PDF - Portable Document Format
PDF is the format that makes a document look the same everywhere. It carries pages, fonts, images and a surprising amount of information about who made it and with what.
PDF stands for Portable Document Format. Adobe created it so that a document would arrive looking exactly as its author intended, on any computer, with any printer, whether or not the reader owned the fonts it was set in. That promise is the reason PDF became the format for contracts, invoices, manuals, forms, tickets and scientific papers. Adobe handed the specification to the International Organization for Standardization in 2008, so PDF is now an open standard rather than one company's format.
How a PDF stores a page
A PDF is not a picture of a page and it is not a word processor document. It is a collection of numbered objects — pages, fonts, images, colour definitions, annotations — plus a table at the end of the file that says where each object begins. A page object holds a list of drawing instructions: put this glyph at these coordinates, fill this rectangle with that colour, place this image in that box.
Because the instructions are geometric rather than pixel-based, text in a PDF stays sharp at any zoom and can usually be selected and searched. The exception is a scanned PDF, which is a photograph of paper wrapped in a page object — there is no text in it at all until someone runs optical character recognition over it.
Fonts are normally embedded, meaning a copy of the typeface travels inside the file. That is what keeps the layout identical on a machine that has never seen the font, and it is also why a PDF of a one-page letter can be several megabytes. A PDF may compress its objects individually, and from PDF 1.5 onward it can pack many small objects into a single compressed stream, which is why two files with identical content can differ greatly in size.
What metadata a PDF carries
PDF carries more metadata than most people expect, in two places at once.
The first is the document information dictionary, the classic set of fields a reader shows under document properties. You will see Title, Author, Subject, Keywords, Creator — the application the document was written in — and Producer, the software that actually wrote the PDF. Alongside them sit CreationDate and ModifyDate, typically stamped in the form D:20240418103122+02'00', which encodes the time zone as well as the moment.
The second is XMP, Adobe's XML-based metadata packet. XMP can repeat the same fields in a structured form, and it is where a document declares things like its PDF/A conformance level, a persistent document identifier that survives editing, and a history of which applications touched it.
Two of these fields are worth a second look before sharing a file. Author is often filled in automatically from the account name of whoever was signed in to the computer, and Producer names your software down to the version. A PDF exported from a phone or a camera app can also inherit metadata from the images inside it. None of that is visible on the page.
The structure itself reports on the document too: the PDF version that tells a reader which features to expect, the page count, the page size in points, the page rotation, whether the file is tagged for screen readers, and whether it is encrypted or digitally signed.
When to use PDF
Use PDF when the layout matters and the document is finished. It is the right answer for anything that will be printed, signed, filed or read by someone you cannot ask to install software.
It is the wrong answer while you are still writing. PDF is awkward to edit — you are moving objects on a page, not reflowing a paragraph — so keep the editable original in a word processor format and export to PDF at the end. It also reads badly on a phone, because a fixed page cannot reflow to a narrow screen; an ebook format handles that far better. For long-term archiving, the PDF/A profile exists precisely to forbid the features that make an ordinary PDF risky to open in twenty years.