[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Please test: tarsnap 1.0.38

Everything looks good on OpenBSD -current. Should be a quick and simple
update in the ports tree.

On Fri, Jul 07, 2017 at 09:25:16PM +0000, Colin Percival wrote:
> Hash: SHA1
> Hi all,
> A potential version 1.0.38 of the Tarsnap client code is ready for testing.
> You can find the new code at
>   https://www.tarsnap.com/download/tarsnap-autoconf-
> and the tarball has SHA256 hash
>   e2b2485ce3afde9c883cd73bf5668d4feb80318d601ed142caad3dcb6da4e1ff
> (Note that the version numbering has changed -- previously this would have
> been "1.0.38a", but that was confusing some tools, so future "alphatest"
> versions are going to be (N-1).99 instead.)
> You can also see all the changes between 1.0.37 and this version in the
> newly-public git repository at
>   https://github.com/Tarsnap/tarsnap
> Potentially (but unlikely) breaking changes in this code compared to tarsnap
> 1.0.37:
> * tarsnap now applies the --humanize-numbers option to the "progress" output
> printed by SIGINFO / SIGUSR1.  If you have scripts which send signals to
> tarsnap and parse its output, they may need to be updated.
> * tarsnap -v now prints 'Deleting archive "foo"' even if only one archive is
> being deleted.  If you have scripts which parse the output of tarsnap -v,
> they may need to be updated.
> * tarsnap now prints a warning if you pass '--configfile /nosuchfile'.  If you
> have scripts which rely on being able to specify nonexistent configuration
> files and not get any warnings from tarsnap, they may need to be updated.
> (Also, if you were doing that, I *really* want to know why.)
> Other substantive changes in this code compared to tarsnap 1.0.37:
> * tarsnap now supports OpenSSL 1.1.
> * tarsnap accepts an --iso-dates option, which causes times to be printed in
> YYYY-MM-DD HH:MM:SS format rather than the traditional unix "ls -l" style.
> * tarsnap accepts a --force-resources option to proceed with decryption of a
> passphrase-encrypted key file even if it is anticipated to use an excessive
> amount of memory or CPU time.  (This may be useful if tarsnap inaccurately
> estimates the amount of memory your system has available.)
> * tarsnap accepts an --archive-names <file> option, which reads a list of
> archive names to operate on when operating in -d and --print-stats modes.
> (This should make some scripts simpler; no more need to construct a command
> line which has '-f' inserted before each archive name.)
> There are also two changes which most users should not encounter:
> * tarsnap now has an --initialize-cachedir mode; this is intended for use by
> the GUI.  (The cache directory will be initialized automatically in normal
> usage of the tarsnap command-line utility.)
> * tarsnap's configure script takes a --with-conf-no-sample option which
> results in the sample tarsnap configuration file being installed as
> "tarsnap.conf" rather than "tarsnap.conf.sample".  This is intended for use in
> some packaging systems which have mechanisms for managing configuration files,
> and should probably not be used by hand (since it will overwrite your existing
> tarsnap.conf).
> As usual, there are also lots of minor build fixes, harmless bug fixes, and
> code cleanups.
> Those of you who are using the tarsnap "experimental" package set should find
> that version will install via the usual apt-get command.  (Please
> let me know if this doesn't happen; in addition to testing tarsnap 1.0.38, I
> want to make sure that the binary package upgrades work as expected.)
> Assuming I don't get any emails complaining that something is broken, I'll
> be releasing this officially as version 1.0.38 in about a week.  As always,
> if you find any newly introduced bugs before the official release, you'll
> be eligible for double the normal Tarsnap bug bounties.
> - -- 
> Colin Percival
> Security Officer Emeritus, FreeBSD | The power to serve
> Founder, Tarsnap | www.tarsnap.com | Online backups for the truly paranoid
> iHMEARECADMWIQTq9Iu6fMd6MP78Dak4zsppDGpqbgUCWV/8NBUcY3BlcmNpdmFA
> dGFyc25hcC5jb20ACgkQOM7KaQxqam5+zgCeKGkJ5ontzGgtc1FEWz8pxVJS7WUA
> n1olqQJnKi+ITxfDnC/YTSJbfwqT
> =5r/Z

James Turner