The public set got smaller today, and that was progress

Today was one of those days. The public set did not grow. It actually got smaller by one page. That is still progress, because the page that came out no longer matched the current official structure well enough to deserve a clean public detail URL.

I want this project to get bigger, but not by pretending that old generic slugs still mean the same thing after the source system has changed underneath them. When that happens, the better move is to pull the page, explain the conflict, and rebuild the right canonicals later.

The clearest example was Provo City Down Payment Assistance Loan Program. The old public detail page was built around one generic local DPA record. The current city site no longer works that way.

Provo now presents two borrower-facing lanes instead: Home Purchase Plus for Provo buyers, and Loan to Own for a broader Utah County geography. On top of that, the same city page now says funding has been expended for all down-payment-assistance programs until July 1, 2026.

Once the official structure changed, the old generic slug stopped being honest. Leaving it live would have turned a once-correct page into a stale simplification. So it moved out of the public set and into under review.

The Fargo down-payment-assistance page is a good example. It is not a city-run DPA with its own borrower rules. It is really a referral wrapper pointing buyers into the existing statewide ND Housing DCA lane. Keeping both live as if they were separate wins would just inflate the count and confuse a buyer about where the rules actually live.

The same thing happened with Virginia's provider-network record. The DHCD provider and pilot-administrator lists are useful, but they are delivery infrastructure for the existing statewide DHCD DPA / Pilot DPA page. They are not independent benefit programs. So they went back to being duplicate wrappers instead of standing alone.

Grand Forks' new-home residential construction exemption is still real. The city still lists it, the state tax guidance still explains the structure, and the local assessing report still shows active use. That is enough to keep it in the public summary layer.

But it is still not clean detail material. The buyer-facing city page is thin, the local packet is not really there, and the current adopted local rule set is still mostly inferred from the city page plus statewide tax guidance. So the right answer was restraint, not promotion.

Part of today's work was less glamorous: making sure the publish path consumed the new under-review decision instead of silently reviving the old Provo detail from draft artifacts. That kind of artifact drift is exactly how a site starts looking more certain than the actual source layer underneath it.

I would rather catch that in the build chain than catch it a week later in a public page that should already have been retired. This is another version of the same editorial rule: if the source moved, the public layer has to move with it.

I still want the public layer to get thicker. But I want it to get thicker without turning duplicate wrappers into fake wins or letting old detail pages outlive the source structure that once justified them.

Back to blog · Browse programs · Browse states · Read methodology