From 01348f04bc92f307f5f61dd9f9f4c8d7746336f5 Mon Sep 17 00:00:00 2001 From: Daniel Baumann Date: Fri, 1 Feb 2019 00:06:40 +0100 Subject: Adding upstream version 0.10. Signed-off-by: Daniel Baumann --- NEWS | 25 ++++++++++++------------- 1 file changed, 12 insertions(+), 13 deletions(-) (limited to 'NEWS') diff --git a/NEWS b/NEWS index cc2b1db..223e417 100644 --- a/NEWS +++ b/NEWS @@ -1,16 +1,15 @@ -Changes in version 0.9: +Changes in version 0.10: -Multi-threaded '-t, --list' has been implemented. See chapter 'Limitations -of parallel tar decoding' in the manual for details. +The new option '--bsolid', which selects per-data-block compression of the +archive, has been added. This option improves compression efficiency for +archives with lots of small files. -The new option '-n, --threads', which sets the number of decompression -threads, has been added. +The new option '-B, --data-size', which sets the size of the input data +blocks for '--bsolid', has been added. -Tarlz now recognizes global pax headers, but for now ignores them. - -Tarlz now decodes numerical fields in headers using length-safe parsers -instead of strtoul to prevent the parser from exceeding the end of the field -if it does not contain a terminating character. - -The new chapter 'Limitations of parallel tar decoding' has been added to the -manual. +If an extended header is required for any reason (for example a file size +larger than 8 GiB or a link name longer than 100 bytes), tarlz now moves the +filename also to the extended header to prevent an ustar tool from trying to +extract the file or link. This also makes easier during parallel extraction +or listing the detection of a tar member split between two lzip members at +the boundary between the extended header and the ustar header. -- cgit v1.2.3