How we verify re-sign rules
Accuracy is not a feature of this project. It is the entire product. A wrong date here does not cost somebody a click, it costs them their position on a list they waited months to climb. So here is exactly how the dataset is built, including the parts that are weak.
Where the data comes from
Every field on a local's page is read from that local's own published referral or hiring-hall page, and we link to that page so you can check our work. We do not scrape aggregators, we do not copy other tools, and we do not take a re-sign window from a forum post or a screenshot.
When a local publishes its rule as a sentence, we quote the sentence verbatim on the page. You should be able to compare what we say against what they say without leaving the tab.
What "last verified" means
It means a person opened that local's referral page on that date and confirmed every non-empty field against it. It does not mean the local has not changed something since. Halls change procedure without announcing it, which is the honest limit of any dataset like this.
Treat a last-verified date the way you would treat a wage sheet: useful, current as of a date, and worth a phone call before you bet your position on it.
What we refuse to publish
This is the part that matters most.
- We never fill a gap with a guess. If a local does not publish its window, the page says "not published" and tells you to call. It would be trivial to assume the common 10th-through-16th window and be right most of the time. Being right most of the time is exactly how somebody gets dropped.
- We do not infer a city or a phone number from an area code or a domain name. Blank beats plausible.
- We do not publish a local at all without a link to its own referral page. No source, no page.
- We do not soften a local's language. Local 106 says there are no exceptions to its receive-by rule. We print that, because a traveler reading a friendlier paraphrase might mail it on the 15th.
Why the exceptions get emphasis
Most locals verified so far run a 10th-through-16th monthly window. That consistency is the trap. Once a traveler has signed at three locals with the same window, they stop checking, and then they sign at a local like 113 in Colorado Springs, which re-signs quarterly with a fifteen-day grace period.
So per-local pages lead with what is different about that local, not with what is normal. The normal parts you will already know. The different part is the one that drops you.
How corrections work
If something here is wrong, send it to hello@dontdropbook.com. What happens next:
- We read your message the same day.
- We re-check the local's own page against what you told us.
- We correct the page, update the last-verified date, and reply to tell you what changed.
- If you want credit on the page for the catch, say so and you get it.
Members catching our mistakes is not a support burden. It is the verification layer, and it is the reason this can stay accurate at a scale one person could not check alone.
What we will not claim
We will not claim this dataset is complete, and we will not claim it is a substitute for your hall. It is a second pair of eyes on a deadline that is easy to lose track of when you are carrying several books and working out of town.
If DontDropBook is ever the only reason you believe your window is a certain week, call the hall.
DontDropBook is an independent project, not affiliated with, endorsed by, or connected to the IBEW or any of its local unions.