The first 30 minutes
A new account opens on a choice screen, not on somebody else’s factory data. You pick how the account starts, and you can change your mind afterwards.
- Import your Excel file. A wizard reads the file, shows you its columns, and asks which column means what. Nothing is written until you have seen a preview.
- Start from sample data. Three plants with a year of orders, so you can click through every screen before committing real numbers. A strip marks it as sample data and carries a one-click delete.
- Start empty. One empty plant. Suitable if you would rather type your first few items by hand.
Whichever you choose, the screen does not come back on the next visit. It appears only while the account still has no data of its own.
Importing your plan from Excel
The wizard has four steps: choose the file, choose the sheet, say which row holds the headers, and map columns to fields. It guesses the mapping from your header names, and you correct whatever it got wrong.
Your headers do not have to match ours. That is the whole point of the mapping step: factories call the same column EXF, Ship date, Ngày giao or 納期, and none of those is wrong.
| Field | Required | What it is |
|---|---|---|
Item | Yes | The style or item code. Everything else is grouped by it. |
Order no. | Yes | The order number this quantity belongs to. Two rows with the same item and order number count as one order. |
Quantity | Yes | Pieces to produce. Blank or zero rows are skipped, and counted for you. |
Ship date | Yes | The date the order must leave the factory. Excel dates and most written formats are accepted. |
PO no. | No | Customer purchase order, kept for reference in reports. |
PO qty | No | Ordered quantity, when it differs from what you will cut. |
Plan type | No | Free text used to group orders in reports. |
The preview before the final step is the screen that matters. It says how many rows were read, how many became orders, and exactly why each skipped row was skipped. Read it: an import that silently drops a quarter of your order book is far worse than one that refuses to run.
Items that appear in the file but are not yet configured in Settings are imported anyway and listed as a warning. They show up in reports but stay out of the recovery-plan engine until you give them a line count and a daily output.
Re-run the import whenever your file changes. The preview then shows what is new, what changed, and what has disappeared from the file — before anything is written.
Sample data, and getting rid of it
Sample data exists so you can judge the software on a full screen instead of an empty one. It is clearly marked: a strip at the top of the app says so on every screen, in every language, and it does not disappear on its own.
The strip carries a delete button. It removes every plant, item, order and production log that came with the sample, and leaves one empty plant behind. Anything you typed yourself afterwards goes with it, so the button asks you to confirm. There is no undo — but there is a backup, see below.
Accounts and roles
Roles are enforced by the server, not by hiding buttons. A viewer who edits the page in developer tools still cannot write anything.
| Role | Can do | Meant for |
|---|---|---|
viewer | Read every screen and export reports. No writes at all. | Management, customers, vendors. Unlimited and free. |
operator | Enter daily output and QC results. | Line leaders and QC staff. |
planner | Everything an operator can do, plus orders, capacity, imports and settings. | The people who own the schedule. These are the paid seats. |
admin | Everything, including account settings. | The person who signed the contract. |
Forgot a password? Use the link on the sign-in screen. The reset link is valid for 60 minutes and works once. Changing a password does not yet end sessions already open on other devices — we say so plainly rather than let you assume otherwise.
The daily loop
Once the orders are in, the routine is short:
- Enter yesterday’s output and defects, by hand or from a file.
- Open the dashboard. Orders that will miss their ship date at the current rate are flagged before they are late, not after.
- For a flagged order, open the recovery plan. It offers two options: the smallest change that saves the nearest missed date, and the change that saves the whole order book. Both are expressed in lines and days, because that is what a factory can act on.
- Apply the one you accept. It becomes the new capacity plan, and every projection updates with it.
Your data and backups
Each customer’s data lives in its own store, keyed by account. No screen, endpoint or report can read across accounts: the account identifier is required by every storage call, so it cannot be left out by accident.
Backups run automatically every day and are kept for 30 days. A backup contains both your production data and the accounts — a restore that brings back the data but not the logins is not a restore.
You can also take your data out at any time without asking us: every table exports to Excel, and Settings has a full export of the whole account. Being able to leave is part of being safe to buy.
Service status and monitoring
The service exposes a public health endpoint. If you run your own monitoring, point it here: it answers in a few milliseconds and touches no database, so polling it every minute costs nothing.
curl https://lineprotrack.com/api/healthA healthy answer looks like this:
{"ok":true,"env":"production","version":"0.1.0","uptimeSec":84213}A second endpoint, /api/health/deep, also checks the database, the disk, how fresh the last backup is and whether mail can be sent, and answers HTTP 503 when something essential is broken. It reports the details only to a caller holding the health token set up by the operator of the instance.
Troubleshooting
The import says a column is missing, but it is right there.
Check the header row number in step three of the wizard. Files exported from ERP systems often carry a title and a blank line above the real headers.
Dates came in wrong, or rows were skipped for an unreadable date.
A date stored as text is ambiguous: 03/04 is two different days depending on who typed it. Format the column as a date in Excel before exporting and the ambiguity disappears.
A number was saved, but another user does not see it.
Two people edited the same account at once. The second save is refused rather than allowed to overwrite, and the app offers to load the newer version. Nothing is lost: unsaved work stays in the browser until it is either saved or discarded.
The app keeps asking me to sign in again.
Sessions refresh silently in the background and last seven days. Being asked again usually means the browser is blocking cookies for the site, or that you signed out on another device.
A screen is blank after an update.
Reload once. If it persists, tell us and include the time — browser errors are recorded on our own servers, so we can usually find yours without asking you to reproduce it.
Anything not covered here: lineprotrack@gmail.com.