Archive

In short: One or more files combined into a single file — usually additionally compressed (see compressing).

In more detail: Formats like ZIP or 7-Zip bundle many files into one archive, which simplifies transport and storage. Not to be confused with archiving — the longer-term, organised keeping of (often archived) data.

In Depth

Bundling vs. compressing

An archive format can do two different things that are often combined but are technically separate steps: bundling (combining several files/folders into a single file, preserving the folder structure, permissions and timestamps) and compressing (reducing file size, see Compressing). The classic Unix tool tar (“Tape Archive”, originally developed for magnetic tapes), for example, only bundles, without compressing itself — hence the common combination .tar.gz, where tar bundles and gzip then compresses:

# Create an archive (c=create, z=gzip, f=filename)
tar -czf project.tar.gz ./project/
 
# Extract the archive again (x=extract)
tar -xzf project.tar.gz

ZIP and 7-Zip do both in one step (bundle AND compress), which makes them more convenient for Windows users, but less flexible than the Unix-style two-step separation — with tar, for example, the same bundle can be compressed with gzip, bzip2 or xz, depending on the desired trade-off between speed and compression ratio.

Practical use

Archives play a central role in backups and software distribution: instead of transferring hundreds of individual files, a single archive file is copied, which makes the transfer more robust (an interruption doesn’t affect individual files halfway through, but can be repeated as a whole) and also consistently preserves metadata like file permissions and timestamps — important, for example, on Unix systems, where individual files’ execute permissions could otherwise be lost during copying. Software packages (e.g. .jar files for Java, .whl files for Python) are also, at their core, nothing other than ZIP archives with a specific internal structure and a different file extension.

Distinction from archiving

Not to be confused with archiving — the longer-term, organised keeping of (often archived) data, e.g. for legal retention requirements. An archive in the technical sense (the file) can be part of an archiving strategy (the process), but isn’t the same thing — you can also archive files without ever creating a .zip/.tar (e.g. simply storing them in a separate, read-only folder).

Integrity checking

Many archive formats also store checksums (e.g. CRC32) for each contained file — when extracting, it’s automatically checked whether the extracted data still matches the checksum, which reliably detects transmission errors or corrupted downloads before you continue working with faulty files. For very important archives (e.g. backups), it’s also worth computing an external checksum of the entire archive file itself (e.g. via sha256sum), to also detect unnoticed corruption of the archive as a whole.

See also: Archiving, ZIP, 7-Zip