Opening a file without the software that made it

Find out what the file actually is before doing anything to it, because the name on it is a suggestion and its first few bytes are the fact. Once you know, almost every file in this situation is a published format that a browser can read on its own: a Word or Excel file is a zip of XML, a HEIC is a photograph in the same kind of box as an MP4 video, and a .msg is a small file system. The guides below take each one in turn. None of them needs the file uploaded, and each says plainly what does not come across.

Why a file will not open

There are three reasons, and they need different answers. The first is a missing reader. A HEIC photograph is compressed with HEVC, a video codec covered by patents, and a program either pays for a licence to decode it or ships without one. Apple ships one, so an iPhone photo opens on a Mac; Windows opens it only with an extension from the Microsoft Store, and Chrome and Firefox do not decode it at all. Nothing is wrong with the photograph. Opening a HEIC file covers what is in the box and how to get a JPEG out of it.

The second is a format that only one program writes. Outlook saves an email dragged to the desktop as a .msg, which uses the same compound file format as an old Word .doc: the subject, the sender, the body and each attachment are separate streams inside it, which is why a text editor shows noise. When Outlook sends a message in its rich-text mode to somebody outside the organisation, the attachments arrive packed inside a winmail.dat, in a format Microsoft calls TNEF, and every other mail program shows that one file instead of what is in it. Both formats are documented, and SATCHEL reads both in the tab. The .msg guide and the winmail.dat guide go through them, including the setting the sender changes so the next email arrives normally.

The third is a refusal that looks like damage. Windows File Explorer can decrypt only the zip encryption scheme from 1989, which was broken in 1994, and not the AES scheme most archivers now write, so a perfectly good encrypted zip is reported as invalid or asks for the password again and again. Separately, Explorer stops with error 0x80010135 when a path stored inside a zip, added to the folder you are extracting into, runs past 260 characters. Neither says what it really means. The AES zip guide and the long path guide explain each and how to get the files out.

Look at the first bytes

Every common format announces itself at the start of the file, and the signature cannot be changed by renaming it. A zip, and therefore every .docx, .xlsx and .pptx, begins with the letters PK. An old .doc, .xls or a .msg begins with the bytes D0 CF 11 E0. A PDF begins with %PDF-, an RTF file with {\rtf1, and a HEIC has ftyp followed by heic a few bytes in. A file called report.docx that begins with {\rtf1 is a rich-text file somebody renamed, and repairing it as a Word document will get nowhere. X-RAY reads the signature and says what a file is, whatever it is called. Why a .docx opens as XML uses exactly this to tell four different problems apart.

When it opens, but wrong

Some files open and then mislead you. A CSV records nothing about which character separates its columns, so Excel does not look: it uses the list separator from the Windows regional settings, and a file written with semicolons lands in one column on a machine that expects commas. The CSV guide shows the fix for one file and for every file, and the sep= line that ends the argument.

A Word document contains no pages. Every machine that opens it lays the pages out again, and a font that is missing and substituted moves one line, which moves every page break after it. A page number in Word is a fact about the machine that printed it, which is why page numbers change on another computer and what to cite instead. Printing adds its own quiet change: a print dialogue set to fit the page shrinks everything by a few per cent, A4 artwork fitted onto Letter comes out at 94 per cent, and nothing on the paper says so. Why a print comes out smaller gives the order to check the settings in, and printing a booklet covers the one print job with a rule of its own.

Why not a converter website

An online converter does not have a secret reader. It runs the same kind of zip library, XML parser or HEVC decoder on its own server, which is why it needs the file sent to it. What is sent is the whole file, with everything in it: the GPS position in the photograph, the author and comments in the document, the internet headers in a .msg, which record the internal servers and addresses the email passed through. It is then kept for as long as that site's policy says, somewhere you cannot check. The readers used here run inside the tab instead. The HEIC decoder, for example, is libheif compiled to WebAssembly, about 1.4 MB, fetched from this site the first time you open a HEIC and kept on your device after that.

What does not come across

Reading a format is not the same as reproducing the program that wrote it, and each guide lists what is lost. Opened in Docs, a .docx keeps its headings, lists, tables, links, footnotes and pictures, but not its tracked changes, comments, fonts, columns or page layout. Opened in Sheets, an .xlsx keeps every sheet with its values, formulas and number formats, but not its charts, pivot tables or macros, and a password-protected workbook cannot be read without the password. A ten-bit HEIC decodes to the eight bits a JPEG holds. An email encrypted with S/MIME can only be opened with the recipient's private key, which lives in their own mail program. Where the loss matters, keep the original and change it where it was made.

Docs and Sheets are part of the Workspace and are free. The single tools named on this page are free to open a file and see what is in it; saving what they make is part of Pro, except where a tool is only handing back your own file, such as taking files out of a zip in STRONGBOX.