[thirdparty/git.git] / Documentation / RelNotes / 2.9.4.txt

Git v2.9.4 Release Notes
========================

Fixes since v2.9.3
------------------

 * There are certain house-keeping tasks that need to be performed at
   the very beginning of any Git program, and programs that are not
   built-in commands had to do them exactly the same way as "git"
   potty does.  It was easy to make mistakes in one-off standalone
   programs (like test helpers).  A common "main()" function that
   calls cmd_main() of individual program has been introduced to
   make it harder to make mistakes.

 * "git merge" with renormalization did not work well with
   merge-recursive, due to "safer crlf" conversion kicking in when it
   shouldn't.

 * The reflog output format is documented better, and a new format
   --date=unix to report the seconds-since-epoch (without timezone)
   has been added.

 * "git push --force-with-lease" already had enough logic to allow
   ensuring that such a push results in creation of a ref (i.e. the
   receiving end did not have another push from sideways that would be
   discarded by our force-pushing), but didn't expose this possibility
   to the users.  It does so now.

 * "import-tars" fast-import script (in contrib/) used to ignore a
   hardlink target and replaced it with an empty file, which has been
   corrected to record the same blob as the other file the hardlink is
   shared with.

 * "git mv dir non-existing-dir/" did not work in some environments
   the same way as existing mainstream platforms.  The code now moves
   "dir" to "non-existing-dir", without relying on rename("A", "B/")
   that strips the trailing slash of '/'.

 * The "t/" hierarchy is prone to get an unusual pathname; "make test"
   has been taught to make sure they do not contain paths that cannot
   be checked out on Windows (and the mechanism can be reusable to
   catch pathnames that are not portable to other platforms as need
   arises).

 * When "git merge-recursive" works on history with many criss-cross
   merges in "verbose" mode, the names the command assigns to the
   virtual merge bases could have overwritten each other by unintended
   reuse of the same piece of memory.

 * "git checkout --detach <branch>" used to give the same advice
   message as that is issued when "git checkout <tag>" (or anything
   that is not a branch name) is given, but asking with "--detach" is
   an explicit enough sign that the user knows what is going on.  The
   advice message has been squelched in this case.

 * "git difftool" by default ignores the error exit from the backend
   commands it spawns, because often they signal that they found
   differences by exiting with a non-zero status code just like "diff"
   does; the exit status codes 126 and above however are special in
   that they are used to signal that the command is not executable,
   does not exist, or killed by a signal.  "git difftool" has been
   taught to notice these exit status codes.

 * On Windows, help.browser configuration variable used to be ignored,
   which has been corrected.

 * The "git -c var[=val] cmd" facility to append a configuration
   variable definition at the end of the search order was described in
   git(1) manual page, but not in git-config(1), which was more likely
   place for people to look for when they ask "can I make a one-shot
   override, and if so how?"

 * The tempfile (hence its user lockfile) API lets the caller to open
   a file descriptor to a temporary file, write into it and then
   finalize it by first closing the filehandle and then either
   removing or renaming the temporary file.  When the process spawns a
   subprocess after obtaining the file descriptor, and if the
   subprocess has not exited when the attempt to remove or rename is
   made, the last step fails on Windows, because the subprocess has
   the file descriptor still open.  Open tempfile with O_CLOEXEC flag
   to avoid this (on Windows, this is mapped to O_NOINHERIT).

 * "git-shell" rejects a request to serve a repository whose name
   begins with a dash, which makes it no longer possible to get it
   confused into spawning service programs like "git-upload-pack" with
   an option like "--help", which in turn would spawn an interactive
   pager, instead of working with the repository user asked to access
   (i.e. the one whose name is "--help").

Also contains minor documentation updates and code clean-ups.
Commit	Line	Data
0202c411 JH	1	Git v2.9.4 Release Notes
	2	========================
	3
	4	Fixes since v2.9.3
	5	------------------
	6
	7	* There are certain house-keeping tasks that need to be performed at
	8	the very beginning of any Git program, and programs that are not
	9	built-in commands had to do them exactly the same way as "git"
	10	potty does. It was easy to make mistakes in one-off standalone
	11	programs (like test helpers). A common "main()" function that
	12	calls cmd_main() of individual program has been introduced to
	13	make it harder to make mistakes.
	14
	15	* "git merge" with renormalization did not work well with
	16	merge-recursive, due to "safer crlf" conversion kicking in when it
	17	shouldn't.
	18
	19	* The reflog output format is documented better, and a new format
	20	--date=unix to report the seconds-since-epoch (without timezone)
	21	has been added.
	22
	23	* "git push --force-with-lease" already had enough logic to allow
	24	ensuring that such a push results in creation of a ref (i.e. the
	25	receiving end did not have another push from sideways that would be
	26	discarded by our force-pushing), but didn't expose this possibility
	27	to the users. It does so now.
	28
	29	* "import-tars" fast-import script (in contrib/) used to ignore a
	30	hardlink target and replaced it with an empty file, which has been
	31	corrected to record the same blob as the other file the hardlink is
	32	shared with.
	33
	34	* "git mv dir non-existing-dir/" did not work in some environments
	35	the same way as existing mainstream platforms. The code now moves
	36	"dir" to "non-existing-dir", without relying on rename("A", "B/")
	37	that strips the trailing slash of '/'.
	38
	39	* The "t/" hierarchy is prone to get an unusual pathname; "make test"
	40	has been taught to make sure they do not contain paths that cannot
	41	be checked out on Windows (and the mechanism can be reusable to
	42	catch pathnames that are not portable to other platforms as need
	43	arises).
	44
	45	* When "git merge-recursive" works on history with many criss-cross
	46	merges in "verbose" mode, the names the command assigns to the
	47	virtual merge bases could have overwritten each other by unintended
	48	reuse of the same piece of memory.
	49
	50	* "git checkout --detach <branch>" used to give the same advice
	51	message as that is issued when "git checkout <tag>" (or anything
	52	that is not a branch name) is given, but asking with "--detach" is
	53	an explicit enough sign that the user knows what is going on. The
	54	advice message has been squelched in this case.
	55
	56	* "git difftool" by default ignores the error exit from the backend
	57	commands it spawns, because often they signal that they found
	58	differences by exiting with a non-zero status code just like "diff"
	59	does; the exit status codes 126 and above however are special in
	60	that they are used to signal that the command is not executable,
	61	does not exist, or killed by a signal. "git difftool" has been
	62	taught to notice these exit status codes.
	63
	64	* On Windows, help.browser configuration variable used to be ignored,
65	which has been corrected.
66
67	* The "git -c var[=val] cmd" facility to append a configuration
68	variable definition at the end of the search order was described in
69	git(1) manual page, but not in git-config(1), which was more likely
70	place for people to look for when they ask "can I make a one-shot
71	override, and if so how?"
72
73	* The tempfile (hence its user lockfile) API lets the caller to open
74	a file descriptor to a temporary file, write into it and then
75	finalize it by first closing the filehandle and then either
76	removing or renaming the temporary file. When the process spawns a
77	subprocess after obtaining the file descriptor, and if the
78	subprocess has not exited when the attempt to remove or rename is
79	made, the last step fails on Windows, because the subprocess has
80	the file descriptor still open. Open tempfile with O_CLOEXEC flag
81	to avoid this (on Windows, this is mapped to O_NOINHERIT).
82
d61226c1 JH	83	* "git-shell" rejects a request to serve a repository whose name
	84	begins with a dash, which makes it no longer possible to get it
	85	confused into spawning service programs like "git-upload-pack" with
	86	an option like "--help", which in turn would spawn an interactive
	87	pager, instead of working with the repository user asked to access
	88	(i.e. the one whose name is "--help").
	89
0202c411	90	Also contains minor documentation updates and code clean-ups.