Package: hashdeep
Version: 4.4-1
Severity: important

Dear Maintainer,

The removal of the symlinks that created the program
names md5deep, sha1deep, etc. destroys the ability of the
multi-call hashdeep progam to output and verify checksums
compatible with the equivalent GNU checksum utilities such
as md5sum, sha1sum, etc.

A simple shell session extract will illustrate this point
clearly:

$ ln -s /usr/bin/hashdeep sha1deep
$ ./sha1deep /usr/bin/hashdeep
cb1e7057d478656834c6862bab88c0870a384474  /usr/bin/hashdeep
$ /usr/bin/hashdeep /usr/bin/hashdeep
%%%% HASHDEEP-1.0
%%%% size,md5,sha256,filename
## Invoked from: [##SANITIZED##]
## $ /usr/bin/hashdeep /usr/bin/hashdeep
##
247960,b090629382bf52b7b49b1427d3b4692b,9948a73e441bedb0e87aaffca4d505dd2fd626796a9c0fdb5ca39c0d6592c3f9,/usr/bin/hashdeep

This feature is also highlighted by the author in a README
comment in Github repository for hashdeep:

"For historical reasons, the program has different
options and features when run with the names 'hashdeep'
and 'md5deep.'"

(https://github.com/jessek/hashdeep/blob/master/README.md)

Please restore the symlinks that were installed in earlier
version of the hashdeep/md5deep package (i.e. make md5deep,
sha1deep, etc. available again for direct command-line
execution)

For more information about multi-call binaries, you may read:
http://www.redbooks.ibm.com/abstracts/tips0092.html?Open

-- System Information:
Debian Release: stretch/sid
  APT prefers testing
  APT policy: (900, 'testing'), (90, 'unstable')
Architecture: amd64 (x86_64)
Foreign Architectures: i386

Kernel: Linux 4.0.0-2-amd64 (SMP w/4 CPU cores)
Shell: /bin/sh linked to /bin/dash
Init: sysvinit (via /sbin/init)

Versions of packages hashdeep depends on:
ii  libc6       2.19-19
ii  libgcc1     1:5.1.1-14
ii  libstdc++6  5.1.1-14

hashdeep recommends no packages.

hashdeep suggests no packages.

-- no debconf information


--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Reply via email to