569fb49fce · matheusbsbs/matheus-bagliano

@@ -8,7 +8,7 @@ When "git push [$there]" does not say what to push, we have used the

8 traditional "matching" semantics so far (all your branches were sent

9 to the remote as long as there already are branches of the same name

10 over there). In Git 2.0, the default will change to the "simple"

11 -semantics that pushes:

11 +semantics, which pushes:

12

13 - only the current branch to the branch with the same name, and only

14 when the current branch is set to integrate with that remote

@@ -53,9 +53,12 @@ Updates since v1.8.4

53

54 Foreign interfaces, subsystems and ports.

55

56 - * "git-svn" used with SVN 1.8.0 when talking over https:// connection

56 + * "git-svn" has been taught to use the serf library, which is the

57 + only option SVN 1.8.0 offers us when talking the HTTP protocol.

58 +

59 + * "git-svn" talking over an https:// connection using the serf library

60 dumped core due to a bug in the serf library that SVN uses. Work

58 - it around on our side, even though the SVN side is being fixed.

61 + around it on our side, even though the SVN side is being fixed.

62

63 * On MacOS X, we detected if the filesystem needs the "pre-composed

64 unicode strings" workaround, but did not automatically enable it.

@@ -65,10 +68,7 @@ Foreign interfaces, subsystems and ports.

68 repository relative to the home directory, e.g. "clone hg::~/there".

69

70 * imap-send ported to OS X uses Apple's security framework instead of

68 - OpenSSL one.

69 -

70 - * Subversion 1.8.0 that was recently released breaks older subversion

71 - clients coming over http/https in various ways.

71 + OpenSSL's.

72

73 * "git fast-import" treats an empty path given to "ls" as the root of

74 the tree.

@@ -79,22 +79,22 @@ UI, Workflows & Features

79 * xdg-open can be used as a browser backend for "git web-browse"

80 (hence to show "git help -w" output), when available.

81

82 - * "git grep" and "git show" pays attention to "--textconv" option

82 + * "git grep" and "git show" pay attention to the "--textconv" option

83 when these commands are told to operate on blob objects (e.g. "git

84 - grep -e pattern HEAD:Makefile").

84 + grep -e pattern --textconv HEAD:Makefile").

85

86 * "git replace" helper no longer allows an object to be replaced with

87 another object of a different type to avoid confusion (you can

88 - still manually craft such replacement using "git update-ref", as an

88 + still manually craft such a replacement using "git update-ref", as an

89 escape hatch).

90

91 - * "git status" no longer prints dirty status information for

91 + * "git status" no longer prints the dirty status information of

92 submodules for which submodule.$name.ignore is set to "all".

93

94 * "git rebase -i" honours core.abbrev when preparing the insn sheet

95 for editing.

96

97 - * "git status" during a cherry-pick shows what original commit is

97 + * "git status" during a cherry-pick shows which original commit is

98 being picked.

99

100 * Instead of typing four capital letters "HEAD", you can say "@" now,

@@ -102,21 +102,21 @@ UI, Workflows & Features

102

103 * "git check-ignore" follows the same rule as "git add" and "git

104 status" in that the ignore/exclude mechanism does not take effect

105 - on paths that are already tracked. With "--no-index" option, it

105 + on paths that are already tracked. With the "--no-index" option, it

106 can be used to diagnose which paths that should have been ignored

107 have been mistakenly added to the index.

108

109 * Some irrelevant "advice" messages that are shared with "git status"

110 output have been removed from the commit log template.

111

112 - * "update-refs" learnt a "--stdin" option to read multiple update

112 + * "update-refs" learned a "--stdin" option to read multiple update

113 requests and perform them in an all-or-none fashion.

114

115 * Just like "make -C <directory>", "git -C <directory> ..." tells Git

116 to go there before doing anything else.

117

118 - * Just like "git checkout -" knows to check out and "git merge -"

119 - knows to merge the branch you were previously on, "git cherry-pick"

118 + * Just like "git checkout -" knows to check out, and "git merge -"

119 + knows to merge, the branch you were previously on, "git cherry-pick"

120 now understands "git cherry-pick -" to pick from the previous

121 branch.

122

@@ -126,56 +126,58 @@ UI, Workflows & Features

126 "git status --porcelain" instead, as its format is stable and easier

127 to parse.

128

129 - * Make "foo^{tag}" to peel a tag to itself, i.e. no-op., and fail if

130 - "foo" is not a tag. "git rev-parse --verify v1.0^{tag}" would be

131 - a more convenient way to say "test $(git cat-file -t v1.0) = tag".

129 + * The ref syntax "foo^{tag}" (with the literal string "{tag}") peels a

130 + tag ref to itself, i.e. it's a no-op., and fails if

131 + "foo" is not a tag. "git rev-parse --verify v1.0^{tag}" is

132 + a more convenient way than "test $(git cat-file -t v1.0) = tag" to

133 + check if v1.0 is a tag.

134

135 * "git branch -v -v" (and "git status") did not distinguish among a

134 - branch that does not build on any other branch, a branch that is in

135 - sync with the branch it builds on, and a branch that is configured

136 - to build on some other branch that no longer exists.

136 + branch that is not based on any other branch, a branch that is in

137 + sync with its upstream branch, and a branch that is configured with an

138 + upstream branch that no longer exists.

139

138 - * Earlier we started rejecting an attempt to add 0{40} object name to

140 + * Earlier we started rejecting any attempt to add the 0{40} object name to

141 the index and to tree objects, but it sometimes is necessary to

140 - allow so to be able to use tools like filter-branch to correct such

141 - broken tree objects. "filter-branch" can again be used to to do

142 - so.

142 + allow this to be able to use tools like filter-branch to correct such

143 + broken tree objects. "filter-branch" can again be used to do this.

144

145 * "git config" did not provide a way to set or access numbers larger

146 than a native "int" on the platform; it now provides 64-bit signed

147 integers on all platforms.

148

149 * "git pull --rebase" always chose to do the bog-standard flattening

149 - rebase. You can tell it to run "rebase --preserve-merges" by

150 + rebase. You can tell it to run "rebase --preserve-merges" with

151 + "git pull --rebase=preserve" or by

152 setting "pull.rebase" configuration to "preserve".

153

154 * "git push --no-thin" actually disables the "thin pack transfer"

155 optimization.

156

155 - * Magic pathspecs like ":(icase)makefile" that matches both Makefile

156 - and makefile and ":(glob)foo/**/bar" that matches "bar" in "foo"

157 - and any subdirectory of "foo" can be used in more places.

157 + * Magic pathspecs like ":(icase)makefile" (matches both Makefile

158 + and makefile) and ":(glob)foo/**/bar" (matches "bar" in "foo"

159 + and any subdirectory of "foo") can be used in more places.

160

159 - * The "http.*" variables can now be specified per URL that the

160 - configuration applies. For example,

161 + * The "http.*" variables can now be specified for individual URLs.

162 + For example,

163

164 [http]

165 sslVerify = true

166 [http "https://weak.example.com/"]

167 sslVerify = false

168

167 - would flip http.sslVerify off only when talking to that specified

169 + would flip http.sslVerify off only when talking to that specific

170 site.

171

170 - * "git mv A B" when moving a submodule A has been taught to

171 - relocate its working tree and to adjust the paths in the

172 + * "git mv A B" when moving a submodule has been taught to

173 + relocate the submodule's working tree and to adjust the paths in the

174 .gitmodules file.

175

176 * "git blame" can now take more than one -L option to discover the

175 - origin of multiple blocks of the lines.

177 + origin of multiple blocks of lines.

178

179 * The http transport clients can optionally ask to save cookies

178 - with http.savecookies configuration variable.

180 + with the http.savecookies configuration variable.

181

182 * "git push" learned a more fine grained control over a blunt

183 "--force" when requesting a non-fast-forward update with the

@@ -197,7 +199,7 @@ UI, Workflows & Features

199

200 * "git whatchanged" may still be used by old timers, but mention of

201 it in documents meant for new users will only waste readers' time

200 - wonderig what the difference is between it and "git log". Make it

202 + wondering what the difference is between it and "git log". Make it

203 less prominent in the general part of the documentation and explain

204 that it is merely a "git log" with different default behaviour in

205 its own document.

@@ -219,23 +221,23 @@ Performance, Internal Implementation, etc.

221 should apply the same "no subprocess or pipe" optimization as we

222 apply to user-supplied GIT_PAGER=cat.

223

222 - * Many commands use --dashed-option as a operation mode selector

223 - (e.g. "git tag --delete") that the user can use at most one

224 - (e.g. "git tag --delete --verify" is a nonsense) and you cannot

225 - negate (e.g. "git tag --no-delete" is a nonsense). parse-options

224 + * Many commands use a --dashed-option as an operation mode selector

225 + (e.g. "git tag --delete") that excludes other operation modes

226 + (e.g. "git tag --delete --verify" is nonsense) and that cannot be

227 + negated (e.g. "git tag --no-delete" is nonsense). The parse-options

228 API learned a new OPT_CMDMODE macro to make it easier to implement

229 such a set of options.

230

229 - * OPT_BOOLEAN() in parse-options API was misdesigned to be "counting

231 + * OPT_BOOLEAN() in the parse-options API was misdesigned to be "counting

232 up" but many subcommands expect it to behave as "on/off". Update

233 them to use OPT_BOOL() which is a proper boolean.

234

233 - * "git gc" exits early without doing a double-work when it detects

235 + * "git gc" exits early without doing any work when it detects

236 that another instance of itself is already running.

237

238 * Under memory pressure and/or file descriptor pressure, we used to

237 - close pack windows that are not used and also closed filehandle to

238 - an open but unused packfiles. These are now controlled separately

239 + close pack windows that are not used and also closed filehandles to

240 + open but unused packfiles. These are now controlled separately

241 to better cope with the load.

242

243 Also contains various documentation updates and code clean-ups.

@@ -245,32 +247,32 @@ Fixes since v1.8.4

247 ------------------

248

249 Unless otherwise noted, all the fixes since v1.8.4 in the maintenance

248 -track are contained in this release (see release notes to them for

250 +track are contained in this release (see the maintenance releases' notes for

251 details).

252

253 * An ancient How-To on serving Git repositories on an HTTP server

252 - lacked a warning that it has been mostly superseded with more

254 + lacked a warning that it has been mostly superseded with a more

255 modern way.

256 (merge 6d52bc3 sc/doc-howto-dumb-http later to maint).

257

256 - * The interaction between use of Perl in our test suite and NO_PERL

258 + * The interaction between the use of Perl in our test suite and NO_PERL

259 has been clarified a bit.

260 (merge f8fc0ee jn/test-prereq-perl-doc later to maint).

261

260 - * The synopsis section of "git unpack-objects" documentation has been

262 + * The synopsis section of the "git unpack-objects" documentation has been

263 clarified a bit.

264 (merge 61e2e22 vd/doc-unpack-objects later to maint).

265

264 - * We did not generate HTML version of documentation to "git subtree"

266 + * We did not generate the HTML version of the documentation to "git subtree"

267 in contrib/.

268 (merge 95c62fb jk/subtree-install-fix later to maint).

269

270 * A fast-import stream expresses a pathname with funny characters by

269 - quoting them in C style; remote-hg remote helper forgot to unquote

271 + quoting them in C style; the remote-hg remote helper forgot to unquote

272 such a path.

273 (merge 1136265 ap/remote-hg-unquote-cquote later to maint).

274

273 - * "git reset -p HEAD" has a codepath to special case it to behave

275 + * "git reset -p HEAD" has a codepath to special-case it to behave

276 differently from resetting to contents of other commits, but a

277 recent change broke it.

278

@@ -289,38 +291,38 @@ details).

291 "HEAD" points at, and the receiving end needed to guess. A new

292 capability has been defined in the pack protocol to convey this

293 information so that cloning from a repository with more than one

292 - branches pointing at the same commit where the HEAD is at now

294 + branch pointing at the same commit where the HEAD is at now

295 reliably sets the initial branch in the resulting repository.

296 (merge 360a326 jc/upload-pack-send-symref later to maint).

297

296 - * We did not handle cases where http transport gets redirected during

298 + * We did not handle cases where the http transport gets redirected during

299 the authorization request (e.g. from http:// to https://).

300 (merge 70900ed jk/http-auth-redirects later to maint).

301

302 * Bash prompting code to deal with an SVN remote as an upstream

301 - were coded in a way not supported by older Bash versions (3.x).

303 + was coded in a way unsupported by older Bash versions (3.x).

304 (merge 52ec889 sg/prompt-svn-remote-fix later to maint).

305

306 * The fall-back parsing of commit objects with broken author or

305 - committer lines were less robust than ideal in picking up the

307 + committer lines was less robust than ideal in picking up the

308 timestamps.

309 (merge 03818a4 jk/split-broken-ident later to maint).

310

309 - * "git rev-list --objects ^v1.0^ v1.0" gave v1.0 tag itself in the

311 + * "git rev-list --objects ^v1.0^ v1.0" gave the v1.0 tag itself in the

312 output, but "git rev-list --objects v1.0^..v1.0" did not.

313 (merge 895c5ba jc/revision-range-unpeel later to maint).

314

313 - * "git clone" gave some progress messages to the standard output, not

314 - to the standard error, and did not allow suppressing them with the

315 + * "git clone" wrote some progress messages to standard output, not

316 + to standard error, and did not suppress them with the

317 --no-progress option.

318 (merge 643f918 jk/clone-progress-to-stderr later to maint).

319

318 - * "format-patch --from=<whom>" forgot to omit unnecessary in-body

320 + * "format-patch --from=<whom>" forgot to omit an unnecessary in-body

321 from line, i.e. when <whom> is the same as the real author.

322 (merge 662cc30 jk/format-patch-from later to maint).

323

324 * "git shortlog" used to choke and die when there is a malformed

323 - commit (e.g. missing authors); it now simply ignore such a commit

325 + commit (e.g. missing authors); it now simply ignores such a commit

326 and keeps going.

327 (merge cd4f09e jk/shortlog-tolerate-broken-commit later to maint).

328

@@ -334,16 +336,16 @@ details).

336 small empty messages to keep the connection alive.

337 (merge 115dedd jk/upload-pack-keepalive later to maint).

338

337 - * "git rebase" had a portability regression in v1.8.4 to trigger a

339 + * "git rebase" had a portability regression in v1.8.4 that triggered a

340 bug in some BSD shell implementations.

341 (merge 99855dd mm/rebase-continue-freebsd-WB later to maint).

342

343 * "git branch --track" had a minor regression in v1.8.3.2 and later

344 that made it impossible to base your local work on anything but a

343 - local branch of the upstream repository you are tracking from.

345 + local branch of the upstream repository you are tracking.

346 (merge b0f49ff jh/checkout-auto-tracking later to maint).

347

346 - * When the webserver responds with "405 Method Not Allowed", "git

348 + * When the web server responds with "405 Method Not Allowed", "git

349 http-backend" should tell the client what methods are allowed with

350 the "Allow" header.

351 (merge 9247be0 bc/http-backend-allow-405 later to maint).

@@ -357,22 +359,22 @@ details).

359 executable files.

360 (merge 1b48d56 jc/cvsserver-perm-bit-fix later to maint).

361

360 - * When send-email comes up with an error message to die with upon

362 + * When send-email obtains an error message to die with upon

363 failure to start an SSL session, it tried to read the error string

364 from a wrong place.

365 (merge 6cb0c88 bc/send-email-ssl-die-message-fix later to maint).

366

365 - * The implementation of "add -i" has a crippling code to work around

367 + * The implementation of "add -i" has some crippling code to work around an

368 ActiveState Perl limitation but it by mistake also triggered on Git

369 for Windows where MSYS perl is used.

370 (merge df17e77 js/add-i-mingw later to maint).

371

370 - * We made sure that we notice the user-supplied GIT_DIR is actually a

372 + * We made sure that we notice when the user-supplied GIT_DIR is actually a

373 gitfile, but did not do the same when the default ".git" is a

374 gitfile.

375 (merge 487a2b7 nd/git-dir-pointing-at-gitfile later to maint).

376

375 - * When an object is not found after checking the packfiles and then

377 + * When an object is not found after checking the packfiles and the

378 loose object directory, read_sha1_file() re-checks the packfiles to

379 prevent racing with a concurrent repacker; teach the same logic to

380 has_sha1_file().

@@ -390,22 +392,22 @@ details).

392 made it unnecessarily inefficient.

393 (merge 680be04 jc/ls-files-killed-optim later to maint).

394

393 - * The commit object names in the insn sheet that was prepared at the

394 - beginning of "rebase -i" session can become ambiguous as the

395 + * The shortened commit object names in the insn sheet that is prepared at the

396 + beginning of a "rebase -i" session can become ambiguous as the

397 rebasing progresses and the repository gains more commits. Make

398 sure the internal record is kept with full 40-hex object names.

399 (merge 75c6976 es/rebase-i-no-abbrev later to maint).

400

401 * "git rebase --preserve-merges" internally used the merge machinery

400 - and as a side effect, left merge summary message in the log, but

401 - when rebasing, there should not be a need for merge summary.

402 + and as a side effect left the merge summary message in the log, but

403 + when rebasing there is no need for the merge summary.

404 (merge a9f739c rt/rebase-p-no-merge-summary later to maint).

405

404 - * A call to xread() was used without a loop around to cope with short

405 - read in the codepath to stream new contents to a pack.

406 + * A call to xread() was used without a loop around it to cope with short

407 + reads in the codepath to stream new contents to a pack.

408 (merge e92527c js/xread-in-full later to maint).

409

408 - * "git rebase -i" forgot that the comment character can be

410 + * "git rebase -i" forgot that the comment character is

411 configurable while reading its insn sheet.

412 (merge 7bca7af es/rebase-i-respect-core-commentchar later to maint).

413

@@ -416,8 +418,8 @@ details).

418 * We used to send a large request to read(2)/write(2) as a single

419 system call, which was bad from the latency point of view when

420 the operation needs to be killed, and also triggered an error on

419 - broken 64-bit systems that refuse to take more than 2GB read or

420 - write in one go.

421 + broken 64-bit systems that refuse to read or write more than 2GB

422 + in one go.

423 (merge a487916 sp/clip-read-write-to-8mb later to maint).

424

425 * "git fetch" that auto-followed tags incorrectly reused the

@@ -431,17 +433,17 @@ details).

433 had a similar problem.

434 (merge 838f9a1 tr/log-full-diff-keep-true-parents later to maint).

435

434 - * Setting submodule.*.path configuration variable to true (without

436 + * Setting a submodule.*.path configuration variable to true (without

437 giving "= value") caused Git to segfault.

438 (merge 4b05440 jl/some-submodule-config-are-not-boolean later to maint).

439

440 * "git rebase -i" (there could be others, as the root cause is pretty

439 - generic) fed a random, data dependeant string to 'echo' and

440 - expects it to come out literally, corrupting its error message.

441 + generic) fed a random, data dependent string to 'echo' and

442 + expected it to come out literally, corrupting its error message.

443 (merge 89b0230 mm/no-shell-escape-in-die-message later to maint).

444

445 * Some people still use rather old versions of bash, which cannot

444 - grok some constructs like 'printf -v varname' the prompt and

446 + grok some constructs like 'printf -v varname' which the prompt and

447 completion code started to use recently.

448 (merge a44aa69 bc/completion-for-bash-3.0 later to maint).

449

RelNotes: spelling & grammar fixes