On 1/26/19 8:25 PM, Eli Schwartz wrote:
Any particular benefit of this method over the other ones we support, by
the way?
lzip has two components, both of which are used: the file format, and
the implementation.
Advantages offered by the file format:
- integrity checking
- easy automated recovery of bit flip errors (most common data
corruption type)
- data recovery capabilities
- very simple (just using the lzip manual, it should be possible to
extract data from it by hand, even once LZMA is obsolete)
- the format itself is copylefted
Advantages offered by the implementation:
- versatile (-0 is similar in speed to gzip, -9 compresses more than
bzip2 on average)
- standardized options and return values (like bzip2, unlike gzip)
- dynamic dictionary size
The goal of the project was to make a new standard general-use archiver
to replace gzip and bzip2 (the author has severe issues with xz[1]).
As far as I'm aware, the format and the specification are technically
excellent, and I agree that it should likely become the default.
Thus, my personal reason for putting it in is as a first step - it can't
ever become the default if no one uses it, and nothing supports it (or
support is purely silent).
[1]: https://www.nongnu.org/lzip/xz_inadequate.html
