]> code.delx.au - gnu-emacs-elpa/blob - NEWS
*** empty log message ***
[gnu-emacs-elpa] / NEWS
1 Changes since 3.9.3:
2
3 * Indentation rules improved. New customizable variable
4 `sml-rightalign-and'. Also `sml-symbol-indent' is now customizable.
5
6 * New add-log support (try C-x 4 a from within an SML function).
7
8 * Imenu support
9
10 * sml-bindings has disappeared.
11
12 * The code skeletons are now abbrevs as well.
13
14 * A new *sml* process is sent the content of sml-config-file (~/.sml-proc.sml)
15 if it exists.
16
17 * `sml-compile' works yet a bit differently. The command can begin
18 with `cd "path";' and it will be replaced by OS.FileSys.chDir.
19
20 * run-sml now pops up the new buffer. It can also run the command on another
21 machine. And it always prompts for the command name. Use a prefix
22 argument if you want to give args or to specify a host on which
23 to run the command.
24
25 * mouse-2 to yank in *sml* should work again (but won't work for next-error
26 any more).
27
28 * New major-modes sml-cm-mode, sml-lex-mode and sml-yacc-mode.
29
30 * sml-load-hook has disappeared as has inferior-sml-load-hook.
31
32 * sml-mode-startup.el is now automatically generated and you're supposed to
33 `load' it from .emacs or site-start.el.
34
35 * Minor bug fixes.
36
37 Changes since 3.3:
38
39 * the sml-drag-* commands have disappeared.
40
41 * added a little bit of `customize' support. Many of the customization
42 variables for indentation are still in flux, so they are not customize'd.
43
44 * proformas have been replaced by skeletons. it's mostly the same as
45 before (the layout has slightly changed, tho). The main difference
46 is that the indentation relies on the major-mode indentation so it
47 is implicitly customized, which makes more sense to me.
48 Also I added an electric space M-SPC that will call the corresponding
49 skeleton if any matches the immediately preceding symbol. Basically
50 that allows you to type `l e t M-SPC' to call the `let' skeleton.
51
52 * M-C-f and M-C-b try to be smart and jump around let..end and such blocks.
53 It's probably either too smart or not smart enough, tho.
54
55 * there is no more sml-<compiler>.el since the code should work for "all"
56 known compilers. If your favorite compiler doesn't seem to work right
57 send me a sample session.
58
59 * hilite support has disappeared and font-lock and menu support is now built-in.
60
61 * the indentation algorithm is inherently much slower. I've tried to ensure
62 the slowness never manifests itself in practice, but if you find a case
63 where the indentation doesn't feel instantaneous, tell me.
64
65 * function arguments get properly indented (yes, madam).
66
67 * the indentation has been majorly reworked. The list of changes is too long.
68 Many customizations have disappeared, some may reappear depending on the
69 feedback I get. The indentation should now "always" work right, so
70 tell me when it doesn't.
71
72 * nested comments are only properly handled if you have a nested-comments
73 aware Emacs (I don't know of any yet) or if you turn on font-lock.
74
75 * provide `sml-compile' which does something similat to `compile' except it
76 passes the command to an inferior-sml process. Also it has an additional
77 hack to look for sml-make-file-name in parent directories and cd to it
78 before sending the command (handy for CM.make() when the sources.cm file
79 is not in the current directory). This hack is very ad-hoc and quite
80 misleading for people who don't use CM. I.e. the default is not safe.
81
82 * sml-send-region and friends now always use a temp file. The temp file
83 management has been made a little more secure.
84
85 * the overlay is now turned off by default. Instead the region is activated,
86 so that transient-mark-mode will end up highlighting the error just like
87 the overlay used to do.
88
89 * sml-proc uses compile.el for error parsing. This mostly means that instead
90 of C-c ` you want to use the standard C-x `. It also means that error
91 formats for any compiler can be added more easily.
92
93 * The special frame handling has been thrown out because it doesn't interact
94 well with Emacs' own similar feature. I believe XEmacs still doesn't provide
95 such a feature, so if you miss it, either switch to Emacs or (convince
96 someone else to) add it to XEmacs.