|
|
1.1 root 1: ============================== release 1.3 ==============================
2:
3: Fri Feb 17 18:49:01 1989 Kevin Layer (layer at snooze)
4:
5: * moved ChangeLog to parent directory of fi/.
6:
7: * include patch43.{68k,sparc} in distribution. This is so that
8: the following feature will work (otherwise the output has
9: duplicate characters, since unpatched Allegro CL echo streams
10: stutter).
11:
12: * a new variable
13: fi:echo-evals-from-buffer-in-listener-p
14: which defaults to nil (since the necessary patch to Allegro CL is
15: not yet widely distributed) and controls whether or not lisp echos
16: what is evaluated from emacs lisp source buffers. Setting this to
17: nil (default is t) will cause the feature to be disabled.
18: Files: sublisp.el, patch43.*
19:
20: Tue Feb 14 17:13:54 1989 Kevin Layer (layer at snooze)
21:
22: >>> * added two new files: clman.el and clman-oblist.el. These are
23: the emacs/on-line manual interface and are now part of this more
24: general interface. Also, there is a directory called `manual'
25: which contains the database for the on-line manual.
26:
27: The entry points are fi:clman and fi:clman-apropos. Nothing other
28: than a binding for these two functions is needed for their use
29: (they are `interactive').
30:
31: * added autoloads for clman stuff
32: File: site-init.el
33:
34: * new function, fi::find-path, takes a string and looks through
35: load-path for that file or directory and returns the pathname
36: (absolute), if found.
37: File: utils.el
38:
39: Sat Nov 26 00:37:01 1988 Kevin Layer (layer at snooze)
40:
41: * new variable, fi:emacs-to-lisp-transaction-directory, which
42: controls where the transaction files are kept. This is especially
43: useful for running lisp on remote machines (i.e., using
44: fi:remote-common-lisp).
45: Files: keys.el and subproc.el
46:
47: Tue Nov 22 20:17:56 1988 Kevin Layer (layer at snooze)
48:
49: * ipc.cl: don't load from the C library unless absolutely
50: necessary! SunOS 4.0 on the Sun4 takes a bogus amount of time to
51: dynamically load under certain conditions.
52:
53: * fix the way fi:XXX-image-name is handled (again, from 11/18/88).
54: Now it is `reduced' before displaying the buffer but it is NOT
55: done when the process already exists.
56: File: subproc.el
57:
58: Mon Nov 21 13:54:17 1988 Kevin Layer (layer at snooze)
59:
60: * fi::subprocess-hack-directory renamed to
61: fi::subprocess-watch-for-special-commands, to which smarts were
62: added about tracking package changes in inferior lisp modes. The
63: new variable fi:common-lisp-package-regexp is used for this
64: purpose.
65: File: modes.el, keys.el and subproc.el
66:
67: * fix fi:lisp-complete-symbol in the case of an explicit package
68: reference.
69: File: keys.el
70:
71: * made the lisp evalserver more bullet-proof (output of more than
72: one process read wasn't doing the right thing).
73: File: tcplisp.el
74:
75: Fri Nov 18 20:34:39 1988 Kevin Layer (layer at snooze)
76:
77: * in fi::make-subprocess, funcall mode after starting process;
78: when fi:XXX-image-name is a cons, it used to be funcall'd after
79: the buffer was displayed. It is now done before the process it
80: created.
81: File: subproc.el
82:
83: * changed default service number from 1123 (which is taken on some
84: machines around here) to 6789.
85: Files: tcplisp.el, spec.n, and ipc.cl
86:
87: Thu Nov 17 12:15:57 1988 Kevin Layer (layer at snooze)
88:
89: * two new files in this directory: ipc.cl and emacs.cl. To ipc.cl
90: added open-network-stream; to emacs.cl added list-all-completions
91: which is used by M-TAB in lisp modes to complete a Common Lisp
92: symbol!
93:
94: * Implemented fi:eval-in-lisp with a lisp eval server over a tcp
95: socket and a new binding for M-TAB (fi:lisp-complete-symbol).
96: Files: tcplisp.el and keys.el
97:
98: * new file clinit.cl, which contains the preferred way to startup
99: the emacs/lisp interface. It requires that EMACSLIBRARY be set to
100: the root of the emacs tree (/usr/local/lib/emacs on almost all
101: machines). The following should be put in .login:
102:
103: setenv EMACSLIBRARY /usr/local/lib/emacs
104:
105: where /usr/local/lib/emacs is replaced by the appropriate
106: pathname.
107:
108: Thu Nov 3 17:08:21 1988 Kevin Layer (layer at snooze)
109:
110: * filename completion as first input to a buffer now works.
111: File: filec.el
112:
113: * new variable `fi:common-lisp-file-types', which is a list of the
114: file extentions for files automatically put in
115: fi:common-lisp-mode. This variable should be set before
116: fi/site-init is loaded.
117: File: modes.el
118:
119: >>> * new variable `fi:rlogin-initial-input', which is the initial
120: input to an rlogin process (after the first prompt is seen)
121: File: rlogin.el
122:
123: Thu Aug 11 21:24:19 1988 Kevin Layer (layer at akbar)
124:
125: * fixed typo in fi:explicit-rlogin; TERM set to `dumb' before
126: calling rlogin.
127: File: rlogin.el
128:
129: Sat Jul 23 12:09:56 1988 Kevin Layer (layer at akbar)
130:
131: * fix another bug in directory tracking code (wasn't handling
132: "pushd ...; ..." correctly.
133: File: subproc.el
134:
135: >>> * added command completion to the file name completion code. The
136: entry point is now fi:shell-do-completion, and the following added
137: to fi:subprocess-mode-hook will do the trick
138:
139: (define-key (current-local-map) "\C-i" 'fi:shell-do-completion)
140:
141: File: filec.el
142:
143: Thu Jul 21 23:47:31 1988 Kevin Layer (layer at akbar)
144:
145: * directory tracking now handles "cd ... ; ..." and ignore
146: (rightly so) "cd ... &" since the latter is executed in a
147: subshell.
148: File: subproc.el
149:
150: * M-. package hacking: do source-file on proper symbol by doing a
151: find-symbol with `fi:package' as the package argument.
152: File: ltags.el
153:
154: Mon Jul 18 11:31:28 1988 Kevin Layer (layer at akbar)
155:
156: * fixed fi:explicit-remote-common-lisp, which was calling
157: fi::make-subprocess with too many arguments.
158: File: subproc.el
159:
160: Sat Jul 16 16:22:02 1988 Kevin Layer (layer at akbar)
161:
162: * fix fi::get-default-symbol to send the proper package info to
163: lisp.
164: File: keys.el
165:
166: ============================== release 1.2 ==============================
167:
168: This release put in ~ftp/pub/fi/gnudist-1.2-tar.Z on ucbvax and ucbarpa
169:
170: =========================================================================
171:
172: Mon Jul 11 23:33:15 1988 Kevin Layer (layer at akbar)
173:
174: * if there is a space in the tag name then don't query lisp. In
175: the future we may handle function specs, but for now only symbols
176: are done correctly.
177: File: ltags.el
178:
179: * fixed up fi::subprocess-hack-directory to handle `pushd +n' and
180: `popd +n'.
181: File: subproc.el
182:
183: * have filename completion do environment variable substitution
184: File: filec.el
185:
186: ============================== release 1.1 ==============================
187:
188: This release on the Allegro CL 3.0 tapes for sun3 and sun4.
189:
190: =========================================================================
191:
192: Tue Jun 7 14:49:17 1988 Kevin Layer (layer at akbar)
193:
194: * reading form for M-W from minibuffer caused an error.
195: File: keys.el
196:
197: Wed May 25 13:48:02 1988 Kevin Layer (layer at akbar)
198:
199: * fix bug (that's been buzzing jdi) in
200: fi:inferior-lisp-newline--RET in the middle of the buffer was
201: copying the wrong input to the end.
202: File: keys.el
203:
204: * fix bug in fi:set-associated-sublisp (" not backquoted in
205: docstring)
206: File: sublisp.el
207:
208: ============================== release 1.0 ==============================
209:
210: This release given to selected customers, and really constitutes a beta
211: release of this software.
212:
213: The compressed tar file is in arpa:~ftp/pub/layer/dist-1.0.tar.Z
214:
215: =========================================================================
216:
217: Wed May 25 10:54:27 1988 Kevin Layer (layer at akbar)
218:
219: * have M-sh-{M,F,D,A,W} commands read a form or symbol from the
220: minibuffer if parsing the current buffer around the point fails.
221: File: keys.el
222:
223: Thu May 19 15:08:00 1988 Kevin Layer (layer at akbar)
224:
225: * new function: fi:remove-all-temporary-lisp-transaction-files
226:
227: * minor bugs
228:
229: Tue May 17 17:22:59 1988 Kevin Layer (layer at akbar)
230:
231: * change in bindings:
232:
233: in all subprocess modes:
234: C-m bound to fi:subprocess-input-region
235:
236: in inferior-lisp mode:
237: RET is bound to a new version of fi:inferior-lisp-newline
238: which, when not at the end of the buffer, sends `input'
239: surrounding the point to the lisp process (also copying it to
240: the end of the buffer).
241:
242: Mon May 16 15:10:02 1988 Kevin Layer (layer at akbar)
243:
244: * fi::make-process and fi::make-tcp-connection use the default
245: directory of the buffer from which they were invoked.
246:
247: Fri May 13 10:27:46 1988 Kevin Layer (layer at akbar)
248:
249: * frontdoor equivalents to M-A, M-D, M-F, M-M commands.
250:
251: Thu May 12 20:01:11 1988 Kevin Layer (layer at akbar)
252:
253: * if fi:MUMBLE-image-name is not a string, then it is funcall'd to
254: yield the real image name.
255:
256: * fi:pop-input and fi:push-input arguments are now optional.
257:
258: * new variable `fi:pop-to-sublisp-buffer-after-lisp-eval'
259: controls whether or not `eval' commands cause one to be put in the
260: sublisp buffer.
261:
262: * `common-lisp-mode' and `lisp-mode' are outta here. To do this,
263: though, we redefine `set-auto-mode' so we can capture the mode in
264: the -*- line.
265:
266: * we now downcase the package name in the -*- line and the package
267: can be a list, in which case we just take the `car' of it.
268:
269: * lisp tag stuff revamped so that package info doesn't hang
270: around.
271:
272: * New function: fi:lisp-find-tag-other-window.
273:
274: * TAB, in subprocess modes, is no longer bound to
275: fi:shell-file-name-completion, but rather is bound to
276: lisp-indent-line.
277:
278: * fi:inferior-lisp-newline does a fi:inferior-lisp-send-sexp-input
279: instead of a fi:subprocess-send-input when in the middle of a
280: sublisp buffer.
281:
282: Wed May 11 14:07:51 1988 Kevin Layer (layer at akbar)
283:
284: * fi:lisp-find-tag, if given a prefix argument, will do a
285: tags-loop-continue.
286:
287: * fi::get-cl-symbol renamed to fi::get-default-symbol, greatly
288: repaired in the process.
289:
290: * fi:lisp-macroexpand and fi:lisp-walk now parse the sexp in the
291: current buffer (at the point) like fi:inferior-lisp-send-sexp-input.
292:
293: * fi:set-associated-sublisp-buffer now sets the sublisp buffer
294: for eval commands for all buffers which are in the same mode as
295: the current buffer.
296:
297: * fi:subprocess-kill-output no longer insert `input' of
298: "[output flushed]" into subprocess buffer.
299:
300: * `fi:inferior-lisp-send-sexp-input' and
301: `fi:inferior-lisp-send-list-input' again have bindings in
302: fi:inferior-common-lisp and fi:tcp-common-lisp modes (ESC RET and
303: C-x RET, respectively).
304:
305: * The fi:another-MUMBLE functions have been removed. In
306: their place fi:MUMBLE has been enhanced and fi:explicit-MUMBLE has
307: been added.
308:
309: * fi:explicit-MUMBLE-file-name and
310: fi:explicit-MUMBLE-image-arguments are outta here.
311:
312: * the fi:default-MUMBLE-file-name and
313: fi:default-MUMBLE-image-arguments are now fi:MUMBLE-image-name
314: (Note: `file-name' -> `image-name') and fi:MUMBLE-image-arguments.
315:
316: * fi:remote-MUMBLE and fi:explicit-remote-MUMBLE functions added.
317:
318: * the TCP lisp functionality of fi:common-lisp was moved to
319: fi:tcp-common and fi:explicit-tcp-common.
320:
321: * As you have undoubtedly noticed, the new form of the interface
322: is:
323: fi:common-lisp
324: fi:remote-common-lisp
325: fi:tcp-common-lisp
326:
327: fi:explicit-common-lisp
328: fi:explicit-remote-common-lisp
329: fi:explicit-tcp-common-lisp
330:
331: all of the above functions take a prefix argument, a number, which
332: is the buffer number. If the buffer number is > 0, then the buffer
333: named `*MUMBLE-N*' is used. The default is 1, in which case
334: the `-1' is dropped. If the buffer number is < 0, then N is
335: chosen to be as small as possible while not pointing to an
336: existing buffer with a running process.
337:
338: * the functions which start subprocesses and connections do not
339: reset the input ring.
340:
341: * internal name change: tcp-lisp changed to tcp-common-lisp.
342: There was mention of fi:tcp-lisp-mode in the document, and that is
343: now fi:tcp-common-lisp-mode.
344:
345: Mon May 9 17:43:22 1988 Kevin Layer (layer at akbar)
346:
347: * remove all remaining indenting stuff from this interface
348: (actually, move it to ../local/indent.el).
349: File: modes.el
350:
351: * add binding for who-calls (M-C, fi:lisp-who-calls)
352: File: modes.el
353: Note: this change will not work with allegro CL 3.0 (who calls is
354: not in that lisp)
355:
356: Fri Apr 29 15:37:18 1988 Kevin Layer (layer at akbar)
357:
358: * removed from fi::make-process the "image name from the
359: environment" hack. Not even Harry, the original author of this
360: feature, uses it.
361: File: subproc.el
362:
363: * cl-indent is no longer loaded, since This interface has nothing
364: to do with indenting.
365: File: site-init.el
366:
367: Mon Apr 25 15:14:59 1988 Kevin Layer (layer at akbar)
368:
369: * fi::unix-domain-socket renamed to fi:unix-domain-socket
370: (exported)
371: File: subproc.el
372:
373: * fi:emacs-lisp-mode moved from file modes.el to elisp.el
374:
375: * The functions fi:inferior-lisp-send-sexp-input and
376: fi:inferior-lisp-send-list-input are no longer bound in inferior
377: lisp modes, since they didn't work (at least I couldn't figure
378: them out). They were previously bound to the keys M-RET and
379: C-x RET.
380:
381: * fi::package renamed fi:package (exported)
382:
383: * generic `lisp' mode removed (since nothing referred to it
384:
385: * lisp-mode and common-lisp-mode are the same as
386: fi:common-lisp-mode (this is for -*- mode defs)
387:
This archive runs on limited infrastructure. Preserving old code on modern bandwidth. Automated agents are requested to crawl responsibly.