| 1 | @c This is part of the Emacs manual. |
| 2 | @c Copyright (C) 1985, 1986, 1987, 1993, 1994, 1995, 1997, 2000, 2001, |
| 3 | @c 2002, 2003, 2004, 2005, 2006, 2007, 2008 Free Software Foundation, Inc. |
| 4 | @c See file emacs.texi for copying conditions. |
| 5 | @node Microsoft Windows, Manifesto, Mac OS, Top |
| 6 | @appendix Emacs and Microsoft Windows/MS-DOS |
| 7 | @cindex Microsoft Windows |
| 8 | @cindex MS-Windows, Emacs peculiarities |
| 9 | |
| 10 | This section describes peculiarities of using Emacs on Microsoft |
| 11 | Windows. Some of these peculiarities are also relevant to Microsoft's |
| 12 | older MS-DOS ``operating system'' (also known as ``MS-DOG''). |
| 13 | However, Emacs features that are relevant @emph{only} to MS-DOS are |
| 14 | described in a separate |
| 15 | @iftex |
| 16 | manual (@pxref{MS-DOS,,, emacs-xtra, Specialized Emacs Features}). |
| 17 | @end iftex |
| 18 | @ifnottex |
| 19 | section (@pxref{MS-DOS}). |
| 20 | @end ifnottex |
| 21 | |
| 22 | |
| 23 | The behavior of Emacs on MS-Windows is reasonably similar to what is |
| 24 | documented in the rest of the manual, including support for long file |
| 25 | names, multiple frames, scroll bars, mouse menus, and subprocesses. |
| 26 | However, a few special considerations apply, and they are described |
| 27 | here. |
| 28 | |
| 29 | @menu |
| 30 | * Text and Binary:: Text files use CRLF to terminate lines. |
| 31 | * Windows Files:: File-name conventions on Windows. |
| 32 | * ls in Lisp:: Emulation of @code{ls} for Dired. |
| 33 | * Windows HOME:: Where Emacs looks for your @file{.emacs}. |
| 34 | * Windows Keyboard:: Windows-specific keyboard features. |
| 35 | * Windows Mouse:: Windows-specific mouse features. |
| 36 | * Windows Processes:: Running subprocesses on Windows. |
| 37 | * Windows Printing:: How to specify the printer on MS-Windows. |
| 38 | * Windows Fonts:: Specifying fonts on MS-Windows. |
| 39 | * Windows Misc:: Miscellaneous Windows features. |
| 40 | @ifnottex |
| 41 | * MS-DOS:: Using Emacs on MS-DOS (otherwise known as @dfn{MS-DOG}). |
| 42 | @end ifnottex |
| 43 | @end menu |
| 44 | |
| 45 | @node Text and Binary |
| 46 | @section Text Files and Binary Files |
| 47 | @cindex text and binary files on MS-DOS/MS-Windows |
| 48 | |
| 49 | GNU Emacs uses newline characters to separate text lines. This is the |
| 50 | convention used on GNU, Unix, and other Posix-compliant systems. |
| 51 | |
| 52 | @cindex end-of-line conversion on MS-DOS/MS-Windows |
| 53 | By contrast, MS-DOS and MS-Windows normally use carriage-return linefeed, |
| 54 | a two-character sequence, to separate text lines. (Linefeed is the same |
| 55 | character as newline.) Therefore, convenient editing of typical files |
| 56 | with Emacs requires conversion of these end-of-line (EOL) sequences. |
| 57 | And that is what Emacs normally does: it converts carriage-return |
| 58 | linefeed into newline when reading files, and converts newline into |
| 59 | carriage-return linefeed when writing files. The same mechanism that |
| 60 | handles conversion of international character codes does this conversion |
| 61 | also (@pxref{Coding Systems}). |
| 62 | |
| 63 | @cindex cursor location, on MS-DOS |
| 64 | @cindex point location, on MS-DOS |
| 65 | One consequence of this special format-conversion of most files is |
| 66 | that character positions as reported by Emacs (@pxref{Position Info}) do |
| 67 | not agree with the file size information known to the operating system. |
| 68 | |
| 69 | In addition, if Emacs recognizes from a file's contents that it uses |
| 70 | newline rather than carriage-return linefeed as its line separator, it |
| 71 | does not perform EOL conversion when reading or writing that file. |
| 72 | Thus, you can read and edit files from GNU and Unix systems on MS-DOS |
| 73 | with no special effort, and they will retain their Unix-style |
| 74 | end-of-line convention after you edit them. |
| 75 | |
| 76 | The mode line indicates whether end-of-line translation was used for |
| 77 | the current buffer. If MS-DOS end-of-line translation is in use for the |
| 78 | buffer, the MS-Windows build of Emacs displays a backslash @samp{\} after |
| 79 | the coding system mnemonic near the beginning of the mode line |
| 80 | (@pxref{Mode Line}). If no EOL translation was performed, the string |
| 81 | @samp{(Unix)} is displayed instead of the backslash, to alert you that the |
| 82 | file's EOL format is not the usual carriage-return linefeed. |
| 83 | |
| 84 | @cindex DOS-to-Unix conversion of files |
| 85 | To visit a file and specify whether it uses DOS-style or Unix-style |
| 86 | end-of-line, specify a coding system (@pxref{Text Coding}). For |
| 87 | example, @kbd{C-x @key{RET} c unix @key{RET} C-x C-f foobar.txt} |
| 88 | visits the file @file{foobar.txt} without converting the EOLs; if some |
| 89 | line ends with a carriage-return linefeed pair, Emacs will display |
| 90 | @samp{^M} at the end of that line. Similarly, you can direct Emacs to |
| 91 | save a buffer in a specified EOL format with the @kbd{C-x @key{RET} f} |
| 92 | command. For example, to save a buffer with Unix EOL format, type |
| 93 | @kbd{C-x @key{RET} f unix @key{RET} C-x C-s}. If you visit a file |
| 94 | with DOS EOL conversion, then save it with Unix EOL format, that |
| 95 | effectively converts the file to Unix EOL style, like @code{dos2unix}. |
| 96 | |
| 97 | @cindex untranslated file system |
| 98 | @findex add-untranslated-filesystem |
| 99 | When you use NFS, Samba, or some other similar method to access file |
| 100 | systems that reside on computers using GNU or Unix systems, Emacs |
| 101 | should not perform end-of-line translation on any files in these file |
| 102 | systems---not even when you create a new file. To request this, |
| 103 | designate these file systems as @dfn{untranslated} file systems by |
| 104 | calling the function @code{add-untranslated-filesystem}. It takes one |
| 105 | argument: the file system name, including a drive letter and |
| 106 | optionally a directory. For example, |
| 107 | |
| 108 | @example |
| 109 | (add-untranslated-filesystem "Z:") |
| 110 | @end example |
| 111 | |
| 112 | @noindent |
| 113 | designates drive Z as an untranslated file system, and |
| 114 | |
| 115 | @example |
| 116 | (add-untranslated-filesystem "Z:\\foo") |
| 117 | @end example |
| 118 | |
| 119 | @noindent |
| 120 | designates directory @file{\foo} on drive Z as an untranslated file |
| 121 | system. |
| 122 | |
| 123 | Most often you would use @code{add-untranslated-filesystem} in your |
| 124 | @file{.emacs} file, or in @file{site-start.el} so that all the users at |
| 125 | your site get the benefit of it. |
| 126 | |
| 127 | @findex remove-untranslated-filesystem |
| 128 | To countermand the effect of @code{add-untranslated-filesystem}, use |
| 129 | the function @code{remove-untranslated-filesystem}. This function takes |
| 130 | one argument, which should be a string just like the one that was used |
| 131 | previously with @code{add-untranslated-filesystem}. |
| 132 | |
| 133 | Designating a file system as untranslated does not affect character |
| 134 | set conversion, only end-of-line conversion. Essentially, it directs |
| 135 | Emacs to create new files with the Unix-style convention of using |
| 136 | newline at the end of a line. @xref{Coding Systems}. |
| 137 | |
| 138 | @vindex file-name-buffer-file-type-alist |
| 139 | @cindex binary files, on MS-DOS/MS-Windows |
| 140 | Some kinds of files should not be converted at all, because their |
| 141 | contents are not really text. Therefore, Emacs on MS-Windows distinguishes |
| 142 | certain files as @dfn{binary files}. (This distinction is not part of |
| 143 | MS-Windows; it is made by Emacs only.) Binary files include executable |
| 144 | programs, compressed archives, etc. Emacs uses the file name to decide |
| 145 | whether to treat a file as binary: the variable |
| 146 | @code{file-name-buffer-file-type-alist} defines the file-name patterns |
| 147 | that indicate binary files. If a file name matches one of the patterns |
| 148 | for binary files (those whose associations are of the type |
| 149 | @code{(@var{pattern} . t)}, Emacs reads and writes that file using the |
| 150 | @code{no-conversion} coding system (@pxref{Coding Systems}) which turns |
| 151 | off @emph{all} coding-system conversions, not only the EOL conversion. |
| 152 | @code{file-name-buffer-file-type-alist} also includes file-name patterns |
| 153 | for files which are known to be Windows-style text files with |
| 154 | carriage-return linefeed EOL format, such as @file{CONFIG.SYS}; Emacs |
| 155 | always writes those files with Windows-style EOLs. |
| 156 | |
| 157 | If a file which belongs to an untranslated file system matches one of |
| 158 | the file-name patterns in @code{file-name-buffer-file-type-alist}, the |
| 159 | EOL conversion is determined by @code{file-name-buffer-file-type-alist}. |
| 160 | |
| 161 | @node Windows Files |
| 162 | @section File Names on MS-Windows |
| 163 | @cindex file names on MS-Windows |
| 164 | |
| 165 | MS-Windows and MS-DOS normally use a backslash, @samp{\}, to |
| 166 | separate name units within a file name, instead of the slash used on |
| 167 | other systems. Emacs on MS-DOS/MS-Windows permits use of either slash or |
| 168 | backslash, and also knows about drive letters in file names. |
| 169 | |
| 170 | @cindex file-name completion, on MS-Windows |
| 171 | On MS-DOS/MS-Windows, file names are case-insensitive, so Emacs by |
| 172 | default ignores letter-case in file names during completion. |
| 173 | |
| 174 | @vindex w32-get-true-file-attributes |
| 175 | The variable @code{w32-get-true-file-attributes} controls whether |
| 176 | Emacs should issue additional system calls to determine more |
| 177 | accurately file attributes in primitives like @code{file-attributes} |
| 178 | and @code{directory-files-and-attributes}. These additional calls are |
| 179 | needed to report correct file ownership, link counts and file types |
| 180 | for special files such as pipes. Without these system calls, file |
| 181 | ownership will be attributed to the current user, link counts will be |
| 182 | always reported as 1, and special files will be reported as regular |
| 183 | files. |
| 184 | |
| 185 | If the value of this variable is @code{local} (the default), Emacs |
| 186 | will issue these additional system calls only for files on local fixed |
| 187 | drives. Any other non-@code{nil} value means do this even for |
| 188 | removable and remote volumes, where this could potentially slow down |
| 189 | Dired and other related features. The value of @code{nil} means never |
| 190 | issue those system calls. Non-@code{nil} values are only useful on |
| 191 | NTFS volumes, which support hard links and file security. |
| 192 | |
| 193 | @node ls in Lisp |
| 194 | @section Emulation of @code{ls} on MS-Windows |
| 195 | @cindex Dired, and MS-Windows/MS-DOS |
| 196 | @cindex @code{ls} emulation |
| 197 | |
| 198 | Dired normally uses the external program @code{ls} (or its close |
| 199 | work-alike) to produce the directory listing displayed in Dired |
| 200 | buffers (@pxref{Dired}). However, MS-Windows and MS-DOS systems don't |
| 201 | come with such a program, although several ports of @sc{gnu} @code{ls} |
| 202 | are available. Therefore, Emacs on those systems @emph{emulates} |
| 203 | @code{ls} in Lisp, by using the @file{ls-lisp.el} package. While |
| 204 | @file{ls-lisp.el} provides a reasonably full emulation of @code{ls}, |
| 205 | there are some options and features peculiar to that emulation; |
| 206 | @iftex |
| 207 | for more details, see the documentation of the variables whose names |
| 208 | begin with @code{ls-lisp}. |
| 209 | @end iftex |
| 210 | @ifnottex |
| 211 | they are described in this section. |
| 212 | |
| 213 | The @code{ls} emulation supports many of the @code{ls} switches, but |
| 214 | it doesn't support all of them. Here's the list of the switches it |
| 215 | does support: @option{-A}, @option{-a}, @option{-B}, @option{-C}, |
| 216 | @option{-c}, @option{-i}, @option{-G}, @option{-g}, @option{-R}, |
| 217 | @option{-r}, @option{-S}, @option{-s}, @option{-t}, @option{-U}, |
| 218 | @option{-u}, and @option{-X}. The @option{-F} switch is partially |
| 219 | supported (it appends the character that classifies the file, but does |
| 220 | not prevent symlink following). |
| 221 | |
| 222 | @vindex ls-lisp-use-insert-directory-program |
| 223 | On MS-Windows and MS-DOS, @file{ls-lisp.el} is preloaded when Emacs |
| 224 | is built, so the Lisp emulation of @code{ls} is always used on those |
| 225 | platforms. If you have a ported @code{ls}, setting |
| 226 | @code{ls-lisp-use-insert-directory-program} to a non-@code{nil} value |
| 227 | will revert to using an external program named by the variable |
| 228 | @code{insert-directory-program}. |
| 229 | |
| 230 | @vindex ls-lisp-ignore-case |
| 231 | By default, @file{ls-lisp.el} uses a case-sensitive sort order for |
| 232 | the directory listing it produces; this is so the listing looks the |
| 233 | same as on other platforms. If you wish that the files be sorted in |
| 234 | case-insensitive order, set the variable @code{ls-lisp-ignore-case} to |
| 235 | a non-@code{nil} value. |
| 236 | |
| 237 | @vindex ls-lisp-dirs-first |
| 238 | By default, files and subdirectories are sorted together, to emulate |
| 239 | the behavior of @code{ls}. However, native MS-Windows/MS-DOS file |
| 240 | managers list the directories before the files; if you want that |
| 241 | behavior, customize the option @code{ls-lisp-dirs-first} to a |
| 242 | non-@code{nil} value. |
| 243 | |
| 244 | @vindex ls-lisp-verbosity |
| 245 | The variable @code{ls-lisp-verbosity} controls the file attributes |
| 246 | that @file{ls-lisp.el} displays. The value should be a list that |
| 247 | contains one or more of the symbols @code{links}, @code{uid}, and |
| 248 | @code{gid}. @code{links} means display the count of different file |
| 249 | names that are associated with (a.k.a.@: @dfn{links to}) the file's |
| 250 | data; this is only useful on NTFS volumes. @code{uid} means display |
| 251 | the numerical identifier of the user who owns the file. @code{gid} |
| 252 | means display the numerical identifier of the file owner's group. The |
| 253 | default value is @code{(links uid gid)} i.e.@: all the 3 optional |
| 254 | attributes are displayed. |
| 255 | |
| 256 | @vindex ls-lisp-emulation |
| 257 | The variable @code{ls-lisp-emulation} controls the flavour of the |
| 258 | @code{ls} emulation by setting the defaults for the 3 options |
| 259 | described above: @code{ls-lisp-ignore-case}, |
| 260 | @code{ls-lisp-dirs-first}, and @code{ls-lisp-verbosity}. The value of |
| 261 | this option can be one of the following symbols: |
| 262 | |
| 263 | @table @code |
| 264 | @item GNU |
| 265 | @itemx nil |
| 266 | Emulate @sc{gnu} systems; this is the default. This sets |
| 267 | @code{ls-lisp-ignore-case} and @code{ls-lisp-dirs-first} to |
| 268 | @code{nil}, and @code{ls-lisp-verbosity} to @code{(links uid gid)}. |
| 269 | @item UNIX |
| 270 | Emulate Unix systems. Like @code{GNU}, but sets |
| 271 | @code{ls-lisp-verbosity} to @code{(links uid)}. |
| 272 | @item MacOS |
| 273 | Emulate MacOS. Sets @code{ls-lisp-ignore-case} to @code{t}, and |
| 274 | @code{ls-lisp-dirs-first} and @code{ls-lisp-verbosity} to @code{nil}. |
| 275 | @item MS-Windows |
| 276 | Emulate MS-Windows. Sets @code{ls-lisp-ignore-case} and |
| 277 | @code{ls-lisp-dirs-first} to @code{t}, and @code{ls-lisp-verbosity} to |
| 278 | @code{(links)} on Windows NT/2K/XP/2K3 and to @code{nil} on Windows 9X. |
| 279 | Note that the default emulation is @emph{not} @code{MS-Windows}, even |
| 280 | on Windows, since many users of Emacs on those platforms prefer the |
| 281 | @sc{gnu} defaults. |
| 282 | @end table |
| 283 | |
| 284 | @noindent |
| 285 | Any other value of @code{ls-lisp-emulation} means the same as |
| 286 | @code{GNU}. Note that this option needs to be set @emph{before} |
| 287 | @file{ls-lisp.el} is loaded, which means that on MS-Windows and MS-DOS |
| 288 | you will have to set the value from your @file{.emacs} file and then |
| 289 | restart Emacs, since @file{ls-lisp.el} is preloaded. |
| 290 | |
| 291 | @vindex ls-lisp-support-shell-wildcards |
| 292 | The variable @code{ls-lisp-support-shell-wildcards} controls how |
| 293 | file-name patterns are supported: if it is non-@code{nil} (the |
| 294 | default), they are treated as shell-style wildcards; otherwise they |
| 295 | are treated as Emacs regular expressions. |
| 296 | |
| 297 | @vindex ls-lisp-format-time-list |
| 298 | The variable @code{ls-lisp-format-time-list} defines how to format |
| 299 | the date and time of files. @emph{The value of this variable is |
| 300 | ignored}, unless Emacs cannot determine the current locale. (However, |
| 301 | if the value of @code{ls-lisp-use-localized-time-format} is |
| 302 | non-@code{nil}, Emacs obeys @code{ls-lisp-format-time-list} even if |
| 303 | the current locale is available; see below.) |
| 304 | |
| 305 | The value of @code{ls-lisp-format-time-list} is a list of 2 strings. |
| 306 | The first string is used if the file was modified within the current |
| 307 | year, while the second string is used for older files. In each of |
| 308 | these two strings you can use @samp{%}-sequences to substitute parts |
| 309 | of the time. For example: |
| 310 | @lisp |
| 311 | ("%b %e %H:%M" "%b %e %Y") |
| 312 | @end lisp |
| 313 | |
| 314 | @noindent |
| 315 | Note that the strings substituted for these @samp{%}-sequences depend |
| 316 | on the current locale. @xref{Time Parsing,,, elisp, The Emacs Lisp |
| 317 | Reference Manual}, for more about format time specs. |
| 318 | |
| 319 | @vindex ls-lisp-use-localized-time-format |
| 320 | Normally, Emacs formats the file time stamps in either traditional |
| 321 | or ISO-style time format. However, if the value of the variable |
| 322 | @code{ls-lisp-use-localized-time-format} is non-@code{nil}, Emacs |
| 323 | formats file time stamps according to what |
| 324 | @code{ls-lisp-format-time-list} specifies. The @samp{%}-sequences in |
| 325 | @code{ls-lisp-format-time-list} produce locale-dependent month and day |
| 326 | names, which might cause misalignment of columns in Dired display. |
| 327 | @end ifnottex |
| 328 | |
| 329 | @node Windows HOME |
| 330 | @section HOME Directory on MS-Windows |
| 331 | @cindex @code{HOME} directory on MS-Windows |
| 332 | |
| 333 | The Windows equivalent of the @code{HOME} directory is the |
| 334 | @dfn{user-specific application data directory}. The actual location |
| 335 | depends on your Windows version and system configuration; typical values |
| 336 | are @file{C:\Documents and Settings\@var{username}\Application Data} on |
| 337 | Windows 2K/XP and later, and either @file{C:\WINDOWS\Application Data} |
| 338 | or @file{C:\WINDOWS\Profiles\@var{username}\Application Data} on the |
| 339 | older Windows 9X/ME systems. |
| 340 | |
| 341 | @cindex init file @file{.emacs} on MS-Windows |
| 342 | The home directory is where your init file @file{.emacs} is stored. |
| 343 | When Emacs starts, it first checks whether the environment variable |
| 344 | @env{HOME} is set. If it is, it looks for the init file in the |
| 345 | directory pointed by @env{HOME}. If @env{HOME} is not defined, Emacs |
| 346 | checks for an existing @file{.emacs} file in @file{C:\}, the root |
| 347 | directory of drive @file{C:}@footnote{ |
| 348 | The check in @file{C:\} is for compatibility with older versions of Emacs, |
| 349 | which didn't check the application data directory. |
| 350 | }. If there's no such file in @file{C:\}, Emacs next uses the Windows |
| 351 | system calls to find out the exact location of your application data |
| 352 | directory. If that system call fails, Emacs falls back to @file{C:\}. |
| 353 | |
| 354 | Whatever the final place is, Emacs sets the value of the @env{HOME} |
| 355 | environment variable to point to it, and it will use that location for |
| 356 | other files and directories it normally creates in the user's home |
| 357 | directory. |
| 358 | |
| 359 | You can always find out where Emacs thinks is your home directory's |
| 360 | location by typing @kbd{C-x d ~/ @key{RET}}. This should present the |
| 361 | list of files in the home directory, and show its full name on the |
| 362 | first line. Likewise, to visit your init file, type @kbd{C-x C-f |
| 363 | ~/.emacs @key{RET}}. |
| 364 | |
| 365 | @cindex @file{_emacs} init file, MS-Windows |
| 366 | Because MS-DOS does not allow file names with leading dots, and |
| 367 | because older Windows systems made it hard to create files with such |
| 368 | names, the Windows port of Emacs supports an alternative name |
| 369 | @file{_emacs} as a fallback, if such a file exists in the home |
| 370 | directory, whereas @file{.emacs} does not. |
| 371 | |
| 372 | @node Windows Keyboard |
| 373 | @section Keyboard Usage on MS-Windows |
| 374 | @cindex keyboard, MS-Windows |
| 375 | |
| 376 | This section describes the Windows-specific features related to |
| 377 | keyboard input in Emacs. |
| 378 | |
| 379 | @cindex MS-Windows keyboard shortcuts |
| 380 | Many key combinations (known as ``keyboard shortcuts'') that have |
| 381 | conventional uses in MS-Windows programs conflict with traditional |
| 382 | Emacs key bindings. (These Emacs key bindings were established years |
| 383 | before Microsoft was founded.) Examples of conflicts include |
| 384 | @kbd{C-c}, @kbd{C-x}, @kbd{C-z}, @kbd{C-a}, and @kbd{W-@key{SPC}}. |
| 385 | You can redefine some of them with meanings more like the MS-Windows |
| 386 | meanings by enabling CUA Mode (@pxref{CUA Bindings}). |
| 387 | |
| 388 | @kindex F10 @r{(MS-Windows)} |
| 389 | @cindex menu bar access using keyboard @r{(MS-Windows)} |
| 390 | The @key{F10} key on Windows activates the menu bar in a way that |
| 391 | makes it possible to use the menus without a mouse. In this mode, the |
| 392 | arrow keys traverse the menus, @key{RET} selects a highlighted menu |
| 393 | item, and @key{ESC} closes the menu. |
| 394 | |
| 395 | @iftex |
| 396 | @inforef{Windows Keyboard, , emacs}, for information about additional |
| 397 | Windows-specific variables in this category. |
| 398 | @end iftex |
| 399 | @ifnottex |
| 400 | @vindex w32-alt-is-meta |
| 401 | @cindex @code{Alt} key (MS-Windows) |
| 402 | By default, the key labeled @key{Alt} is mapped as the @key{META} |
| 403 | key. If you wish it to produce the @code{Alt} modifier instead, set |
| 404 | the variable @code{w32-alt-is-meta} to a @code{nil} value. |
| 405 | |
| 406 | @findex w32-register-hot-key |
| 407 | @findex w32-unregister-hot-key |
| 408 | MS-Windows reserves certain key combinations, such as |
| 409 | @kbd{Alt-@key{TAB}}, for its own use. These key combinations are |
| 410 | intercepted by the system before Emacs can see them. You can use the |
| 411 | @code{w32-register-hot-key} function to allow a key sequence to be |
| 412 | seen by Emacs instead of being grabbed by Windows. This functions |
| 413 | registers a key sequence as a @dfn{hot key}, overriding the special |
| 414 | meaning of that key sequence for Windows. (MS-Windows is told that |
| 415 | the key sequence is a hot key only when one of the Emacs windows has |
| 416 | focus, so that the special keys still have their usual meaning for |
| 417 | other Windows applications.) |
| 418 | |
| 419 | The argument to @code{w32-register-hot-key} must be a single key, |
| 420 | with or without modifiers, in vector form that would be acceptable to |
| 421 | @code{define-key}. The meta modifier is interpreted as the @key{ALT} |
| 422 | key if @code{w32-alt-is-meta} is @code{t} (the default), and the hyper |
| 423 | modifier is always interpreted as the Windows key (usually labeled |
| 424 | with @key{start} and the Windows logo). If the function succeeds in |
| 425 | registering the key sequence, it returns the hotkey ID, a number; |
| 426 | otherwise it returns @code{nil}. |
| 427 | |
| 428 | @kindex M-TAB@r{, (MS-Windows)} |
| 429 | @cindex @kbd{M-@key{TAB}} vs @kbd{Alt-@key{TAB}} (MS-Windows) |
| 430 | @cindex @kbd{Alt-@key{TAB}} vs @kbd{M-@key{TAB}} (MS-Windows) |
| 431 | For example, @code{(w32-register-hot-key [M-tab])} lets you use |
| 432 | @kbd{M-TAB} normally in Emacs, for instance, to complete the word or |
| 433 | symbol at point at top level, or to complete the current search string |
| 434 | against previously sought strings during incremental search. |
| 435 | |
| 436 | The function @code{w32-unregister-hot-key} reverses the effect of |
| 437 | @code{w32-register-hot-key} for its argument key sequence. |
| 438 | |
| 439 | @vindex w32-capslock-is-shiftlock |
| 440 | By default, the @key{CapsLock} key only affects normal character |
| 441 | keys (it converts lower-case characters to their upper-case |
| 442 | variants). However, if you set the variable |
| 443 | @code{w32-capslock-is-shiftlock} to a non-@code{nil} value, the |
| 444 | @key{CapsLock} key will affect non-character keys as well, as if you |
| 445 | pressed the @key{Shift} key while typing the non-character key. |
| 446 | |
| 447 | @vindex w32-enable-caps-lock |
| 448 | If the variable @code{w32-enable-caps-lock} is set to a @code{nil} |
| 449 | value, the @key{CapsLock} key produces the symbol @code{capslock} |
| 450 | instead of the shifted version of they keys. The default value is |
| 451 | @code{t}. |
| 452 | |
| 453 | @vindex w32-enable-num-lock |
| 454 | @cindex keypad keys (MS-Windows) |
| 455 | Similarly, if @code{w32-enable-num-lock} is @code{nil}, the |
| 456 | @key{NumLock} key will produce the symbol @code{kp-numlock}. The |
| 457 | default is @code{t}, which causes @key{NumLock} to work as expected: |
| 458 | toggle the meaning of the keys on the numeric keypad. |
| 459 | @end ifnottex |
| 460 | |
| 461 | @vindex w32-apps-modifier |
| 462 | The variable @code{w32-apps-modifier} controls the effect of the |
| 463 | @key{Apps} key (usually located between the right @key{Alt} and the |
| 464 | right @key{Ctrl} keys). Its value can be one of the symbols |
| 465 | @code{hyper}, @code{super}, @code{meta}, @code{alt}, @code{control}, |
| 466 | or @code{shift} for the respective modifier, or @code{nil} to appear |
| 467 | as the key @code{apps}. The default is @code{nil}. |
| 468 | |
| 469 | @vindex w32-lwindow-modifier |
| 470 | @vindex w32-rwindow-modifier |
| 471 | @vindex w32-scroll-lock-modifier |
| 472 | The variable @code{w32-lwindow-modifier} determines the effect of |
| 473 | the left Windows key (usually labeled with @key{start} and the Windows |
| 474 | logo). If its value is @code{nil} (the default), the key will produce |
| 475 | the symbol @code{lwindow}. Setting it to one of the symbols |
| 476 | @code{hyper}, @code{super}, @code{meta}, @code{alt}, @code{control}, |
| 477 | or @code{shift} will produce the respective modifier. A similar |
| 478 | variable @code{w32-rwindow-modifier} controls the effect of the right |
| 479 | Windows key, and @code{w32-scroll-lock-modifier} does the same for the |
| 480 | @key{ScrLock} key. If these variables are set to @code{nil}, the |
| 481 | right Windows key produces the symbol @code{rwindow} and @key{ScrLock} |
| 482 | produces the symbol @code{scroll}. |
| 483 | |
| 484 | @vindex w32-pass-alt-to-system |
| 485 | @cindex Windows system menu |
| 486 | @cindex @code{Alt} key invokes menu (Windows) |
| 487 | Emacs compiled as a native Windows application normally turns off |
| 488 | the Windows feature that tapping the @key{ALT} key invokes the Windows |
| 489 | menu. The reason is that the @key{ALT} serves as @key{META} in Emacs. |
| 490 | When using Emacs, users often press the @key{META} key temporarily and |
| 491 | then change their minds; if this has the effect of bringing up the |
| 492 | Windows menu, it alters the meaning of subsequent commands. Many |
| 493 | users find this frustrating. |
| 494 | |
| 495 | You can re-enable Windows' default handling of tapping the @key{ALT} |
| 496 | key by setting @code{w32-pass-alt-to-system} to a non-@code{nil} |
| 497 | value. |
| 498 | |
| 499 | @ifnottex |
| 500 | @vindex w32-pass-lwindow-to-system |
| 501 | @vindex w32-pass-rwindow-to-system |
| 502 | The variables @code{w32-pass-lwindow-to-system} and |
| 503 | @code{w32-pass-rwindow-to-system} determine whether the respective |
| 504 | keys are passed to Windows or swallowed by Emacs. If the value is |
| 505 | @code{nil}, the respective key is silently swallowed by Emacs, |
| 506 | otherwise it is passed to Windows. The default is @code{t} for both |
| 507 | of these variables. Passing each of these keys to Windows produces |
| 508 | its normal effect: for example, @kbd{@key{Lwindow}} opens the |
| 509 | @code{Start} menu, etc.@footnote{ |
| 510 | Some combinations of the ``Windows'' keys with other keys are caught |
| 511 | by Windows at low level in a way that Emacs currently cannot prevent. |
| 512 | For example, @kbd{@key{Lwindow} r} always pops up the Windows |
| 513 | @samp{Run} dialog. Customizing the value of |
| 514 | @code{w32-phantom-key-code} might help in some cases, though.} |
| 515 | |
| 516 | @vindex w32-recognize-altgr |
| 517 | @kindex AltGr @r{(MS-Windows)} |
| 518 | @cindex AltGr key (MS-Windows) |
| 519 | The variable @code{w32-recognize-altgr} controls whether the |
| 520 | @key{AltGr} key (if it exists on your keyboard), or its equivalent, |
| 521 | the combination of the right @key{Alt} and left @key{Ctrl} keys |
| 522 | pressed together, is recognized as the @key{AltGr} key. The default |
| 523 | is @code{t}, which means these keys produce @code{AltGr}; setting it |
| 524 | to @code{nil} causes @key{AltGr} or the equivalent key combination to |
| 525 | be interpreted as the combination of @key{CTRL} and @key{META} |
| 526 | modifiers. |
| 527 | @end ifnottex |
| 528 | |
| 529 | @node Windows Mouse |
| 530 | @section Mouse Usage on MS-Windows |
| 531 | @cindex mouse, and MS-Windows |
| 532 | |
| 533 | This section describes the Windows-specific variables related to |
| 534 | mouse. |
| 535 | |
| 536 | @vindex w32-mouse-button-tolerance |
| 537 | @cindex simulation of middle mouse button |
| 538 | The variable @code{w32-mouse-button-tolerance} specifies the |
| 539 | time interval, in milliseconds, for faking middle mouse button press |
| 540 | on 2-button mice. If both mouse buttons are depressed within this |
| 541 | time interval, Emacs generates a middle mouse button click event |
| 542 | instead of a double click on one of the buttons. |
| 543 | |
| 544 | @vindex w32-pass-extra-mouse-buttons-to-system |
| 545 | If the variable @code{w32-pass-extra-mouse-buttons-to-system} is |
| 546 | non-@code{nil}, Emacs passes the fourth and fifth mouse buttons to |
| 547 | Windows. |
| 548 | |
| 549 | @vindex w32-swap-mouse-buttons |
| 550 | The variable @code{w32-swap-mouse-buttons} controls which of the 3 |
| 551 | mouse buttons generates the @kbd{mouse-2} events. When it is |
| 552 | @code{nil} (the default), the middle button generates @kbd{mouse-2} |
| 553 | and the right button generates @kbd{mouse-3} events. If this variable |
| 554 | is non-@code{nil}, the roles of these two buttons are reversed. |
| 555 | |
| 556 | @node Windows Processes |
| 557 | @section Subprocesses on Windows 9X/ME and Windows NT/2K/XP |
| 558 | @cindex subprocesses on MS-Windows |
| 559 | |
| 560 | @cindex DOS applications, running from Emacs |
| 561 | Emacs compiled as a native Windows application (as opposed to the DOS |
| 562 | version) includes full support for asynchronous subprocesses. |
| 563 | In the Windows version, synchronous and asynchronous subprocesses work |
| 564 | fine on both |
| 565 | Windows 9X/ME and Windows NT/2K/XP as long as you run only 32-bit Windows |
| 566 | applications. However, when you run a DOS application in a subprocess, |
| 567 | you may encounter problems or be unable to run the application at all; |
| 568 | and if you run two DOS applications at the same time in two |
| 569 | subprocesses, you may have to reboot your system. |
| 570 | |
| 571 | Since the standard command interpreter (and most command line utilities) |
| 572 | on Windows 9X are DOS applications, these problems are significant when |
| 573 | using that system. But there's nothing we can do about them; only |
| 574 | Microsoft can fix them. |
| 575 | |
| 576 | If you run just one DOS application subprocess, the subprocess should |
| 577 | work as expected as long as it is ``well-behaved'' and does not perform |
| 578 | direct screen access or other unusual actions. If you have a CPU |
| 579 | monitor application, your machine will appear to be 100% busy even when |
| 580 | the DOS application is idle, but this is only an artifact of the way CPU |
| 581 | monitors measure processor load. |
| 582 | |
| 583 | You must terminate the DOS application before you start any other DOS |
| 584 | application in a different subprocess. Emacs is unable to interrupt or |
| 585 | terminate a DOS subprocess. The only way you can terminate such a |
| 586 | subprocess is by giving it a command that tells its program to exit. |
| 587 | |
| 588 | If you attempt to run two DOS applications at the same time in separate |
| 589 | subprocesses, the second one that is started will be suspended until the |
| 590 | first one finishes, even if either or both of them are asynchronous. |
| 591 | |
| 592 | @cindex kill DOS application |
| 593 | If you can go to the first subprocess, and tell it to exit, the second |
| 594 | subprocess should continue normally. However, if the second subprocess |
| 595 | is synchronous, Emacs itself will be hung until the first subprocess |
| 596 | finishes. If it will not finish without user input, then you have no |
| 597 | choice but to reboot if you are running on Windows 9X. If you are |
| 598 | running on Windows NT/2K/XP, you can use a process viewer application to kill |
| 599 | the appropriate instance of NTVDM instead (this will terminate both DOS |
| 600 | subprocesses). |
| 601 | |
| 602 | If you have to reboot Windows 9X in this situation, do not use the |
| 603 | @code{Shutdown} command on the @code{Start} menu; that usually hangs the |
| 604 | system. Instead, type @kbd{CTL-ALT-@key{DEL}} and then choose |
| 605 | @code{Shutdown}. That usually works, although it may take a few minutes |
| 606 | to do its job. |
| 607 | |
| 608 | @vindex w32-quote-process-args |
| 609 | The variable @code{w32-quote-process-args} controls how Emacs quotes |
| 610 | the process arguments. Non-@code{nil} means quote with the @code{"} |
| 611 | character. If the value is a character, use that character to escape |
| 612 | any quote characters that appear; otherwise chose a suitable escape |
| 613 | character based on the type of the program. |
| 614 | |
| 615 | @ifnottex |
| 616 | @findex w32-shell-execute |
| 617 | The function @code{w32-shell-execute} can be useful for writing |
| 618 | customized commands that run MS-Windows applications registered to |
| 619 | handle a certain standard Windows operation for a specific type of |
| 620 | document or file. This function is a wrapper around the Windows |
| 621 | @code{ShellExecute} API. See the MS-Windows API documentation for |
| 622 | more details. |
| 623 | @end ifnottex |
| 624 | |
| 625 | @node Windows Printing |
| 626 | @section Printing and MS-Windows |
| 627 | |
| 628 | Printing commands, such as @code{lpr-buffer} (@pxref{Printing}) and |
| 629 | @code{ps-print-buffer} (@pxref{PostScript}) work in MS-DOS and |
| 630 | MS-Windows by sending the output to one of the printer ports, if a |
| 631 | Posix-style @code{lpr} program is unavailable. The same Emacs |
| 632 | variables control printing on all systems, but in some cases they have |
| 633 | different default values on MS-DOS and MS-Windows. |
| 634 | |
| 635 | Emacs on Windows automatically determines your default printer and |
| 636 | sets the variable @code{printer-name} to that printer's name. But in |
| 637 | some rare cases this can fail, or you may wish to use a different |
| 638 | printer from within Emacs. The rest of this section explains how to |
| 639 | tell Emacs which printer to use. |
| 640 | |
| 641 | @vindex printer-name@r{, (MS-DOS/MS-Windows)} |
| 642 | If you want to use your local printer, then set the Lisp variable |
| 643 | @code{lpr-command} to @code{""} (its default value on Windows) and |
| 644 | @code{printer-name} to the name of the printer port---for example, |
| 645 | @code{"PRN"}, the usual local printer port or @code{"LPT2"}, or |
| 646 | @code{"COM1"} for a serial printer. You can also set |
| 647 | @code{printer-name} to a file name, in which case ``printed'' output |
| 648 | is actually appended to that file. If you set @code{printer-name} to |
| 649 | @code{"NUL"}, printed output is silently discarded (sent to the system |
| 650 | null device). |
| 651 | |
| 652 | You can also use a printer shared by another machine by setting |
| 653 | @code{printer-name} to the UNC share name for that printer---for |
| 654 | example, @code{"//joes_pc/hp4si"}. (It doesn't matter whether you use |
| 655 | forward slashes or backslashes here.) To find out the names of shared |
| 656 | printers, run the command @samp{net view} from the command prompt to |
| 657 | obtain a list of servers, and @samp{net view @var{server-name}} to see |
| 658 | the names of printers (and directories) shared by that server. |
| 659 | Alternatively, click the @samp{Network Neighborhood} icon on your |
| 660 | desktop, and look for machines which share their printers via the |
| 661 | network. |
| 662 | |
| 663 | @cindex @samp{net use}, and printing on MS-Windows |
| 664 | @cindex networked printers (MS-Windows) |
| 665 | If the printer doesn't appear in the output of @samp{net view}, or |
| 666 | if setting @code{printer-name} to the UNC share name doesn't produce a |
| 667 | hardcopy on that printer, you can use the @samp{net use} command to |
| 668 | connect a local print port such as @code{"LPT2"} to the networked |
| 669 | printer. For example, typing @kbd{net use LPT2: \\joes_pc\hp4si}@footnote{ |
| 670 | Note that the @samp{net use} command requires the UNC share name to be |
| 671 | typed with the Windows-style backslashes, while the value of |
| 672 | @code{printer-name} can be set with either forward- or backslashes.} |
| 673 | causes Windows to @dfn{capture} the @code{LPT2} port and redirect the |
| 674 | printed material to the printer connected to the machine @code{joes_pc}. |
| 675 | After this command, setting @code{printer-name} to @code{"LPT2"} |
| 676 | should produce the hardcopy on the networked printer. |
| 677 | |
| 678 | With some varieties of Windows network software, you can instruct |
| 679 | Windows to capture a specific printer port such as @code{"LPT2"}, and |
| 680 | redirect it to a networked printer via the @w{@code{Control |
| 681 | Panel->Printers}} applet instead of @samp{net use}. |
| 682 | |
| 683 | If you set @code{printer-name} to a file name, it's best to use an |
| 684 | absolute file name. Emacs changes the working directory according to |
| 685 | the default directory of the current buffer, so if the file name in |
| 686 | @code{printer-name} is relative, you will end up with several such |
| 687 | files, each one in the directory of the buffer from which the printing |
| 688 | was done. |
| 689 | |
| 690 | If the value of @code{printer-name} is correct, but printing does |
| 691 | not produce the hardcopy on your printer, it is possible that your |
| 692 | printer does not support printing plain text (some cheap printers omit |
| 693 | this functionality). In that case, try the PostScript print commands, |
| 694 | described below. |
| 695 | |
| 696 | @findex print-buffer @r{(MS-DOS)} |
| 697 | @findex print-region @r{(MS-DOS)} |
| 698 | @vindex lpr-headers-switches @r{(MS-DOS)} |
| 699 | The commands @code{print-buffer} and @code{print-region} call the |
| 700 | @code{pr} program, or use special switches to the @code{lpr} program, to |
| 701 | produce headers on each printed page. MS-DOS and MS-Windows don't |
| 702 | normally have these programs, so by default, the variable |
| 703 | @code{lpr-headers-switches} is set so that the requests to print page |
| 704 | headers are silently ignored. Thus, @code{print-buffer} and |
| 705 | @code{print-region} produce the same output as @code{lpr-buffer} and |
| 706 | @code{lpr-region}, respectively. If you do have a suitable @code{pr} |
| 707 | program (for example, from GNU Coreutils), set |
| 708 | @code{lpr-headers-switches} to @code{nil}; Emacs will then call |
| 709 | @code{pr} to produce the page headers, and print the resulting output as |
| 710 | specified by @code{printer-name}. |
| 711 | |
| 712 | @vindex print-region-function @r{(MS-DOS)} |
| 713 | @cindex lpr usage under MS-DOS |
| 714 | @vindex lpr-command @r{(MS-DOS)} |
| 715 | @vindex lpr-switches @r{(MS-DOS)} |
| 716 | Finally, if you do have an @code{lpr} work-alike, you can set the |
| 717 | variable @code{lpr-command} to @code{"lpr"}. Then Emacs will use |
| 718 | @code{lpr} for printing, as on other systems. (If the name of the |
| 719 | program isn't @code{lpr}, set @code{lpr-command} to specify where to |
| 720 | find it.) The variable @code{lpr-switches} has its standard meaning |
| 721 | when @code{lpr-command} is not @code{""}. If the variable |
| 722 | @code{printer-name} has a string value, it is used as the value for the |
| 723 | @code{-P} option to @code{lpr}, as on Unix. |
| 724 | |
| 725 | @findex ps-print-buffer @r{(MS-DOS)} |
| 726 | @findex ps-spool-buffer @r{(MS-DOS)} |
| 727 | @vindex ps-printer-name @r{(MS-DOS)} |
| 728 | @vindex ps-lpr-command @r{(MS-DOS)} |
| 729 | @vindex ps-lpr-switches @r{(MS-DOS)} |
| 730 | A parallel set of variables, @code{ps-lpr-command}, |
| 731 | @code{ps-lpr-switches}, and @code{ps-printer-name} (@pxref{PostScript |
| 732 | Variables}), defines how PostScript files should be printed. These |
| 733 | variables are used in the same way as the corresponding variables |
| 734 | described above for non-PostScript printing. Thus, the value of |
| 735 | @code{ps-printer-name} is used as the name of the device (or file) to |
| 736 | which PostScript output is sent, just as @code{printer-name} is used |
| 737 | for non-PostScript printing. (There are two distinct sets of |
| 738 | variables in case you have two printers attached to two different |
| 739 | ports, and only one of them is a PostScript printer.) |
| 740 | |
| 741 | The default value of the variable @code{ps-lpr-command} is @code{""}, |
| 742 | which causes PostScript output to be sent to the printer port specified |
| 743 | by @code{ps-printer-name}, but @code{ps-lpr-command} can also be set to |
| 744 | the name of a program which will accept PostScript files. Thus, if you |
| 745 | have a non-PostScript printer, you can set this variable to the name of |
| 746 | a PostScript interpreter program (such as Ghostscript). Any switches |
| 747 | that need to be passed to the interpreter program are specified using |
| 748 | @code{ps-lpr-switches}. (If the value of @code{ps-printer-name} is a |
| 749 | string, it will be added to the list of switches as the value for the |
| 750 | @code{-P} option. This is probably only useful if you are using |
| 751 | @code{lpr}, so when using an interpreter typically you would set |
| 752 | @code{ps-printer-name} to something other than a string so it is |
| 753 | ignored.) |
| 754 | |
| 755 | For example, to use Ghostscript for printing on the system's default |
| 756 | printer, put this in your @file{.emacs} file: |
| 757 | |
| 758 | @example |
| 759 | (setq ps-printer-name t) |
| 760 | (setq ps-lpr-command "D:/gs6.01/bin/gswin32c.exe") |
| 761 | (setq ps-lpr-switches '("-q" "-dNOPAUSE" "-dBATCH" |
| 762 | "-sDEVICE=mswinpr2" |
| 763 | "-sPAPERSIZE=a4")) |
| 764 | @end example |
| 765 | |
| 766 | @noindent |
| 767 | (This assumes that Ghostscript is installed in the |
| 768 | @file{D:/gs6.01} directory.) |
| 769 | |
| 770 | @node Windows Fonts |
| 771 | @section Specifying Fonts on MS-Windows |
| 772 | @cindex font specification (MS Windows) |
| 773 | |
| 774 | Starting with Emacs 23, fonts are specified by their name, size |
| 775 | and optional properties. The format for specifying fonts comes from the |
| 776 | fontconfig library used in modern Free desktops: |
| 777 | |
| 778 | @example |
| 779 | [Family[-PointSize]][:Option1=Value1[:Option2=Value2[...]]] |
| 780 | @end example |
| 781 | |
| 782 | The old XLFD based format is also supported for backwards compatibility. |
| 783 | |
| 784 | Emacs 23 supports a number of backends. Currently, the @code{gdi} |
| 785 | font backend is supported on Windows. |
| 786 | |
| 787 | @cindex font properties (MS Windows) |
| 788 | @noindent |
| 789 | Optional properties common to all font backends on MS-Windows are: |
| 790 | |
| 791 | @table @code |
| 792 | |
| 793 | @vindex font-weight-table @r{(MS-Windows)} |
| 794 | @item weight |
| 795 | Specifies the weight of the font. Special values @code{light}, |
| 796 | @code{medium}, @code{demibold}, @code{bold}, and @code{black} can be specified |
| 797 | without @code{weight=} (e.g., @kbd{Courier New-12:bold}). Otherwise, |
| 798 | the weight should be a numeric value between 100 and 900, or one of the |
| 799 | named weights in @code{font-weight-table}. If unspecified, a regular font |
| 800 | is assumed. |
| 801 | |
| 802 | @vindex font-slant-table @r{(MS-Windows)} |
| 803 | @item slant |
| 804 | Specifies whether the font is italic. Special values |
| 805 | @code{roman}, @code{italic} and @code{oblique} can be specified |
| 806 | without @code{slant=} (e.g., @kbd{Courier New-12:italic}). |
| 807 | Otherwise, the slant should be a numeric value, or one of the named |
| 808 | slants in @code{font-slant-table}. On Windows, any slant above 150 is |
| 809 | treated as italics, and anything below as roman. |
| 810 | |
| 811 | @item family |
| 812 | Specifies the font family, but normally this will be specified |
| 813 | at the start of the font name. |
| 814 | |
| 815 | @item pixelsize |
| 816 | Specifies the font size in pixels. This can be used instead |
| 817 | of the point size specified after the family name. |
| 818 | |
| 819 | @item adstyle |
| 820 | Specifies additional style information for the font. |
| 821 | On MS-Windows, the values @code{mono}, @code{sans}, @code{serif}, |
| 822 | @code{script} and @code{decorative} are recognized. These are most useful |
| 823 | as a fallback with the font family left unspecified. |
| 824 | |
| 825 | @vindex w32-charset-info-alist |
| 826 | @item registry |
| 827 | Specifies the character set registry that the font is |
| 828 | expected to cover. Most Truetype and Opentype fonts will be unicode fonts |
| 829 | that cover several national character sets, but you can narrow down the |
| 830 | selection of fonts to those that support a particular character set by |
| 831 | using a specific registry from @code{w32-charset-info-alist} here. |
| 832 | |
| 833 | @item spacing |
| 834 | Specifies how the font is spaced. The @code{p} spacing specifies |
| 835 | a proportional font, and @code{m} or @code{c} specify a monospaced font. |
| 836 | |
| 837 | @item foundry |
| 838 | Not used on Windows, but for informational purposes and to |
| 839 | prevent problems with code that expects it to be set, is set internally to |
| 840 | @code{raster} for bitmapped fonts, @code{outline} for scalable fonts, |
| 841 | or @code{unknown} if the type cannot be determined as one of those. |
| 842 | @end table |
| 843 | |
| 844 | @cindex font properties (MS Windows gdi backend) |
| 845 | Options specific to @code{GDI} fonts: |
| 846 | |
| 847 | @table @code |
| 848 | |
| 849 | @cindex font scripts (MS Windows) |
| 850 | @cindex font unicode subranges (MS Windows) |
| 851 | @item script |
| 852 | Specifies a unicode subrange the font should support. |
| 853 | |
| 854 | The following scripts are recognized on Windows: @code{latin}, @code{greek}, |
| 855 | @code{coptic}, @code{cyrillic}, @code{armenian}, @code{hebrew}, @code{arabic}, |
| 856 | @code{syriac}, @code{nko}, @code{thaana}, @code{devanagari}, @code{bengali}, |
| 857 | @code{gurmukhi}, @code{gujarati}, @code{oriya}, @code{tamil}, @code{telugu}, |
| 858 | @code{kannada}, @code{malayam}, @code{sinhala}, @code{thai}, @code{lao}, |
| 859 | @code{tibetan}, @code{myanmar}, @code{georgian}, @code{hangul}, |
| 860 | @code{ethiopic}, @code{cherokee}, @code{canadian-aboriginal}, @code{ogham}, |
| 861 | @code{runic}, @code{khmer}, @code{mongolian}, @code{symbol}, @code{braille}, |
| 862 | @code{han}, @code{ideographic-description}, @code{cjk-misc}, @code{kana}, |
| 863 | @code{bopomofo}, @code{kanbun}, @code{yi}, @code{byzantine-musical-symbol}, |
| 864 | @code{musical-symbol}, and @code{mathematical}. |
| 865 | |
| 866 | @cindex font antialiasing (MS Windows) |
| 867 | @item antialias |
| 868 | Specifies the antialiasing to use for the font. The value @code{none} |
| 869 | means no antialiasing, @code{standard} means use standard antialiasing, |
| 870 | @code{subpixel} means use subpixel antialiasing (known as Cleartype on Windows), |
| 871 | and @code{natural} means use subpixel antialiasing with adjusted spacing between |
| 872 | letters. If unspecified, the font will use the system default antialiasing. |
| 873 | @end table |
| 874 | |
| 875 | @node Windows Misc |
| 876 | @section Miscellaneous Windows-specific features |
| 877 | |
| 878 | This section describes miscellaneous Windows-specific features. |
| 879 | |
| 880 | @vindex w32-use-visible-system-caret |
| 881 | @cindex screen reader software, MS-Windows |
| 882 | The variable @code{w32-use-visible-system-caret} is a flag that |
| 883 | determines whether to make the system caret visible. The default is |
| 884 | @code{nil}, which means Emacs draws its own cursor to indicate the |
| 885 | position of point. A non-@code{nil} value means Emacs will indicate |
| 886 | point location by the system caret; this facilitates use of screen |
| 887 | reader software. When this variable is non-@code{nil}, other |
| 888 | variables affecting the cursor display have no effect. |
| 889 | |
| 890 | @iftex |
| 891 | @inforef{Windows Misc, , emacs}, for information about additional |
| 892 | Windows-specific variables in this category. |
| 893 | @end iftex |
| 894 | |
| 895 | @ifnottex |
| 896 | @vindex w32-grab-focus-on-raise |
| 897 | @cindex frame focus policy, MS-Windows |
| 898 | The variable @code{w32-grab-focus-on-raise}, if set to a |
| 899 | non-@code{nil} value causes a frame to grab focus when it is raised. |
| 900 | The default is @code{t}, which fits well with the Windows default |
| 901 | click-to-focus policy. |
| 902 | |
| 903 | @vindex w32-list-proportional-fonts |
| 904 | The variable @code{w32-list-proportional-fonts} controls whether |
| 905 | proportional fonts are included in the font selection dialog. If its |
| 906 | value is non-@code{nil}, these fonts will be included. The default is |
| 907 | @code{nil}. |
| 908 | @end ifnottex |
| 909 | |
| 910 | @ifnottex |
| 911 | @include msdog-xtra.texi |
| 912 | @end ifnottex |
| 913 | |
| 914 | @ignore |
| 915 | arch-tag: f39d2590-5dcc-4318-88d9-0eb73ca10fa2 |
| 916 | @end ignore |