 | TurboSFV - Release history for version 11 |
| 2026-FEB-07 - v11.00 |
 |
|
| New hash algorithms: XXH3-64 and XXH3-128 |
TurboSFV now supports the latest variants from the xxHash family, called XXH3-64 and XXH3-128. XXH3
better supports small input data than the older variants XXH32 and XXH64: Depending on the amount
of data, different, specific optimized algorithms will be used.
But also for larger input, the computation is more efficient, as the XXH3 algorithm is better designed for SSE
based instructions (Streaming SIMD Extensions) on 64-bit CPUs. For 32-bit CPUs, the preferred algorithm
is still XXH32, because it operates with 32-bit numbers. Note, that a checksum produced by XXH3-64
is equivalent to the lower 64 bits of a XXH3-128 checksum.
In TurboSFV, the new algorithms XXH3-64 and XXH3-128 as well as the already implemented algorithms
XXH32 and XXH64 are now grouped together and available in a new menu group named xxHash family.
This group is available in the explorer context menu of a shell object, in the main menu of TurboSFV
and on the shell property sheet page "Hash" of a file, where these algorithms can be selected
for calculating checksums on the fly.
Checksums produced by the algorithms XXH3-64 and XXH3-128 are stored in a checksum file with the new
extension xh3. This file extension can be associated with TurboSFV during the installation, but also
in the configuration on page "Shell (Admin)", if TurboSFV was started with administration
rights. For XXH32 and XXH64 checksums, the file extension xxh is still used.
The new hash algorithms are available in all editions of TurboSFV, including the command-line versions
and the service application in TurboSFV XE: For service jobs, the XXH3 algorithms can be selected as
the hash type for the creation of hash files in the service job details.
Users, who are updating TurboSFV from the previous version 10, might first select the new algorithms in the
list of available algorithms in the configuration on page "Shell (User)", in order to see
them in the menus.
|
| Multipart checksum files |
Multipart checksum files are checksum files, where one big checksum file will be split into multiple
parts, and each part contains the same number of checksums, except the last part. In case of a huge
amount of files to hash, splitting the job into multiple parts can take much less resources than an
all-in-one job: TurboSFV calculates the checksums for a certain amount of files, writes the checksums
into a file with a unique part number in the file name, and then continues the job with the next part.
The resulting checksum files are then easier to handle: They are not only smaller, the validation can
be also done single-wise part by part, or all together, like one big checksum file with an overall result.
This new feature is available in the command-line versions, which are included in TurboSFV CE, PE and XE,
as well as in the service application of TurboSFV XE. The command-line version "Create" got
a new switch "/m", which expects a value, in order to specify the number of checksums per file, as a
power of 10. For example, "/m:3" instructs the program, to split the hash file into multiple parts,
each containing 10^3 = 1000 checksums. Each part gets its own part number, starting with 1, which
will be added to the file name together with a separator, for example "_1".
The validation of a multipart checksum file with the command-line version "Validate" works
similar: By specifying the /m switch (here without a value), the program expects the file name of the first
part of a multipart checksum file as the source argument, and then validates all checksums part by part,
as if it was one big checksum file. Without /m switch, the command-line version treats the source argument
as a single hash file, as usual.
For the service application included in TurboSFV XE, similar instructions can be set in the job details
of creation, validation and analysis service jobs.
Note that a multipart checksum file must be encoded in Unicode, for example UTF-8, to avoid problems
while writing file names with Unicode characters to the checksum file.
|
| Other changes and improvements |
- Problem fixed, that some controls and icons in the main program didn't scale properly.
- Numerous technical improvements.
|
| Installation notes |
The setup of TurboSFV version 11 now allows the update from version 10 - for the same
edition (CE, LE, PE, XE). It's no longer necessary to uninstall first the previous version.
As mentioned above, update users might select first the new XXH3 algorithms in the TurboSFV shell
configuration, in order to see them in the various menus. Also note, that registration keys
for version 10 are not valid in version 11 - below follow some hints for upgrading.
|
| Upgrade information for version 11 |
The registration keys for version 10 are not valid in version 11. Customers, who
purchased a license after 2025-OCT-01, can upgrade for free by following the instructions below.
For other customers, who acquired a license for version 10, there is a 50% discount available. If you haven't
yet received a coupon by e-mail, then you can request one.
In both cases, the acquired amount of licenses for the respective edition (CE, LE, PE or XE) can be upgraded.
If you are eligible, then please send an e-mail to the address below, providing the following information:
- Registration ID as indicated in the about box of TurboSFV: Navigate to the TurboSFV application menu,
select Help and then About. Do not send your personal registration key!
- The e-mail address which you have used for the purchase.
|
E-Mail:
|
|
This offer is valid until 2026-APR-30.
|
|
|
|
|