-
Notifications
You must be signed in to change notification settings - Fork 0
/
Changes
979 lines (653 loc) · 35.5 KB
/
Changes
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
Changes in 2.5 PL8
- fixed a hang when hitting certain 8-bit chars in the From line
- fixed a file descriptor leak (from Andreas Luik)
- removed all Expires: processing
Changes in 2.5 PL7
- a few display issues fixed (from Kenneth Reek)
- buffer overflow in frm fixed
Changes in 2.5 PL6
- Fixed message-id in the In-Reply-To header.
- Actually show the -r option in the usage
- Whacked a few more places where we might follow a symlink in tmpdir
- Fastmail didn't grok commas correctly.
- Change lock.c to not use errno unless there's an errno
Changes in 2.5 PL5
- A few more buffer overruns fixed.
Changes in 2.5 PL4
- little bug in aliasdb.c (from Scot W. Hetzel <scot@genroco.com>)
- Message-Id: could overflow (from Darin Wayrynen <darin@deru.net>)
- Fixed the permissions setting problem with the edit mailbox function
- Added a ResizeScreen() call after show_msg() calls to hopefully deal wit
SIGWINCH problems (inspired from Yuke Zhuge <zhuge@stone.pdx.informix.com>)
- Fixed problems under SunOS 4.1.3 (from Paul J. Menchini <mench@mench.com>)
- Fixed content encoding (from David Woolley <david@djwhome.demon.co.uk>)
- Attempt to stop crashes with long recipient list (from Gilles Detillieux
<grdetil@scrc.umanitoba.ca>)
Changes in 2.5 PL3
- Fixed the declaration of outbound_from_name (from Richard Jackson
<rjackson@gmu.edu>)
- Fixed the leapyear macro (from Gerhard Niklasch)
- Fixed editing of a long CC line (from Derek Godfrey <djg@sjc.ascend.com>)
- Mail that lands in the sentmail with attachments now has the
Content-disposition header text added to the message (from Dave
Spensley)
- Made allto a bit longer (suggested by Dave Spensley
<davesp@typhoon.xnet.com>)
- Fixed Configure to find (again) fcntl on Linux
- can_access will now allow you to write to a char device (/dev/null)
Changes in 2.5 PL2
- Fixed (again, maybe?) savebyalias from Adam Benjamin <aebenjam@opentext.com>
- Fixed bug were an alias delete would crash elm if the mailbox was empty
Changes in 2.5 PL1
- Fixed savebyalias from Michael D Hughes <mdh@logcabin.woods.bridge.com>
and Robert Schwartz <rjs@elt.com>
- Changes for 8 bit clean from H.J. Lu <hjl@varesearch.com>
Changes in 2.5 PL0
- One minor change to make an easier compile on SparcLinux.
Changes in 2.5 PL0pre8
- Y2K stuff from Ed Symanzik <zik@msu.edu>
Changes in 2.5 PL0pre7
- minor change to deal with egcs
Changes in 2.5 PL0pre6
- newalias nows does a geteuid() instead of getuid(). This is to lets
SysAdmins do "su <user> -c 'newalias'"
Changes in 2.5 PL0pre5
- Fix for GECOS entries with comma-delimited sub-fields from Chip
Rosenthal <chip@unicom.com>.
- Fix? for keys getting screw up after a ^Z from Chip Rosenthal
<chip@unicom.com>.
Changes in 2.5 PL0pre4
- A few changes for BSD from Chip Rosenthal <chip@unicom.com>.
- A few changes to Configure suggested by Chip Rosenthal <chip@unicom.com>.
Changes in 2.5 PL0pre3
- Configure didn't see fcntl locking on Linux. Fix from Sean McDermott
<spm@intel.com>
Changes in 2.5 beta 3
- Fixed a problem where elm would try to close a file twice when
sending an attachment (from reichler@ews.uiuc.edu)
- Added docs for environment variables to the man page (from Eric
Haszlakiewicz <nimenees@uaccess.net>)
Changes in 2.5 beta 2
- IncomingFolders no longer set Magic Mode on for everything
- Configure should now deal with Linux.
- Make //pattern honor current limit (from Keith Neufeld<neufeld@pvi.org>)
- year 2000 fix.....?
- Building on Solaris requires /usr/ccs/bin to be in the PATH
Changes in 2.5 beta 1
- b)ounce honor DONT_ADD_FROM (from Keith Neufeld<neufeld@pvi.org>)
- removed [encode] display and reply (from Keith Neufeld<neufeld@pvi.org>)
- make right/left arrows work in the file browser the same as in the
message index. (from Keith Neufeld<neufeld@pvi.org>)
- fixed a bug in the incomingfolders code (it wouldn't save) (from Bob Schwartz
<rjs@elt.com>)
- fixed the meaning of noheader for forwards (from Keith Neufeld<neufeld@pvi.org>)
Changes in 2.5 alpha 11
- fixed a bug in the alias code (thanks to Keith Neufeld<neufeld@pvi.org>)
- fixed the generation of temp file names in newmbox.c
- allowed longer Cc: lines. This was due to a crash with a group reply.
- added incomingfolders to the elmrc. This is a list of folders that will
automatically be opened with "magic mode" on. (inspired by a patch from
Cristophe Kalt <kalt@ensta.fr>)
Changes in 2.5 alpha 10
- Press u on a message that is read but not deleted will mark the message as
unread (from Larry Schwimmer <rosebud@cyclone.Stanford.EDU>).
- mmencode.1 man page is now installed with others (from Gregory Margo
<gmargo@newton.vip.best.com>).
- Magic mode stuff is now in help and NLS (from Gregory Margo
<gmargo@newton.vip.best.com>).
- allow elimination of aliases from larger aliases (To: bigalias -subalias).
Based on incomplete patch by Waldi Ravens <waldi@moacs.indiv.nl.net>. Also
removes redundant declarations of get_alias_address(), which is declared in
elm_globals.h (from Keith Neufeld<neufeld@pvi.org>)
- On an entry field that normally clears to be retyped from scratch as soon
as the user presses a key (Subject: on header edit screen), allow right
arrow and delete to begin editing the existing default answer. (from Keith
Neufeld<neufeld@pvi.org>)
- Add a warning to Configure about disabling shell escapes (from Keith
Neufeld<neufeld@pvi.org>)
- Eliminate complaints in doc/Makefile about duplicate makefile entries if
mansrc and catmansrc are both "none" (from Keith Neufeld<neufeld@pvi.org>)
- Allow entry of uppercase commands on header display/edit screens, since
that's how they're parenthesized on the display. (from Bob Schwartz
<rjs@elt.com> and Keith Neufeld <neufeld@pvi.org>
- Make 2+ -line error messages write over more of the screen instead of
scrolling it. (from Bob Schwartz <rjs@elt.com> and Keith Neufeld
<neufeld@pvi.org>
- Don't try to send a message if the user erased all the recipients on the
header screen. (from Bob Schwartz <rjs@elt.com> and Keith Neufeld
<neufeld@pvi.org>
- Extend save-by-name behavior to allow save-by-alias (if the user has an
alias for this address). (from Bob Schwartz <rjs@elt.com> and Keith Neufeld
<neufeld@pvi.org>
- Elm munged the permissions of a mailbox when not using system call locking
and the mailbox was ordinary in all regards except for being on another
filesystem than the temp mailbox. Fixed. (from Bob Schwartz <rjs@elt.com>
and Keith Neufeld <neufeld@pvi.org>. Various other folks have also
submitted patches that address this problem.
- Fix a bug in frm error message (from Bob Schwartz <rjs@elt.com>
and Keith Neufeld <neufeld@pvi.org>
Changes in 2.5 alpha 9
======================
- Fixed a write call in lock.c (from Paul Close <pdc@lunch.engr.sgi.com>)
- When creating a new file, always make it so that symlinks aren't followed.
This probably should only be done for temp files...
- Add full name comments to the Cc: line of a group reply, and removes the
To: recipient from the Cc: list (from Keith Neufeld<neufeld@pvi.org>)
- Make the keys on the status change screen change the status, instead of
taking you up into the row to change the status with the space bar. Also
add he status change configuration to "elm -v" (from Keith
Neufeld<neufeld@pvi.org>)
- In fastmail: fix the argument list generated to invoke the mailer to be
space-separated rather than comma-separated; and change access() to
can_access() (from Keith Neufeld<neufeld@pvi.org>)
- Fixed a problem with alt editor introduced in alpha 8 (from Robert
Schwartz <rjs@elt.com>)
- Added vi-style `H'' and `L'' commands to go to the top |and bottom of the
current indexed (Index, Alias) screen (from Keith Neufeld<neufeld@pvi.org>)
- More changes by Chip Rosenthal:
Totally new curses subsystem.
Encapsulated folder state into a (FOLDER) datatype. Removes a
gadzillion globals, and paves the way towards multiple
folder support.
Maintain alias state separately, introducing stuff like num_aliases
and curr_alias, to avoid the incestuous sharing of globals
between folders and aliases. This required some ugly
changes to the menu stuff since they can no longer, for
instance, just look at current -- they need to distinguish
whether being run in folder or alias mode. This likely is
an interim step, Keith has mumbled stuff about replacing
the wretched menu code.
Total rewrite of the Elm startup code. Of particular note is that
Elm entry into and exit from raw in a more controlled and
consistent fashion.
Total rewrite of the Elm shutdown code. Replaced the dozen or so separate
procedures that all terminate Elm in slightly different ways,
with a single leave() procedure.
New debug menu entered with '@' or '#' (the two are now equivalent).
Added debug info on terminal (curses) settings.
The display_helpfile() program is now passed a just a topic (say "options")
instead of a full pathname ("/usr/local/lib/elm/help-options").
File selection browser now has smart directory/pattern changing. Instead
of separate commands, just type in what you want (directory, pattern,
or both) and it just figures it out.
When leaving the header menu for the ready-to-send screen, the display
is updated properly.
Fixed the annoying screen-redraw-every-five-minutes.
MIME_SEND has been removed. The MIME sending stuff is always on.
The MIME_RECV stuff remains an option, controlled by the
Configure answer to the MIME question.
Instead of just assuming base64 on binary data, see if quoted-printable
will produce smaller encoded output.
New Makefiles. There now is a Make.defs.SH file that includes all of
the definitions (CC, BINDIR, etc.) plus the standard rules
(all, clean, etc.). The Makefile.SH files just define the
list of targets, and the specific rules to build/install them.
The Makefile.SH includes Make.defs plus depend.out to generate
the resulting Makefile. Thus Makefile.SH should be much smaller
and simpler, and the subdirectory makefiles should now work
consistently.
New "allow_setuid" setting. This must be enabled in the global elmrc in
order for Elm to operate setuid. This prevents naive setuid
installation without being too heavy handed.
Depreciated "keypad" setting. New curses system knows whether or not
to use HP keypad by termcap settings.
Reorganization of main header files. The elmutil.h header is now
obsolete. All portions of the Elm package include elm_defs.h,
and the main Elm program (but not utilities) include elm_globals.h.
Moved many definitions specific to Elm (and not used by the
utilities) from elm_defs.h to elm_globals.h. Placed declaration
of the dozen-or-so widely used globals in elm_defs.h.
Major overhaul of the global namespace. Terminal state is now encapsulated
into (struct term_info). Folder state is now encapsulated
into (FOLDER). Most parameters incestuously shared between
folder and alias now split apart (e.g. added num_aliases,
curr_aliases). New "opmode" parameter indicates Elm
operating mode (OPMODE_READ, OPMODE_SEND, etc.).
New hdrs/port_termios.h encapsulates terminal setting commands in
a POSIX wrapper.
sysdefs.h no longer defines individual help files, but rather a
single "system_help_dir".
New lib/fast_getpw.c. Does a "fast" getpwnam() by caching recent results.
Replaced gethostname() and gethostdomain() in lib/opt_utils.c with new
get_hostname() and get_hostdomain() in new lib/gethostname.c.
Added new init_common() to perform initializations common to Elm and
utilities in a consistent fashion.
get_word() not treats a (parenthesized comment) as a single word.
Added file selection browser and attachments menu to the message catalog.
New doc/help-selfolder with help on folder selection commands.
Reduced redundant macros, procedures, and definitions. Stuff like
use isspace() instead of whitespace(), streq() instead
of equals(), TRUE instead of YES, strchr() instead of index(),
etc.
All of the utilities now call initialize_common() rather than re-implementing
the initialization code in an inconsistent fashion.
fastmail was broke. Rewrote it.
Fixed readmsg so that when weeding is done, the default action is the
opposite of the last action. That is, with a weed list of
"Foo: Bar:" the default is reject all other headers. With
a weed list of "Foo: !Bar:" the default is accept all other
headers.
Changes in 2.5 alpha 8
======================
Lots of changes by Chip Rosenthal <chip@unicom.com>, including:
--- EXTERNAL CHANGES ---
New "make setup" target in base Makefile ensures .SH files up to date.
New "make tags" target in base Makefile creates tags for src and lib dirs.
New "mime-types" file added to $libdir. (The mime-types file I've
used basically is munged from my Gopher/HTTP server. It might
not be appropriate for the final release.)
Renamed "mlists" to "mlists.eg" in $libdir.
Renamed all help files from stuff like "elm-help.0" to "help-main".
Major overhaul of the elmrc file. Options are now grouped together
logically rather than appearing in alphabetical order. (See next
item for more info.)
Major change to $libdir/elmrc-info. Previously, this was used simply
as comments for annotating the elmrc. Now, it directs the production
of the elmrc. The old elmrc-info format was:
option-name
# Comments to associate with the option.
The new format is:
# Comments to associate with the option.
option-name
Basically, just passes the elmrc-info file through, substituting
in the "option-name = value" wherever it encounters an option
name. Appropriate checks are included so that settings missing
from elmrc-info and extraneous settings in elmrc-info are noted
with warnings. If elmrc-info is missing, we fall back on a
raw alphabetical listing.
Added "mmencode" to utils and "mmencode.1" to doc. This is an
encoder/decoder for base64 and quoted-printable. It's based on
the metamail version, which is freely redistributable.
"readmsg" now accepts "-w" to specify header weeding. The default
works like: readmsg -w 'Subject: From: To: Cc: Apparently- Date:'
The list entries are left-anchored "patmatch" patterns.
Added "printhdrs" elmrc option that allows header weeding to be
specified when printing. The value of this option becomes the
"-w" arg to "readmsg".
Before, when printing a message, Elm discarded stdout/stderr if the
"print" option had "%s" in it. Now stdout/stderr is always left
attached to the tty.
The "site hiding" stuff has been deleted.
Fixed potential problem where Elm was too aggressive stripping outside
quotes from fullnames such as these:
joe@acme.com ("Joe User :-)")
joe@acme.com ("Joe" User "IV")
Pattern matching now allows \-quoting, such as "name\*with\*stars".
Fixed NLS utilities to compile on ANSIishy systems.
Improved SIGWINCH/SIGCONT support. The places that already handled
SIGWINCH (main menu and alias menu) now redraw automatically
on SIGCONT as well. The verify_transmission menu, file browser,
and attachments menu do SIGWINCH/SIGCONT too. Also, if you do a
screen resize during a suspend, Elm now will notice.
Fixed possible problem with character reading. (The problem was that
ReadCh() was doing a read(2), while the rest of the package uses
stdio. Fixed ReadCh() to use getchar(3) instead.)
Temporarily ifdef'ed out the access() test on the ~e editor. (I'd
like to delete this entirely.) This allows you to say
"easyeditor = emacs" and not have to specify a full pathname.
Under the s)ave command, the cookie to mean "same as the last folder
I used" has been changed from "." to "!!".
Under the s)ave command, if you specify a pattern you now get the
new file selection browser instead of an "ls -C" listing. The
file selection browser also is started if you give a directory
name.
Under the c)hange command, if you specify a pattern you now get the
new file selection browser instead of an "ls -C" listing. The
file selection browser also is started if you give a directory
name.
The "fwdattribution" setting now understands "%[begin|end]" formatting.
Stuff to the left of the bar is printed in the opening attribution,
stuff to the right in the close. The %A escape is depreciated
and is slated for removal.
Elm will try to use $PAGER or sel_pager to display help files, before
falling back to its built-in help pager.
The chance of funky behavior during signals has been reduced. (In
other words, Elm was doing tons of things in its signal handlers
that are dangerous. Most, but not all, of them have been removed.)
Before, if you tried to f)orward a form Elm would allow it, but wouldn't
let you edit or anything. Now, it won't even do that. It just
gives a message prompting you to use b)ounce instead.
Elm is now MIME-sending capable right out of the box. The answer that
you give Configure with respect to the MIME question affects just
the reading side of Elm.
Outbound mail messages (and MIME multipart/mixed parts) no longer
have Content-Length: headers. Content-Length: is generated only
when messages are saved to a local folder.
The saved copy of a message will not have a Content-Type header
unless the message is the source to a form.
If you omit the encoding from an [include ...] line, Elm will guess
an appropriate encoding rather than assuming no encoding.
The content-type is now optional to [include ...]. That is, you
can say just [include filename] and Elm will guess both the content
type and encoding.
New [attach ...] command. Use exactly the same way as [include ...].
Difference is inclusions appear in-line with the message; attachments
appear at the end.
New attachments menu. To enter, after composing the message, strike
"a" at the verify_transmission menu.
The MIME headers of outbound message have been changed. Before
the headers on a part would look something like:
Content-Type: image/gif
Content-Transfer-Encoding: base64
Content-Name: /pub/archive/images/Drawing/bill.gif
Content-Length: 5376
Now they will appear something like:
Content-Type: image/gif
Content-Transfer-Encoding: base64
Content-Description: GIF picture - version 87a 640 x 350, 16 colors
Content-Disposition: attachment; filename="bill.gif"
The "textencoding" elmrc option has been dropped. Elm now determines
on its own whether a message is "8bit" or "7bit". This setting
will be ignored silently if it appears in elmrc.
Accordingly, Configure no longer asks about "defencoding".
The "charset" elmrc option is now used only if the message has 8-bit
characters. If the message contains solely 7-bit characters then
this setting is ignored and US-ASCII is used.
Accordingly, Configure now offers ISO-8859-1 and not US-ASCII as the
default "defcharset".
--- INTERNAL CHANGES - header files ---
Major reorganization of the header files. Basically here is the mapping:
old way --> new way
defs.h --> elm_defs.h
header.h + elm.h --> elm_globals.h
Here is the general idea: everything and anything in the Elm
package should include "elm_defs.h". It includes all of the
definitions, inclusions, and portability hackups. The *ONLY*
files that should include "elm_globals.h" are those in the "src"
subdirectory. It defines the globals, and contains some of the
definitions/declarations used with in Elm. (Note that there are
one or two routines in the library that require "elm_globals.h".
I'd argue either that should be fixed, or they don't belong in
the library.)
The "elm_defs.h" provides for ANSI prototyping. It says:
#ifndef P_
# ifdef ANSI_C
# define P_(ARGS) ARGS
# else
# define P_(ARGS) ()
# define const
# endif
#endif
For example, the cvt_timestr_to_hhmmss() procedure (from
lib/date_util.c) should be declared with something like:
int cvt_timestr_to_hhmmss P_((const char *, int *, int *, int *));
A new header file, "elm_lib.h" has been introduced. It contains
declarations (and prototypes) for *every* entry in libutil.a.
This header needs to be updated as changes are made to the library.
This header is included automatically via "elm_defs.h", thus should
not be included explicitly. (Exception: components of the library
that have their own header files, such as ndbz.c, are not included
in "elm_lib.h".)
Renamed "curses.h" to "elm_curses.h" to avoid name collision/confusion.
Added "port_wait.h" header. Anything that needs to use wait(2) should
include this. Then, declare a "waitstatus_t" to hold the return
status, and use the POSIX macros to test the status.
New headers "sndhdrs.h" and "sndparts.h" support the new mail sending
subsystem.
A lot of stuff from "mime.h" has been dropped, retaining basically just
that which is used by the receiving portion of Elm.
The MIME capability has been split into MIME_SEND and MIME_RECV.
MIME_SEND is hardwired on. MIME_RECV follows the Configure answer.
(It is my hope that MIME_SEND will be deleted before release, and
this stuff will just be standard. As currently written, things will
not compile if you try with MIME_SEND turned off.)
The "SHORT" macro has been dropped. It was overloaded, used both as
a string length as well as an argument to sort name types. Instead,
use TLEN (as in teeny length) to declare 10-byte strings. As far
as sort names (used to be SHORT/FULL/PAD), the padded sort names have
been dropped (I fixed the menus to do the padding themselves), so
the sort_name() procedures now accept a TRUE/FALSE boolean to
return long versus short names.
I've dropped most of the happy macros for character names. If you
want to use ':' then say that, not COLON.
Renamed INCOMING/OUTGOING to LOCK_INCOMING/LOCK_OUTGOING.
Dropped JANUARY, FEBRUARY, ... definitions from elm_defs.h.
The MAIN_HELP, OPTIONS_HELP, etc. definitions have been changed from
numbers to help file names and moved into sysdefs.h.
Rewrote the dprint() macro to fix subtle bug. This means stuff like:
if (something)
dprint();
else
foobar();
will work correctly now. (Before, foobar() would be run if
"something && !debug".)
Things in sysdefs.h/elm_defs.h/elm_globals.h were juggled about to their
appropriate home. For instance, the declaration of strstr() does
not belong in sysdefs.h. I put it in the new elm_lib.h header.
Deleted nearly all #includes from headers.h when converting over to
"elm_globals.h". This stuff belongs in "elm_defs.h".
Previously "headers.h" was used to declare the extern globals and
"elm.h" was used to create/initialize them. The new "elm_globals.h"
does both. Just declare INTERN before including the file (done
in elm.c) to create them.
Drop explicit initialization of empty strings. This means they now will
be created in BSS rather than DATA, saving dozens of kilobytes of
space in the executable and speeding startup.
All of the outbound message header globals are gone (to, subject, etc.)
The new SEND_HEADERS datatype does this.
The "forwarding" and "text_encoding" globals are gone.
The "resize_screen" global has been renamed "caught_screen_change_sig" to
reflect that it is used in SIGCONT handling as well as SIGWINCH.
One exception to the "all globals in elm_globals.h" rule: the "debug"
and "debugfile" global declarations have been moved into "elm_defs.h"
since so many things use them.
--- INTERNAL CHANGES - lib/* code changes ---
The entire library directory has been ANSIfied and checked with
"gcc -Wall". (Well...actually almost -Wall. Due to the:
/*#undef ... /* ... */
stuff, I had to turn off the nested comment checks.
The can_access() procedure has been rewritten. It now avoids an
expensive fork when unnecessary. Also, it now follows access(2)
semantics (setting proper errno status rather than returning an
errno value). (All calls to can_access() have been updated to
deal with this change.)
The error_description() routine has been dropped, and the entire
package has been modified to use strerror(). lib/errno.c has
been modified to provide a strerror() for those systems that
lack it.
figure_out_addressee() routine has been modified so that you now
have to pass the username as the second argument. Previously
it obtained this info from a global variable, which may or may
not be initialized properly if figure_out_addressee() is called
from a utility program.
The "next_field" value from parse_arpa_mailbox() now points to the
character after scanning left off rather than where the next field
begins. What this essentially means is that whitespace is no
longer skipped before setting the "next_field" value.
The patmatch() procedure now requires a third "opts" argument.
Added the following (possibly useful) routines to the library:
basename(), trim_trailing_slashes(), trim_trailing_spaces(),
and trim_quotes().
--- INTERNAL CHANGES - src/* code changes ---
The sort_name() and alias_sort_name() procedures used to take a
FULL, SHORT, or PAD argument. The menus were fixed to pad the
name displays properly, so PAD is no longer necessary. These
routines now take a TRUE (display full name) or FALSE (display
short name) argument.
Added ResizeScreen() to src/curses.c. This may be called to effect
a screen resize, and to update the associated globals.
Calling semantics to elm_date_str() [file src/date.c] changed.
New src/file_ops.c added that contains things such as file_open(),
file_close(), file_seek(), file_rename(), etc. They do exactly
what you think they do -- plus they add error checking and
print messages on failure. In other words, rather than spending
a half screenful of code to fopen() a file, just file_open() it.
The calling semantics to copy_message() have changed. It no longer
accepts a "prefix" argument. New usage is:
copy_message(FILE *dest_file, int cm_options);
The following new options have been added:
CM_PREFIX - Add "prefixchars" to each line.
CM_FORWARDING - Replaces the (now removed) "forwarding" global.
CM_ATTRIBUTION - Add an attribution line.
Since attribution generation has been merged into this routine,
the makeAttString() procedure has been moved into src/fileio.c
Calling semantics to display_helpfile() [file src/help.c] changed.
Before, it used to take an integer to select a helpfile. Now, it
takes a helpfile pathname. (These pathnames are now defined in
sysdefs.h.)
Calling semantics to read_number() [file src/in_utils.c] changed.
It used to have a hardwired screen line number and got its
default response from a magic global variable. Now you have
to pass both of these as arguments.
Calling semantics to GetPrompt() [file src/in_utils.c] changed.
Now you need to pass the timeout as a parameter, rather than using
the global "timeout". Say "0" for infinite timeout. Also, if a
SIGWINCH/SIGCONT occurs while in this procedure, the screen resize
is done automatically. Thus, the calling routine merely needs to
do a redraw. In other words, any menu that says:
cmd = ReadCh();
can get instant SIGWINCH/SIGCONT support by changing that line to:
if ((cmd = ReadCh()) == NO_OP_COMMAND) {
do_redraw = TRUE; /* or however you force a screen redraw */
continue;
}
All of the signal initializations moved out of src/init.c and into
the new initialize_signals() [file src/signals.c] routine.
The error message handling [file src/out_utils.c] is a bit
smarter. Now, for instance, the clear_error() routine performs
output only if there is an error to clear. It also returns TRUE
when it does so, telling you that the cursor has moved.
The remail() procedure (implements "b)ounce") has been moved into
src/reply.c.
The strip_commas() routine [file src/strings.c] has been modified to
elide extraneous whitespace -- not create *more* of it!
The argv_zero() routine [file src/strings.c] has been dropped. Use
the new basename() library routine.
The system_call() procedure [file src/syscall.c] will set errno
properly in the event of failure. (By failure, I mean that something
went wrong, and not that the subprocess had a non-zero status.)
If it has anything to do with sending mail, then chances are 87.43%
that it has been rewritten, replaced, deleted, or moved to
some other file.
All of the mail sending has been unified behind a new routine:
int send_message(given_to, given_cc, given_subject, mssgtype)
const char *given_to, *given_cc, *given_subject;
int mssgtype;
Supported values for the "mssgtype" argument are:
#define SM_ORIGINAL 0 /* this is an original message */
#define SM_REPLY 1 /* this is a response to the current mssg */
#define SM_FORWARD 2 /* the current mssg is being forwarded */
#define SM_FWDUNQUOTE 3 /* forward - without quoting or editing */
#define SM_FORMRESP 4 /* this is a response to a fill-out form */
- Configure now looks for /var/spool/lock (from Bernd Ernesti
<bernd@arresum.inka.de>)
- regexp.h was removed from the distribution (from Keith Neufeld<neufeld@pvi.org> )
Changes in 2.5 alpha 7
======================
- Minor tweaks to eliminate superfluous strlen() calls (from Chip
Rosenthal <chip@unicom.com>)
- Significant rewrite of rfc822_toklen() (from Chip Rosenthal <chip@unicom.com>)
- Add %)F and %>F to attribution, for from name and from address (from Paul
Close <pdc@lunch.engr.sgi.com>)
- Add header fields to the send menu. Only list to/cc/subject and any
non-null fields (from Paul Close <pdc@lunch.engr.sgi.com>)
- More efficiency changes (from Paul Close <pdc@lunch.engr.sgi.com>)
- Don't use _down termcap if it is \n, since that doesn't work most
places without mucking with stty (from Paul Close <pdc@lunch.engr.sgi.com>)
- elm won't allow the tmp mailbox to be a symlink (from Syd Weinstein
<syd@Myxa.COM>)
- removed move_old_files_to_new() since it is no long needed (from Syd Weinstein
<syd@Myxa.COM>)
- added the Status change screen (from Keith Neufeld<neufeld@pvi.org> )
- fixed a bug with the timezone stuff (from Chip Rosenthal <chip@unicom.com>)
- fixed a small bug in fwdattribution (from Chip Rosenthal <chip@unicom.com>)
- fixed readmsg to deal with long header lines (from Chip Rosenthal
<chip@unicom.com>)
Changes in 2.5 alpha 6
======================
- added handling of Apparently-To: (from Paul Close <pdc@lunch.engr.sgi.com>
- fixed a bug with MMDF & MIME (from Michael Elkins <elkins@aero.org>).
- fixed up the behaviour of ask & friends (from Keith Neufeld<neufeld@pvi.org> )
- added the ability to make entries of the form "!attribute" in a DT_MLT
list or as a DT_SYN (which makes it an antonym, but it's still processed by
DT_SYN). It does _not_ allow "!attribute" in the elmrc, which would be
silly. It reimplements autocopy as a DT_MLT of !askreplycopy and replycopy,
and it fixes alwaysleave to be a DT_MLT of alwayskeep and !alwaysstore,
removing the `kludge that should be part of the "machine"'' from
src/read_rc.c. (from Keith Neufeld<neufeld@pvi.org> )
- prlong moved from $libdir to $bindir
- corrects inconsistent names used for header file inclusion.
While most files include "foo.h", a couple used "../hdrs/foo.h".
- add a "fwdattribution" elmrc variable that enhances the handling of
forwarded messages. If undefined (the default), the current behaviour is
preserved (standard attribution and prefix added on edited messages). If
this variable is defined, the prefix is never added, and instead of the
standard attribution the inclusion is bracketed like:
Changes in 2.5 alpha 5
======================
- autocopy expanded into replycopy and askreplycopy (in elmrc). replycopy
determines whether the default should be yes or no, and askreplycopy
determines whether it should happen automatically. (from Keith Neufeld
<neufeld@pvi.org> )
- askstore can be used to separately prompt to move read messages to
=received when you leave your mailbox. I see its main use for people who
don't want to use =received but still like confirmation of deletion. (from
Keith Neufeld <neufeld@pvi.org> )
- Slight change to the behavior of resolv. (from Keith Neufeld < neufeld@pvi.org>)
- If you delete the last message in a folder that is a symlink, the folder will
be emptied (this didn't work before). (from Keith Neufeld < neufeld@pvi.org>)
- Added Chip Rosenthal's <chip@unicom.com> change to allow builds in a separate
directory.
- Elm will now no longer stomp on folders if the user is over quota (from Larry
Schwimmer <rosebud@cyclone.Stanford.EDU>).
- Changed the test for the Ultrix fflush bug.
Changes in 2.5 alpha 3
======================
- Added a change from Keith Neufeld <neufeld@pvi.org> to make the confirm for
saving tagged messages a user configurable option (confirmtagsave in elmrc).
- Flush the edit buffer after an edit and now make it a user config thing
(editflush in elmrc).
Changes in 2.5 alpha 1
======================
- Added the changes that allow other folders to be treated as spool folders
(the "Magic Mode" change). Hitting '&' toggles this as well as the use of
the command line option -M. The -M option also works on frm/nfrm.
- Added the "replied" code. You must set showreply=ON in your elmrc for this
to work.
- Added a key aliases to make emacs users happy:
cntl-V is same as +
cntl-N is down
cntl-P is up
- Added the Ultrix fflush bug fix.
- Added some extra attribution code (inspired by a patch from
serini@ghost.sm.dsi.unimi.it (Piero Serini)). The attribution string
may now contain:
%F who the message is from (this is the same as %s)
%D date the message was created (the from message)
%I message ID
%S message subject
- Added patch to allow embedded date in folder names (from Mike Kenney
<mike@wavelet.apl.washington.edu>). This allows:
%h month name ( 3 letter abbreviation )
%y last 2 digits of year
%m month number
%d day of the month
%j day of the year
This patch also allows environment vars to be enclosed in {}.
- Added Wayne Davison's patch to confirm on saving tagged messages.
- If DONT_ADD_FROM is defined the Your name option doesn't do anything so it
now doesn't show up (from Larry Schwimmer <rosebud@cyclone.Stanford.EDU>).
- There were a couple places in mailmsg2 that strincmp was called with only 2
parameters.
- The wildcard '*' is now allowed in the alternatives list. For example,
wfp5p@*virginia.edu matches user wfp5p for all machines at virginia.edu.
(from Michael Elkins <elkins@aero.org>).
- Added the mailing list stuff done by Paul Close <pdc@sgi.com>
- Allow the entry of aliases like the To: line
(from Keith Neufeld <neufeld@pvi.org>)
- Change to deal with parenthesized comments in when eliminating members from
an alias. (from Keith Neufeld <neufeld@pvi.org>)
- Allow a shell escape from the alias screen (just like from
the index screen). It does not put the shell escape onto the alias
screen menu. (from Keith Neufeld <neufeld@pvi.org>)
- Allow the use of "T" from the builtin pager. (from Keith Neufeld
<neufeld@pvi.org>)
- Changes over to enhanced mode in for the print command. From Chip Rosenthal
<chip@unicom.com>
- Readmsg fixed to not truncate messages when called from within elm (it'll
still have the the problem if run standalone). From Chip Rosenthal
<chip@unicom.com>.
- Added fix for utime() from Chip Rosenthal <chip@unicom.com>
- Flush the edit buffer after an edit and now make it a user config thing
(editflush in elmrc).
- Made a few Configure changes