Part V · THE PEOPLE WHO KEPT THIS · Chapter 20

Let's Keep Each Other Posted

Portland, Oregon · a server, and a box in the cloud · 2000–2026

In the record Al Krewson (1943–2006) · Michael V. Krewson (b. 1971) · Elizabeth Cregier · and everyone in the record

"I went on a search for my ties starting the later 1960s and had some pretty good luck," Al Krewson wrote to a stranger named Marcia Shaw on 20 January 2005, copying his nephew [1]. That is the whole of his method as he described it: a letter to a cousin, a bus to Hammond, a night at Aunt Anna's, a trip to Centerville, a picture recognised, a phone call that went wrong, and thirty-five years of keeping what he had been told. He had no Cruise; he thought the family had come "with the Pennsylvania Dutch" in the 1500s and never learned otherwise [2]. He reversed his grandfather's given names — "Franklyn Charles" for Charles Franklyn — and got his own father's birthplace wrong until a cousin corrected him with a question [1] [3]. And he was, by the standard this book has held every document to, an honest witness: he separated what he remembered from what he was told, named who told him, said when two accounts disagreed, and refused to judge the man who could not answer. "I have the facts, he has his notions," he wrote of the Cherokee argument with his brother, and then: "It's not worth it to me to argue with him over it … We'll leave it there" [4]. Every letter ended the same way, and the last chapter takes its title from it.

What the record is

On 20 January 2005 Michael asked for "a web based visual presentation of what he/she knows so that my son's generation and his son's generation can add on to it" [5]. Twenty-one years later it exists, and this chapter describes it plainly, because a record that is not described is a record that will be lost the way Raike's seven volumes were.

It is a private database on a server Michael runs, called — in the family's habit of naming its tools — MiSTeRGenealogy. On the day this chapter was written it held 182 people, 36 family stories, 367 files (327 photographs, four birth certificates, four marriage certificates, ten DNA files), and six DNA matches [6]. Its cards are the ones this book has quoted: Galen's 1943 certificate with "KREWISON" misspelled on it; Margery's GRO certificate from Purley; the c. 1890 photograph with Earl's ten names on the back; Al's emails, every one; the Family Finder match to Annette Maichen at 466 centimorgans that proved the Michiela line; the myOrigins panels that found no Cherokee. Beside it, in a private store in the cloud, is the whole of Cruise's Volume I — every page scanned, and 282 files of machine-read text — from which every will and deed in this book was quoted by page and, when the reading was doubtful, checked against the scan [7]. The record has a rule, printed on the family's own pages before this book adopted it: documented, inferred, or legend, and say which. It has thirty-three dated backup copies. And it has, on thirty-four of its cards, a small blue book — 📘 — marking what this book found and wrote back into it, so that the record and the book cannot drift apart [6].

The day it lost twenty-four stories

It is also fragile, and the book will not pretend otherwise. On 11 September 2026 — the first day of writing — the record's family stories were counted and there were eight. There had been thirty-two. A week earlier, a routine save from a page that had been open too long had posted the eight stories it knew about over the thirty-two that existed, and the software, which replaced each list wholesale, had let it. Twenty-four write-ups of the Croesen book — the Tree Thief, the Voorlezer's house, Lydia, the burned will — were gone, and no one had noticed [8]. They were recovered the same day from the last full backup, whose eight shared stories were byte-identical to the current ones, so the merge was clean; and the software was changed that evening so that a save which would shrink a list by more than one item is refused unless the person saving says, explicitly, that they mean to delete. Seven tests pin the rule. The lesson was written into the record in the words the family's own history had taught: absence in a write must never destroy presence in the record. A Bucks County son burned a will in 1758; a Dutch family tree went missing in the East Indies in 1897; a web page overwrote twenty-four stories in 2026. The keeper's job is the same in every century, and the book's first act was to do it.

What the book found

Because the record was kept, the book could find things, and because the book found things, the record grew. The short list, for whoever comes next. That both of Elizabeth Cregier's grandfathers signed the surrender of New Amsterdam in 1664, and that she is the last common ancestor of Michael and a four-star general [9]. That the family's first named enslaved person was Fortune, in 1729, and that Jude, Striffin and Lydia were to buy their freedom under a will of 1783 — four people the record had not held [10]. That the Labadists walked across Staten Island in October 1679, not June 1676. That "Krewson" appears on page 229 of Cruise and the OCR was right. That Earl W. Krewson of Portland, who wrote the names on the photograph, was born in 1905, was a sergeant in the Second World War, and is the same Earl who took Al in. That Al was a seaman in the Navy, Vietnam era, and died on 24 January 2006, a year and three days after the letters. That the Berkeley grandparents were Margery's father, not Félice's parents; that 1948 was a bell-tent at Lancaster; that the Irish estate went to cousins by the operation of a law repealed three years later; that the Military Cross could not have been won in the Boer War; that Frank King Swain lived in Henry Mercer's castle. And, from Michael's own memory, that Al could play the piano, and tried to teach him.

Then, on the evening the last chapter was published, the record grew by a third in one night, and the book had to go back into three chapters. A relative on the Knight side sent a typed letter of 1919 and an obituary of 1973; Michael exported the Ancestry tree he had been keeping for years, and then its gallery. The short list of what that night found: that Granny's people were Cuban, Asturian and Maltese, that her father died in Paris when she was two, and that the "Spanish royalty" had a documented house behind it (chapter 14). That Grandpa Marshall was a Rhodes Scholar, an Old Contemptible of 1914 and a Gallipoli captain, that he worked with Ricardo on the octane number, and that he and his second wife raised Margery in California from the age of six (chapter 13). That Galen's mother was Charlotte "Lottie" Hall, that his two brothers died in 1913, that he was wounded in France, that he lived as "Bill Small," and that he died in a federal prison hospital after a mail-fraud conviction the family never knew of (chapters 12 and 13). That Herbert Mitchell was born about 1895 and the Boer War was never his (chapter 14). Seventy-five people, five stories and sixty-seven documents were added to the record between the two evenings, and the counts in the paragraphs above are left as they stood, dated, because that is what the record's own rule requires [12].

What is still open

The book ends with its debts listed, because a list is the only thing that gets paid.

1. Lewis Krewson's Civil War. Eighteen in April 1861, in Ohio; the record's largest single blank. The 1850–1880 censuses, the Ohio and Iowa rosters, the 1890 veterans' schedule and the pension index have never been searched.
2. Charles Franklyn Krewson's marriages, and Galen's mother. Settled, 11 September 2026: Charlotte "Lottie" Melissa Hall, married 25 December 1895 (chapter 12, postscript). What is open now is the "L. Marshall, 1888–1956" of Annette's chart — a misreading, or a second wife after 1917.
3. Félice's mother, and Brussels. Henry Faustin Knight died in Paris in 1891; who he married, and why the child was born in Brussels in 1889, no document yet says. The Paris civil register that recorded his death is the lead.
4. Galen's death certificate, read in full. The tree has it (Missouri, 1957); the institution on it would confirm or refute the book's inference of the Medical Center for Federal Prisoners. And the 1956 case file, United States v. Krewson, Southern District of California, is at the National Archives at Riverside.
5. Who sent the Knight papers. The relative on Facebook whose tree is in Spanish is not yet named in the record.
6. Simon Krewson's Revolution. Born at Bustleton in 1755; no service, no pension, no roster yet.
7. Lydia. The Bucks County slave register of 1780, the Pennsylvania Abolition Society's manumission books, the 1790 census — whether she paid her thirty pounds.
8. Anne Raike's seven volumes, Katherine Weaver's papers, and Warren Cruise's own. Last known in Denver in 1998. The 1700s that exist nowhere else are in them.
9. Herbert Mitchell. Now also 1962 — the tree's date. The Irish Calendar of Wills and Administrations for 1962–64 (a letter to query@nationalarchives.ie); a London Gazette search for the Military Cross; and the video of Glenn saying the castle's name, which no one has transcribed.
10. Marjorie. Her grave — most likely Lincoln Memorial Park — and her service record; which branch sent the honour guard.
11. Three SF-180s — Al, Marjorie, Earl — drafted and not yet mailed to St. Louis [11].
12. Gary A. Krewson (1943–2024) and Leonard Miles Krewson (1923–2012), two Oregon veterans in the government's list, relation unknown.
13. The Indian relative. Al: "I do have pictures of this Indian relative somewhere" — Blackfoot or Crow, he thought, not Cherokee; the DNA says whoever she was, she is not in the direct line. The pictures, if his children have them.
14. Garret's Holland. Winschoten, in Groningen, is the family's best guess and nobody's proof; the general looked with an army and found nothing before 1650.
15. Volume II. Never written; but the box it would have been made from is item 5.

The keepers

Katherine Breck Weaver, from 1915. Frank King Swain, from the 1920s, in his castle. Anne Raike, from 1932. A Mr. Stols in The Hague in 1943. Marie Metzgar, fifty years. Wayne Nelson. Shirley Cruzen Stevens. William Bogardus. General Kroesen, from 1960. Warren Cruise, twenty years, and his mother before him. Earl, with his ballpoint pen on the back of a photograph. Al, from the late 1960s to 2005. Michael, from 2000, when he first wrote to his uncle, to the server he keeps now. Not one of them was paid; not one of them was, by Al's diagnosis, part of a family that "built up what other families did." What they built was this — a record that a Los Angeles child of 1946, a sergeant of 1944, a Bucks County lieutenant of 1776 and a girl at a font in 1662 all stand in together, on the same page, with the fences between them written down and therefore, for the first time, visible.

The title of the book is a distance: the eight miles of water between Brooklyn and Staten Island that the family crossed first, in the 1680s, because the land was cheaper on the other side. Every generation after crossed something — the Delaware, the Alleghenies, the Mississippi, the Rockies, an ocean back the other way, a wall, a war — and left less behind than it should have, and someone came after and picked up what was left. This book is one more of those. It is not finished, because the list above is not finished, and the record it was written from was built to be added to. That was the request in 2005, and this is the reply, in the words the family's genealogist used to close every letter he wrote:

1. Al Krewson to Marcia Shaw, cc. Michael, 20 January 2005, 9:20 p.m. (info-on-krewson-fathers-names-and-iowa-2005-01-20.htm).
2. Al Krewson to Michael, 26 June 2002, 1:52 p.m. (family-history-uncle-al-2002-06-26.htm): "They came here in the mid to late 1500s with the Pennsylvania Dutch (Deutsch) posing as Germans."
3. Al Krewson to Michael, 21 January 2005, 11:01 a.m.: "Cincinnati, Ohio … could it have been Cincinnati, Iowa." See chapter 17.
4. Al Krewson to Michael, 21 January 2005, 11:55 a.m.: the Indian legend, "I have the facts, he has his notions."
5. Michael Krewson to Al Krewson, 20 January 2005, 8:51 p.m., and Al's reply, 11:40 p.m. (more-rifts-in-the-family-2005-01-20.htm). See chapter 17.
6. MiSTeRGenealogy, the family's private record on Michael's server, counted 12 September 2026: 182 people; 36 family stories; 367 files (327 photographs, 21 other documents, 4 marriage certificates, 4 birth certificates, 10 DNA files, 1 census/immigration record); 6 DNA matches; 34 person-cards carrying the book's 📘 write-back tag; 33 dated backup copies. The documented / inferred / legend rule is the record's own.
7. The family's private archive store: the complete scan and OCR of Cruise, Vol. I (282 text files, CROESEN-FAMILIES-OF-AMERICA-P###-###-*.TXT, with page images), plus the correspondence and photograph files cited throughout this book by filename.
8. The record's own log for 11 September 2026: 32 family stories reduced to 8 by a save on 4 September 2026; 24 restored from the backup of 22 August 2026; the write guard added the same day to the record's software (a list may not shrink by more than one item without an explicit override; seven tests), documented in the record-keeping software's own notes. Chapter 18 tells the 1758 and 1897 losses.
9. Chapters 1 and 19 and their sources (Cruise p. 213; Cruise p. ix).
10. Chapters 7 and 9 and their sources (Cruise pp. 191–193, 301–305).
11. Standard Form 180 worksheets for Alfred G. M. Krewson, Marjorie Krewson and Earl W. Krewson, prepared 12 September 2026 for the National Personnel Records Center, St. Louis; see chapter 16, note 4.
12. The record's own log for 11–12 September 2026: 182 → 257 people, 36 → 41 stories, 367 → 407 files, across five write-backs (the Knight papers; the GEDCOM; the Ancestry gallery in three parts), each with a dated backup; the Knight/López Calleja line added as a fourth branch on the Branches tab. See chapters 12, 13 and 14 and their sources for every document.