The Manuals Library, Two Collections That Refuse to Guess

The Hours Lost Looking Things Up

A tech has a car on the lift and needs one number. A torque, a fluid capacity, a gap. The manual they own does not cover this model. A forum post has three answers. A video skips the step they need. They go looking, and the car sits.

An RV owner has a different version of the same problem. The rig is fifteen years old, the maker has changed hands twice, and nobody can say which recalls ever applied to it. They call a shop to ask, and the shop goes looking too.

Say a tech looks up specs five times a day and each hunt takes ten minutes. That is close to an hour a day, per tech, spent searching instead of working. And the worst case is not the lost hour. It is the wrong number that looked right.

The RV Collection

System manuals. There are 22 of them. Each one covers one system on a rig, such as the water system, the slides or the brakes. It explains how the system works, what fails, and how to check it in a sensible order. We wrote these ourselves. Nothing is copied from a maker's manual.

Model manuals. There are 200 RV model manuals. Each one is built for one model line. It pulls in the system manuals that fit that kind of rig. A travel trailer and a motorhome do not get the same set. Then it lists every public recall on file for the line, printed as filed with its campaign number, because recall records are public. Where we confirmed the maker still publishes an owner's manual, the page points to it.

The kind of rig for each line was checked by hand. Guessing it from recall text turned out to be wrong too often to trust.

The Car Collection

The car service manuals are built for the tech at the lift. Each model has one spec file, split into engine, transmission and chassis. Each block lists specs per job, steps that belong to that engine, and known faults.

Every number on the page carries a link to where it came from. Engine internal figures, like head bolt torque, need two sources that agree before they go live. If they do not agree, the figure stays out and the page says so.

Tests run over every spec file before a build. They fail the build if any of these are true:

• A number has no source link.
• An engine internal torque has fewer than two sources, or the sources disagree.
• A step mentions a number with a unit that is not backed by a sourced spec.

Both collections build as static pages, the way we describe on the static site builds page.

What Broke and What We Changed

Failed Calls Written as Clean Records

The first RV recall fetcher saved a failed call as a model with no recalls. The page looked tidy and was wrong. The fetcher now keeps every answer, paces itself, waits out a refusal, and stops if a call never comes back. It will not write no recalls for a model it never heard back about.

Dates Read the Wrong Way Round

The public recall data writes the day before the month. Read the usual way, some dates came out wrong or not at all. We parse them the way they are written.

Manuals That Said Look It Up Elsewhere

The first car manuals told the reader to pull specs from the factory manual. That is not help. It is a shrug. We rebuilt them with a spec file per model, each figure sourced, so the manual answers the question it is opened for.

A Torque Where the Sources Split

For one common engine, two good sources gave two different head bolt torques. Picking one would have been a guess. We dropped the figure. A wrong head bolt torque ruins an engine, and a missing one only sends the tech to the factory book. Any figure read from a badly scanned table is dropped the same way.

What the Library Deliberately Does Not Do

• It does not copy a maker's manual. The system writing is ours, and the recall text is public.
• It does not average two numbers. Two sources that disagree mean no number, not a middle one.
• It does not claim full coverage. A car model without a sourced spec says the spec is coming, not a guess.

These limits make the library look less complete than a site that fills every blank. That is the trade we chose. A tech who finds a blank goes and checks. A tech who finds a confident wrong number uses it.

What It Means When You Switch It On

An RV owner opens one page for their rig and sees how its systems work and which recalls apply. A tech opens one page for the car in front of them and sees the specs with their sources. When a number is missing, the page says so, and they check the factory book for that one figure instead of trusting a forum.

The hunt shrinks from minutes to seconds, and the number they use has been checked. For how this fits the rest of the shop, read about Shop Desk.