Creating an archive
The create dialog: sources, name, compression, password and extra options.
Starting
| Way | What happens |
|---|---|
| Drag a file or folder into the window | The dialog opens with that source already filled in |
| File → Create Archive… (Ctrl+N) | Empty dialog; you pick the sources |
| Right-click a file or folder → Create Archive… | The selection becomes the source |
All three work in both profiles: the menu is the map of what the program can do, not a reflection of the current focus. See Two profiles.
Select Sources… opens a chooser with the file system on the left and your list on the right. There may be several sources. Delete removes the selected source from the list.
General
Archive file
The File archive: field plus an Insert date button that puts the current date and time into the file name. Name templates:
| Template | Substitutes |
|---|---|
%year, %month, %day | Year, month, day |
%hour, %min, %sec | Hour, minute, second |
%date | The whole date |
%time | The whole time |
What happens if the file exists
The program does not overwrite the archive, it adds to it: a new version is created. This is stated under the file name field. If that is not what you want, give a different name or delete the old archive.
Compression level
| Level | Meaning |
|---|---|
| m0 | No compression |
| m1 | Fast |
| m2 | Standard, recommended |
| m3 | High |
| m4 | Very high |
| m5 | Extreme |
The higher the level, the smaller the archive and the longer it takes. m2 is a sensible compromise for most cases.
File hash
The algorithm the engine uses to checksum file names and contents.
The default is xxhash: fast and good enough. Others include
xxh3, sha256, sha3,
blake3, sha1, crc32,
md5, windate, whirlpool, the
variants suffixed b (compatible with other zpaq
implementations), and the highway family.
Changing the algorithm makes an existing archive incompatible with the previous way of checking it, so change it deliberately and before you start working with the archive.
Password
Two fields: the password and its repetition. Creation will not start until the second matches the first — that guards against a typo that would make the archive impossible to open.
The password encrypts the archive contents. An encrypted archive is visible in the tree, but its contents cannot be read without the password: the program asks when you open or extract it.
A password cannot be recovered
If the password is forgotten, the contents cannot be recovered by any means. Keep the password stored separately from the archive.
Comment
Free text stored in the archive with the version and shown in the version list. Useful for noting what changed.
Include and exclude masks
| Field | Meaning |
|---|---|
| Include only | Only files matching the masks go into the archive; the rest are skipped |
| Exclude | Everything goes in except files matching the masks |
Several masks are separated by a semicolon: *.txt; *.doc.
Handy for keeping temporary and working files out of the archive.
Post-creation check
| Option | What it does |
|---|---|
| No check | Checks nothing; creation finishes immediately |
| Test (full) | Full integrity test of the new archive |
| Verify against filesystem | Compares the archive contents with the original files |
| Paranoid | The strictest check, and the slowest |
Checking after creation is more reliable than assuming everything went well: it catches damage while the archive is still fresh.
Compression
This tab fine-tunes the method. For ordinary use the level on the General tab is enough — these settings are best left alone unless you have a specific reason.
| Field | What it sets |
|---|---|
| Context models | Which models predict the content while compressing: MATCH (a), MIX2 (t), ISSE (i), SSE (s), Context (c), Words (w), Mixing (m) |
| Mode | Journaling with deduplication, or streaming without it |
| Preprocessing | Transformation of the data before compression |
| Block size | Size of the block being processed |
| Search depth | How hard to look for repeats |
The bottom shows what the chosen combination amounts to. A deeper search and more models give better compression and noticeably more running time.
Advanced
Working with files
| Option | Why |
|---|---|
| Support long paths (>255) | Paths deeper than the standard Windows limit |
| Store file creation date | Restores the original dates on extraction |
| Do not recurse into subfolders | Archive only what sits directly in the source |
| Store NTFS ADS | Alternate data streams (Zone.Identifier and such) |
| Skip OneDrive placeholders | Do not archive the stub files OneDrive keeps on disk |
| Ignore file access errors | Carry on when a file cannot be read instead of stopping |
Archive contents
| Option | Why |
|---|---|
| Do not force re-adding XLS/PPT | Do not add again what has not changed since last time |
| Disable extra checksums | Faster, but less reliable |
| Disable deduplication | Skip the search for repeats: faster, but worse |
Destination, copies, media
| Option | Why |
|---|---|
| Split into volumes, size | Multi-volume archive. Give a size such as 1GB |
| Check free space | The program checks that there is enough room before starting |
| Second copy to folder | Save a copy of the archive elsewhere at once, e.g. a network drive |
| Create self-extracting EXE | The archive as an executable file |
| Use Volume Shadow Copies (VSS) | Lets you archive files that are in use, e.g. a document open in Word |
| Shut down PC on completion | For a night task: power off when the work is done |
A separate field takes an error log file. If you fill it in, the engine writes the details of its work there, which helps when investigating failures.
Progress and result
Once started, the dialog shows progress: files and folders processed, remaining, speed and elapsed time. A long operation can be cancelled.
When it finishes, a window reports the outcome:
| Field | Meaning |
|---|---|
| Original size / Size in archive | Before and after — this is where the compression shows |
| Compression level | The final percentage |
| Files, New version, Total versions | What was added and how many versions the archive has |
| Parameters, Options | Which settings the archive was made with |
| Integrity check | Passed, or completed with errors |
Save the settings as a task
Save as Task… stores all the parameters in a .rzt
file. You can then run that file without the dialog and put it on a
schedule — see Tasks and scheduling.