: two registers cited at 81 facts give “2 sources, 81 citations”. The tree keeps a source for each page cited;
  • Mark the research's facts with a note (on by default, only when the tree has facts from the research): a fact the research added or corrected gets the note “Derived by the Strom research (per: …)”;
  • otherwise program IDs are written under Strom's own tags (_STROM_IMPORT), so that they come back on the next load into Strom.
• Controls: Export → Export GEDCOM → What the file is for; from the card What next the link Back to {program}: export with sources… (opens the export with the choice back).
• Since: app 3.26.0
• From app 3.27.0: a family's photo from another program's file goes back at the family in Back to {program} as the file wrote it; a reference to a media record goes back as a photo written directly at the family (A2.2).

S6.7 A family tree loaded again and merging

A family tree of the same program loaded again and merged with the tree finds its people by program ID. A person who has the same ID but otherwise does not match is only offered.

• Enforced by the code:
  • a match by ID counts as certain only when the name is similar, the sex is not man against woman, the years of birth (otherwise of baptism) differ by at most 3 and no other global ID (for example FamilySearch) differs. Otherwise it is only a proposed match with a score of at most 79. The reason for the match is Same ID in the file;
  • _UID is paired with UID, _FSFTID with EXID with the address familysearch.org;
  • the record identifier and RIN add 25 points to a name match, only for files of the same program; by themselves they make no match;
  • an undecided match is confirmed by itself in a merge only from a score of 80; a match of 50–79 is only offered (?) and without a person's decision it is not joined (A8.3);
  • a note of a person, a relationship and a source is not doubled in a merge;
  • siblings known only by surname are not added a second time when the same file is merged again.
• A tree loaded by an older version of the app: merging the same file gives the same tree and the same Back to {program} as a new import: families (also a family with one named parent) are not doubled and no marriage the file does not have is added. A person with only a given name and no dates is paired with the person of the same name in the tree through a matched relative (the reason for example Child of matched person, Sibling of matched person, In-law of matched person); score 50, always only for review. No pairing when the relative has two people of that name, the sex differs or the dates of birth or death differ.
• Limits:
  • families are not paired by ID;
  • an ID that more than one person in the tree carries does not make a match by itself, it only offers it (A8.2);
  • the marks From family tree are not added by merging into a tree from an older version: its values stay a person's facts. A new import of the file gives the marks;
  • doubles from an earlier merge are not removed by merging; they are deleted by hand.
• A manual merge of two people (the window Merge people) joins the notes of both people and of both joined families, each part once.
• Controls: Import → Import GEDCOM → Merge with existing (A8).
• Since: app 3.26.0; 3.27.0: an ID of more people in the tree only offers; a manual merge of people joins notes (before it did not)

S6.8 A family tree in the research

The research takes over a family tree from another program with the command strom intake as leads (R11.3) and keeps the origin of each fact (S6.3).

• Enforced by the code:
  • program IDs are kept: the same family tree taken over a second time (or another program's family tree) finds the people by them, adds nobody twice, and new facts are added as leads;
  • a person who is only possibly the same (another name, another FamilySearch ID, or only an ID without _UID as with Gramps) is not joined: the result says “Maybe the same person as one of the research by an ID of the program (not joined, to be decided): …” (in --json candidates);
  • what the file has and is not taken over (media records, links to media, the program's own tags) the result counts: “Not taken from the file: …” (in --json notTaken);
  • a file in the Windows-1250 or Windows-1252 encoding (“ANSI”) is read with the right letters; a character that cannot be read is reported, never silently replaced;
  • the IDs go back with the research into the Strom app;
  • research set up earlier marks itself once after the update (which facts came from the family tree and which are a person's); no value changes. Where this is not certain, the fact counts as a person's. A value from a family tree from the Strom app is told by the research's history: it was there right after the family tree was taken in, not before, and it is the same today; only such a value is the user's family tree's (S6.3). strom fixes undo origins takes the marks back, strom fixes run origins then sets them again. Even before the marking, a family tree taken over again finds people by the IDs of earlier imports.
• Controls: terminal: strom intake rodokmen.ged, strom conflict list, strom event revert, strom person revert, strom fixes show origins (what the fix is and what it found here), strom fixes undo origins, strom fixes run origins.
• Since: research 1.30.0