- * Most of the I/O is now buffered, which results in a pretty
- large speedup when running under MS Windows. (Craig Barratt)
-
- * Optimizations to the name-handling/comparing code have made
- some significant reductions in user-CPU time for large file
- sets.
-
- * Some variable-type cleanup that makes the code more consistent.
-
- * Reduced memory requirements of hard link preservation.
- (J.W. Schultz)
-
- * Implemented a new algorithm for hard-link handling that speeds
- up the code significantly. (J.W. Schultz and Wayne Davison)
-
- * The --hard-link option now uses the first existing file in the
- group of linked files as the basis for the transfer. This
- prevents the sub-optimal transfer of a file's data when a new
- hardlink is added on the sending side and it sorts alphabetically
- earlier in the list than the files that are already present on the
- receiving side.
-
- * Dropped support for protocol versions less than 20
- (2.3.0 released 15 Mar 1999) and activated warnings for
- protocols less than 25 (2.5.0 released 23 Aug 2001)
- (Wayne Davison and J.W. Schultz, severally)
-
- * More optimal data transmission for --hard-links (protocol 28).
-
- * More optimal data transmission for --checksum (protocol 28).
-
- * Less memory is used when --checksum is specified.
-
- * Less memory is used in the file list (a per-file savings).
-
- * The generator is now better about not modifying the file list
- during the transfer in order to avoid a copy-on-write memory
- bifurcation (on systems where fork() uses shared memory).
- Previously, rsync's shared memory would slowly become unshared,
- resulting in real memory usage nearly doubling on the receiving
- side by the end of the transfer. Now, as long as permissions
- are being preserved, the shared memory should remain that way
- for the entire transfer.
-
- * Changed hardlink info and file_struct + strings to use
- allocation pools. This reduces memory use for large file-
- sets and permits freeing memory to the OS. (J.W. Schultz)
-
- * The 2 pipes used between the receiver and generator processes
- (which are forked on the same machine) were reduced to 1 pipe
- and the protocol improved so that (1) it is now impossible to
- have the "redo" pipe fill up and hang rsync, and (2) trailing
- messages from the receiver don't get lost on their way through
- the generator over to the sender (the latter mainly affected
- hard-link messages and verbose --stats output).
-
- * The reading & writing of the file-list in batch-mode is now
- handled by the same code that sends & receives the list over
- the wire. This makes it much easier to maintain.
-
- * Improved the internal uid/gid code to be more portable and
- a little more optimized.
-
- * Device numbers are now sent as separate major/minor values
- with 32-bit accuracy for each one (protocol 28). Previously
- hard-link device data was sent as a single 64-bit number, and
- copied devices were sent as a single 32-bit number. This will
- make inter-operation of 64-bit binaries more compatible with
- their 32-bit brethren. Note that optimizations in the binary
- protocol often sends the device data using fewer bytes than
- before, even though more precision is now available.
+ - Some buffer sizes were expanded a bit, particularly on systems where
+ MAXPATHLEN is overly small (e.g. cygwin).
+
+ - If io_printf() tries to format more data than fits in the buffer, exit
+ with an error instead of transmitting a truncated buffer.
+
+ - If a va_copy macro is defined, lib/snprintf.c will use it when defining
+ the VA_COPY macro.
+
+ - Reduced the amount of stack memory needed for each level of directory
+ recursion by nearly MAXPATHLEN bytes.
+
+ DEVELOPER RELATED:
+
+ - Several diffs in the patches dir now use the proper --enable-FOO
+ configure option instead of --with-FOO to turn on the inclusion of
+ the newly patched feature.
+