The message lives in the code, not on a web page.
Most codes send a phone off to a web page. This one does not. The words you type are stored in the pattern itself, so the phone shows them straight away, with no signal, no page load and no chance of the message going missing when somebody rebuilds the website.
Tip: Ctrl+C or “Right click\Copy”
Displaying the logo on the QR code may take some time. Just wait for it to load.
When some colors are combined on the QR code, it may not work correctly. Don’t
forget to check before posting!
There is no destination here and no redirect. The characters you type are encoded directly, and a scanner reads them back as text. On an iPhone running iOS 11 or later the Camera app shows the words in a banner and offers to copy or search them. Android 9 and later does much the same in its scan sheet, with a copy button. Nothing is fetched, nothing is logged, and the code reads identically in a basement with no bars on the signal meter. The trade-off is that text is inert: a phone number written as text will not dial, and a web address written as text usually has to be copied by hand.
Long text is the main enemy. Every character adds modules, and a dense grid printed small fails on a phone held at a normal distance. The usual print rules matter here too: leave the four-module quiet zone alone, keep dark modules on a light background because inverted codes fail on older readers, and take the SVG or EPS rather than a PNG stretched to fit. Accented letters and emoji cost several times what a plain letter costs, and some older scanner apps show them as boxes, so a label meant for strangers is better off with plain characters. The text cannot be edited once it is printed, and nothing counts the scans. The general guide covers the rest of the print side.
The question is what the code has to survive. A link is a dependency: on signal in the room, on the page still existing, on whoever maintains that site. Text has none of that, which is why it wins in the plant room, on the cold store shelf and on an equipment plate screwed to a machine that will still be running in 2040. It wins on a museum caption in a building with thick walls and no public Wi-Fi, and on a lost-property tag, where the finder is standing in a car park and should not have to trust an unknown web address. It also wins on an allergen note, since nobody wants a page load between a customer and an answer about nuts. A link is the better choice the moment the information changes, or when it runs to more than a paragraph, and then a plain link code is the type you want.
On length, the technical ceiling is 4296 alphanumeric characters, and that number is useless as guidance. Something under 100 characters gives a sparse pattern that reads at 2 cm from arm's length without anyone shuffling forward. Around 300 characters, roughly a short paragraph, the grid is nearing 60 squares a side and wants 4 to 5 cm in print, or a phone brought close. Past about 500 characters you are printing something that has to be held at 10 cm and lit well, which defeats a shelf label. Treat 300 as the working limit for anything printed, and if the text is longer than that, it wanted to be a page all along.
No, at neither end. The characters are encoded in the pattern, so a scanner reads the words off the ink and shows them with mobile data off and no Wi-Fi in range. That is the reason these turn up in warehouses, plant rooms and cold stores where a link would be useless.
A QR code can hold up to 4296 alphanumeric characters, but that is a theoretical limit rather than advice. Keep printed labels under about 300 characters. Beyond that the grid gets dense enough that a phone has to be brought close and the light has to be good.
No. The text is the code, so changing a word means a different pattern and a reprint. If the wording will change, point a code at a page instead and edit the page. For a label that should never change, such as a serial number, that permanence is the point.
They are encoded correctly, and most current phones display them properly. Older scanner apps sometimes show boxes or question marks instead. Each of those characters also takes several times the space of a plain letter, which makes the code denser. On a public label, plain characters are the safer choice.
Not reliably. Text is inert, so a number in it is just characters, and many scanners show it without offering to call. Some phones detect a number and make it tappable, some do not. For a number that dials on every phone, use the phone code type rather than plain text.
No. Nothing is requested from a server when somebody scans it, so there is no scan to record. Counting is only possible on the editable link code, which sends the phone through a redirect on this site and reports totals with the countries they came from.
