MD5 it is then

-pastey-

supports the following hashes: MD2, MD4, MD5, CRC-32, Adler32, Gost,
Haval (128,160,192, 224, 256), SHA (1, 256, 384, 512), Tiger (128,
160, 192), RipeMD (128, 160).

Most used hashes are CRC-32, MD5 and SHA. Because of popularity first
two hash types (MD5 and CRC32) also have its own established file
formats. CRC-32 has a file format with extension .SFV. It was designed
for use in WinSFV software and was adopted by many CRC checkers. MD5
format originates from Linux. It is output of GNU md5sum utility and
very established in Linux community.

If you are wondering what are the numbers that are beside the word,
they are related to the length of the hash calculated. Usually the
longer the hash is the more it is reliable. For more information on
hash algorithms or specific hash algorithm I recommend you search the
internet.

I recommend you use MD5 algorithm. It is more reliable than CRC-32

Flasherly wrote:
-pastey-

supports the following hashes: MD2, MD4, MD5, CRC-32, Adler32, Gost,
Haval (128,160,192, 224, 256), SHA (1, 256, 384, 512), Tiger (128,
160, 192), RipeMD (128, 160).

Most used hashes are CRC-32, MD5 and SHA. Because of popularity first
two hash types (MD5 and CRC32) also have its own established file
formats. CRC-32 has a file format with extension .SFV. It was designed
for use in WinSFV software and was adopted by many CRC checkers. MD5
format originates from Linux. It is output of GNU md5sum utility and
very established in Linux community.

If you are wondering what are the numbers that are beside the word,
they are related to the length of the hash calculated. Usually the
longer the hash is the more it is reliable. For more information on
hash algorithms or specific hash algorithm I recommend you search the
internet.

I recommend you use MD5 algorithm. It is more reliable than CRC-32

The collision properties of MD5 are better than CRC32.

But MD5 is no good for authentication purposes. It's been cracked.

*******

Stick with SHA256 now for authentication.

Better still, is remote attestation by signing files using
the Flasherly Private Key (PKI). That covers cases where
a perpetrator modifies the file with the SHA256 values
stored in it for public consumption. As a user, at my end,
I use the Flasherly Public Key to verify the file was
created by Flasherly. The signature attests to the
genuine properties of the file.

Paul

On Wed, 24 Feb 2021 22:23:19 -0500, Paul
wrote:

The collision properties of MD5 are better than CRC32.

But MD5 is no good for authentication purposes. It's been cracked.

*******

Stick with SHA256 now for authentication.

Better still, is remote attestation by signing files using
the Flasherly Private Key (PKI). That covers cases where
a perpetrator modifies the file with the SHA256 values
stored in it for public consumption. As a user, at my end,
I use the Flasherly Public Key to verify the file was
created by Flasherly. The signature attests to the
genuine properties of the file.

The quote posted from a commercial program over a decade ago. Looks
like a later version, perhaps not much newer though, later was
licensed for terms of free/share distribution.

CDCheck v3.1.3.0
Copyright Mitja Perko
Date [27 November 2004]
[5.00MB]

I'm looking into building hash tables across audio files for
longer-term, collision/veracity purposes. Two copies tentatively
after some initial tests for building a base for later comparative
routines. MD5 still looks to be sufficient for that, a couple "aired"
private volumes, at least for, with hash tables on each volume dated
and built alike an initial opportunity for least potential of
corruption from inception. (Some of those audio files may old
themselves, but that's not say luck will continue always to favor much
anything in a digital storage realm.)

Seems I've recently seen MD5 floating around, presumably, still in
usage for authentication, or missed altogether that note in regard for
SHA256's status (albeit a SHA 512 suffix in the above iteration I
have).