Hi,
Yesterday I downloaded "7-Zip" V9.20 ("7z920.exe") after a
friend recommended it. One of my laptops (Dell) has Windows 7.
WinZip 7 (I used for years) will not install.
So far, I have used 7-Zip once to unzip a .ZIP file and I had
no problems. Note: 7-Zip is free.
Has anyone had any issues using this file(s) compressor?
Thank You in advance, John
Works fine here. Have it installed on several Windows
installs. There is also a version for Linux.
*******
For me, the best part of it, is the ability to read and
traverse various archive formats. That's what I use it
for mostly. So I do more "decompression" work with it
than "compression" work. On the decompression side, it
doesn't recognize the 20 or 30 different "packers" available
for executable code, but otherwise it does a pretty impressive
job of letting you look inside an archive. Anti virus
applications, have to know how to unpack all of those.
An example of a packer would be UPX.
http://en.wikipedia.org/wiki/Upx
*******
This would be for the compression side.
The only thing a bit weird about 7Zip, is the uneven usage
of multiple cores. There are cases where I think it could
benefit from multiple cores, where the author of the program
doesn't want to put the effort into making it work better.
But considering the author offers it for free, well, you
can't complain too much. I can't blame him for putting
more effort into the .7z flavor.
The .7z format provides excellent compression, but
with a heavy computational price. It's not something
that "better C coding" would fix
It's the nature
of the algorithm used. If you want to reduce entropy,
you have to work to get it.
For space-time tradeoffs, I like GZip. GZip isn't the absolutely
fastest compressor, but it's in the middle of the pack here.
Some of the ones that are faster, may have uneven performance
depending on the type of data you're feeding them. My favorite
might be GZip running with a compression level of 3 (on its scale
of 1..9). (And no, I wasn't as methodical as this person, and
test all nine settings. I selected three and went with it.
Maybe four is better?)
http://www.linuxjournal.com/article/8051
In terms of fastest GZip implementation, that would be
the multi-core PIGZ. (7Zip only uses one core for GZip compression.)
http://www.zlib.net/pigz/
There is a Windows port of PIGZ here. The problem I have
with this version, going by memory, was later, when
probing to see what size the files are in the archive,
it doesn't seem to handle things larger than 4GB properly.
So if I need "good quality" archives, I might flip over
to the Linux version and do the compression there. It's the
header information in the archive that may be a bit suspect
with the Windows version. (I recommend testing if you have
any questions. I didn't keep good notes of the differences
or anything.)
http://web.archive.org/web/20100107...r.net/index.php/my-software/58-downloads.html
http://web.archive.org/web/20110723181540/http://www.leszer.net/index.php/my-software.html
It's a fun topic, especially when you're trying to trim
down some 500GB archive files. That's when the
differences matter. I think if I went with the .7z
format as an option, it was threatening to take
all week (7 days).
I've never had any issues with the headers of the
archives made with 7Zip. So in terms of "quality",
no issues there. It was just the PIGZ program, where
you have to compare the Linux and Windows version,
to figure out which one is better.
Paul
