Routes that are closed
A search that runs for months can reach the same dead end again and again, and each time it costs hours to find out. A running list of closed routes guards against that. When a line of enquiry stops, the list notes how far the search actually went, and the specific outside event that would justify opening it again, whether a new holding, a paid order or a visit. On the page the three kinds of "closed" look identical, so the list is useful only if it keeps them apart.
The first kind is a source gap that has been exhausted. To trace who lived at the corner from 1901 to 1940, every available address book was read. Where the street list on a page could not be read, the residents' surnames were looked up in the book's alphabetical name list, and for the years with no edition at all the telephone directory supplies occupancy evidence for 1932, 1933 and 1937. The address book also prints its own limits: whole classes of cheap dwelling were never listed in it at all. For the years when no edition was printed, there is nothing more a desk can search in that series. What would justify a return is set by the recorded retry gate, which makes no universal claim about every archive there might be.
The second kind is a route that is still open and waits only for an order or a visit. The material has been catalogued and located but not obtained, because getting it means paying for a copy or reading it on site. Eight questions carry this order-only verdict explicitly, and seven more are sorted the same way because they are marked as owner action. One of those seven, an on-site question, asks whether the works ever held a factory licence. The digitised newspapers were searched and carry no such notice for no. 169, and the permit series that would settle it has been found but never digitised . Even a reading room can answer only part of that question. The surviving series runs no later than early 1918, and the nuisance-permit records from 1918 into the 1950s have not survived, so a visit would document the works' first years and nothing after.
Three of the eight are plainer purchases. The furniture firms' trademark files are held on paper in a national archive, never digitised, and would have to be searched there in person. A 1952 overhead photograph of the block would need to be ordered as a licensed copy; the frame should look south along the 's-Gravendijkwal towards the Maastunnel, with buildings on the Rochussenstraat in the foreground. And the chain of who has held no. 169 is undocumented from about 1901 down to the block's conversion into apartments in 2009; it can be recovered only from the land registry's historical title records and the deeds behind them . None of these is a dead end. Each is a purchase, or a day in a reading room, that has not yet been spent.
The other rows are not order-only. The generated ledger sorts them into one bounded negative, one exhausted source gap, one backlog of live lookups, one waiting for the owner to clear a CAPTCHA, ten answers already held and three results the project has landed. Fifteen of the 32 open-question records still have no explicit desk verdict and are sorted by their status and channel alone, so this page must not suggest that every question has been closed or given a retry gate. The ledger directs the work, but it cannot prove that an answer does not exist.