TAR Creator

How to use the tar creator

1. Choose the files to include; folder structure is kept where the browser reports it.
2. Create the archive and check anything reported as left out.
3. Download the .tar, and gzip it separately if you want compression.

About this tool

TAR is the oldest archive format still in daily use, and its design shows why it lasted: a 512-byte header, the file's bytes, padding to the next 512-byte boundary, repeat. There is no index and no central directory, so an archive can be written to a tape or a pipe without knowing in advance what will go into it.

This creator writes ustar archives, the POSIX form that every tar implementation reads. Each header carries the name, mode, size, modification time and a checksum, and the checksum is computed the way the standard describes — over the header with the checksum field itself read as spaces, which is the detail a naive implementation gets wrong.

The format's real constraint is the name field: 100 bytes, extended by a 155-byte prefix that must break on a path separator. A long path is split across the two, but a single component longer than 100 bytes cannot be expressed at all. That entry is reported and left out rather than truncated, because a silently shortened filename is a file you cannot find later.

A TAR is not compressed. It concatenates files, each behind a 512-byte header, which is why an archive is slightly larger than the files inside it. Compression is a separate step — conventionally gzip, which is where `.tar.gz` comes from — and separating the two is what lets tar stream to a tape or a pipe without knowing in advance how large the result will be.

Common uses

• Bundling several files into one download for a colleague.
• Producing a .tar for a system that expects that format rather than ZIP.
• Packaging files without a compression step, when they are already compressed.

Frequently asked questions

Why is the .tar larger than the files it contains?

Because TAR does not compress. Each file gets a 512-byte header and is padded up to the next 512-byte boundary, so an archive of many small files carries noticeable overhead.

Can it make a .tar.gz?

Not in one step. Create the .tar here and compress it separately — gzip is a distinct stage in the pipeline, which is exactly what the two extensions in .tar.gz record.

What happens to a very long file path?

It is split across the header's 100-byte name and 155-byte prefix fields. A single path component longer than 100 bytes cannot be represented, so that entry is reported and left out rather than truncated.