1Unoffical Patch #17 (larry@garfield.cs.mun.ca Nov 19, 1991) 2 3 - core dump due to selecting last_article+1 on screen 4 5 - core dump due to consolidating (closing) last subject in a group 6 under certain circumstances. 7 8 - selecting consolidated articles with = command did not highlight 9 the selected consolidated articles correctly. 10 11 - fixed unconditional CLEAR TO END OF LINE that erased the next line on 12 some terminals. 13 14Unofficial Patch #18 (larry@garfield.cs.mun.ca; Dec 1, 1991) 15 16 - possible core dump/strange behavior when using '=' command to 17 return to the menu from article reading mode. 18 (Reported by Olaf Schlueter [olaf@oski.toppoint.de]) 19 20 - articles marked as A_LEAVE_NEXT ('=') are not given any special 21 attention in consolidated menu mode. I considered this incorrect 22 (I leave stuff around all the time :) and modified nn's behavior 23 to notice them. 24 25 - unable to de-select an auto-selected subject when using 26 consolidated-menu mode. 27 28Patch #17 and #18 made official (peter@DIALix.oz.au) 29 30Patch #19 (peter@DIALix.oz.au; 28 Nov 92) 31 32Title: Reduction in startup time 33From: Geoff Collyer <geoff@world.std.com> 34 The initial startup time can be cut in half. 35 36Title: A/UX 3.0 support 37From: Jim Jagielski <jim@jagubox.gsfc.nasa.gov> 38 Support for Apple A/UX 3.0, originally by Alexis Rosen 39 40Title: NeXT-2.2 support 41From: Lele Gaifax <lele@itnsg1.cineca.it> 42 43Title: Misc "static" function declarations corrected in prototypes. 44From: Lele Gaifax <lele@itnsg1.cineca.it> 45 46Title: NNTP name canolicalization in /usr/lib/nntp_server vs gethostname() 47From: Mark R. Levinson <chicken@minerva.npac.syr.edu> 48 The code now converts the name from nntp_server to the official name 49 before comparing it to the gethostname() result. This is useful if 50 nntp_server points to an alias (ie: news.your.domain.name) 51 52Title: Count lines if no "Lines:" field present. 53From: Paul Eggert <eggert@twinsun.com> and Luc Rooijakkers <lwj@cs.kun.nl> 54 If there is no "Lines:" field, NN will count them. 55 56Title: Correct line counting algorithm conflict with NNTP and folders/digests. 57From: Peter Wemm <peter@DIALix.oz.au> 58 The code would attempt to fetch a sub-article from a folder or 59 digest from the NNTP server - which is clearly not possible... 60 61Title: Make line counting optional 62From: Peter Wemm <peter@DIALix.oz.au> 63 #define DONT_COUNT_LINES option to config - it is undesirable to 64 fetch all articles from a remote NNTP server just to display a 65 line count on the menu. 66 67Title: print a "??" instead of "-1" for lines count in some display modes 68From: J. Porter Clark <jpc@avdms8.msfc.nasa.gov> 69 If you change the "layout", it is possible for NN to display a -1 70 in the lines count field. 71 72Title: Timezone misalculations 73From: Paul Eggert <eggert@twinsun.com> 74 NN mishandled non-GMT timezones in sorting resulting in articles being 75 presented out of order. 76 77Title: NNGRAB speedups 78From: Paul Eggert <eggert@twinsun.com> 79 nngrab generated unnecessary files and processes (especially on 80 BSD'ish machines that do not have a built-in "[" in the shell 81 82Title: NN wastes time seeking to find the file size 83From: Paul Eggert <eggert@twinsun.com> 84 Make NN count bytes as they are read rather than seeking all over the 85 place. 86 87Title: Correct NN's header parser to RFC compliant behavior 88From: Paul Eggert <eggert@twinsun.com> 89 NN gets confused on some variations of legitimate header continuation 90 lines. 91 92Title: Remove defunct comp.ai.neural-nets digest special handling 93From: Paul Eggert <eggert@twinsun.com> 94 The comp.ai.neural-nets digests are now in standard form. 95 96Title: Improve ":decode" to recognize more formats 97From: Mark Rawling <Mark.Rawling@mel.dit.csiro.au> 98 Fix for fault in the chtbl code 99 100Title: Add a "ignore-formfeed" variable 101From: Romain King <romain@pryamid.com> 102 Special request: display "^L" instead of a page break. Not on 103 be default. 104 105Title: Fix that darn "NO DIRECTORY (ok)" message that nnmaster is so fond of. 106From: Steve Simmons <scs@lokkur.dexter.mi.us> 107 If nnmaster is going to complain about a missing directory, at least 108 say which one it is.... 109 110Title: get aux script to clean up common editor backup files 111From: Minh Tran-Le <mtranle@paris.intellicorp.com> 112 Emacs makes file~ backups, some make file.bak, etc. Try and clean 113 up a few of them after launching an editor. 114 115Title: make aux script generate control messages that INN will accept. 116From: Peter Wemm <peter@DIALix.oz.au> 117 INN's inews is pretty fussy in what it will accept.. 118 119Title: Resolve conflict NN with INN on SVR4/386 120From: Peter Wemm <peter@DIALix.oz.au> 121 SVR4/386 has an "interesting" feature: You can write to a file 122 with mmap() **without** updating the timestamps.... 123 INN uses this for the active file, nnmaster now has a "-U" switch 124 to force an unconditional scan of the active file after the timeout, 125 rather than if the timestamps changed. 126 127Title: VMS ANU-NEWS compatability 128From: Matt Heffron <heffron@falstaff.css.beckman.com> 129 VMS ANU-NEWS's NNTP returns a different error code on no-such-article 130 131Title: Log group that generated a nntp error 132From: Matt Heffron <heffron@falstaff.css.beckman.com> 133 If a NNTP error is generated, log the group that caused it. 134 135Title: Support terminals without clr_eol capability 136From: Luc Rooijakkers <lwj@cs.kun.nl> 137 Some ancient terminals cannot clear to end-of-line. Simulate it by 138 outputting spaces. This means that printf/puts/putchar/stdout can 139 no longer be used in visual mode, because term.c has to keep track 140 of the cursor position. Use the new tprintf/tputc instead (but only 141 from files that are always linked with term.o). A list of the files 142 that may do printf/puts/putchar/stdout is in `noterm'. 143 144Title: Miscellaneous cleanup of term.c 145From: Luc Rooijakkers <lwj@cs.kun.nl> 146 Use tputs correctly unless BROKEN_TPUTS is #defined (UNISYS?) 147 148Title: Allow backslash to escape semicolons in extra headers 149From: Luc Rooijakkers <lwj@cs.kun.nl> 150 This is necessary to let the user add a MIME Content-Type header 151 via the news-header or mail-header variables. 152 153Title: Fixed annoying cursor position during some messages 154From: Luc Rooijakkers <lwj@cs.kun.nl> 155 Don't reposition the cursor to the prompt line after a message 156 if the message has overwritten the prompt. 157 158Title: Fixed invisible error messages 159From: Luc Rooijakkers <lwj@cs.kun.nl> 160 If terminal has a separate `visual mode' screen, printf/nn_exit 161 sequences don't work. Replaced user_error by nn_exitmsg which has 162 the exit status as additional first argument, and replaced the 163 printf/nn_exit sequences by calls to nn_exitmsg 164 165Title: Added charset variable and 8-bit checking for posts 166From: Luc Rooijakkers <lwj@cs.kun.nl> 167 The new charset variable can be set to the character set of the 168 terminal. It also determines the default value of the 169 data-bits variable. 170 If a message is posted whose body contains 8-bit characters 171 and charset is not set to "unknown", nn will add MIME-Version 172 and Content-XXX headers indicating its character set. 173 Such messages cannot be posted if the current character set 174 is a 7-bit set. 175 To obtain the previous behaviour, do `set charset unknown'. 176 177Title: Cleaned fill/nofill use in nn.1.* files 178From: Luc Rooijakkers 179 Lots of examples in the manual were being filled by *roff. 180 181Title: Sped up aux.sh and made it more portable 182From: Luc Rooijakkers 183 Using case instead of if is faster on shells that don't have if 184 185Title: Preview mode and menu.c 186From: Steven Grimm <koreth@hyperion.com> 187 The symptom: previewing (usually via auto-preview-mode) an article on 188 the next-to-last page of a group menu then going to the last page of 189 the menu caused nn to dump core if the previewed article had a greater 190 article number than the last article on the last screen. For example, 191 viewing article 'h' would cause the problem if the last screen only 192 went up to 'e'. 193 194Title: Correct a small over-optimization in term.c 195From: Luc Rooijakkers 196 My term.c contained one cursor position update too many, causing 197 things to go wrong sometimes. 198 199Title: DecStation/Ultrix 4.2 support 200From: "stolen" from Toivo Pedaste <toivo@uniwa.uwa.edu.au> my Peter Wemm. 201 conf/m-dec3100.h for mips machines running Ultrix. 202 203Title: Misc cleanups of bad code 204From: Paul Eggert <eggert@twinsun.com> 205 The code is loaded with functions that return no value, etc. 206 This is a start at fixing some of the more serious faults. 207 208Title: Macro to call metamail 209From: mikew@kpc.com 210 A simple example macro for passing MIME articles to metamail 211 212Title: Bug in folder code procuces core dump. 213From: Stefan Schwarz <stefans@bauv106.bauv.unibw-muenchen.de> 214 215Title: rfmail support 216From: Bo Kullmar <bk@kullmar.kullmar.se> 217 218Title: Clean up code part 2 219From: Peter Wemm <peter@DIALix.oz.au> 220 Continued on from what Paul Eggert started... AARRGGHH!! It's a mess! 221 222Title: Update help screens 223From: Peter Wemm <peter@DIALix.oz.au> 224 As suggested by Bill Wohler <wohler@sap-ag.de>, sort out the help 225 pages to "clarify" the differences between a few commands, and to 226 add the "(" and ")" keys.. [darn.. It's a tight fit in there.] 227 228Title: Domain/OS fixes and others 229From: Michael T Pins <mtpins@icaen.uiowa.edu> 230 1> Makes nnmaster run much happier on HP/Apollo's running Domain/OS 231 2> Allows nn to compile on variants of HP-UX, among others, that it 232 won't compile on otherwise. 233 234Title: aux.sh fixes 235From: Peter Wemm <peter@DIALix.oz.au>, bug reports from many... 236 The changes to aux.sh to use "here documents" broke a lot of 237 /bin/sh's Hack them out of the code, as not everybody has ksh. 238 239Title: SunOS 4.1 supports window siziing 240From: Paul Eggert <eggert@twinsun.com> 241 Add a change to s-sunos4-0.h 242 243Title: Add support for PCS Cadmus with MIPS cpu 244From: Peter Pichol <pp@informatik.uni-kiel.dbp.de>) 245 Create m-pcs.h 246 247Title: Add support for DEC Alpha's running OSF/1 248From: Mike Wexler <mikew@kpc.com> 249 Create m-alpha.h and s-osf1.h 250 251Title: A way to read even less news! 252From: Keld.Simonsen@dkuug.dk 253 It has been an old desire (viz. the FAQ) to be able to skip 254 followups to subjects, one did not care to read. The following two 255 variables implement a way to omit these followups, plus a way to 256 auto-select the articles that you are most likely to be interested 257 in, namely followups to articles that you have already read, or 258 followups to articles you have posted yourself. 259 260 The killing of uninteresting followups is done without entries in 261 the kill file. The thread is killed if the root article of the 262 thread is a followup itself, and the subject is not selected. 263 264 The default of this patch to nn is no changed behaviour. 265 266 Two new variables are introduced, can be set in the .nn/init file: 267 268 ignore-re (boolean, default false) 269 If set, articles with subjects already seen in a previous 270 invocation of nn or another newsreader - and not auto-selected - 271 are automatically killed. 272 273 auto-select-rw (boolean, default false) 274 If set, a subject of an article read or posted is automatically 275 used for subsequent auto-selecting (if not already selected). 276 277Title: Fixes for non ANSI SunOS cc 278From: Paul Eggert <eggert@twinsun.com> 279 280 NN 6.4.19gamma's term.c contains a cast that assumes that the 281 compiler is an ANSI C compiler. This doesn't work in K&R 282 compilers (e.g. SunOS 4.1.2 with bundled cc). 283 284Title: void vs int mixups in return values 285From: Paul Eggert <eggert@twinsun.com> 286 Here's a patch to NN 6.4.19gamma's term.c that fixes a problem 287 where it returns a value in a void function. 288 289Title: After you `:set record xxx', `:set' misdisplays mail-record 290From: Paul Eggert <eggert@twinsun.com> 291 The :set listing were incorrectly displaying mail-record and 292 news-record as always unset. 293 294Title: Improvement/bug fix to nn: archiving full articles using NNTP 295From: Fred Hucht <fred@hal6000.thp.Uni-Duisburg.DE> 296 If you set nnmaster to archive articles while in NNTP mode, it 297 only archived the headers... 298 299Title: Add a flag to show Originator 300From: Jim Jagielski <jim@jagubox.gsfc.nasa.gov> 301 Add's support for 'J' in news-header to display the "Originator:" 302 303Title: Catch short term error conditions to prevent nnmaster aborting. 304From: J.P.M. van der Jagt <jeanpaul@duteca.et.tudelft.nl> 305 Prevent things like "400 Server load too high, try again later" 306 from making nnmaster exit. 307 308Title: make nn 6.4 PL18 compile (and work ;-)) for hp_ux 8.0. 309From: Bart Muyzer <bartm@cv.ruu.nl> 310 311Title: `nn -xXm' does't clear to end of line (NN 6.4.19gamma) 312From: Paul Eggert <eggert@shadow.twinsun.com> 313 NN wasn't properly clearing to end-of-line after each output 314 line. 315 316From: Paul Eggert <eggert@twinsun.com> 317Title: infelicities in NN 6.4.19gamma term.c 318 The term.c code had some really bizzare conditional tests that 319 were uncessesary, and some were obvious typos. This patch doesn't 320 fix any real bugs, but it removes the needless code. It also 321 introduces a few optimizations in critical places: (This is a 322 particularly big win on most Sparcs, which don't have integer 323 division.) 324 325From: Peter Wemm <peter@zeus.DIALix.oz.au> 326Title: silly bug in nn-use-elm script 327 contrib/nn-use-elm had a leading #/bin/sh instead of #!/bin/sh 328 This causes invocation of csh on many systems... (including mine :-) 329 330From: Peter Wemm <peter@zeus.DIALix.oz.au> 331Title: long standing bug in sort.c 332 non-related articles were often being associated in 333 subject-match-parts mode. This tries to make it less likely. 334 335From: Steve Grimm <koreth@hyperion.com> 336Title: prefix.c fix, HAVE_TRUNCATE for SunOS. 337 prefix.c change #ifdef NNTP to #ifdef INEWS_PATH 338 SunOS has truncate. 339 340From: Michael Masterson <mjm@thumper.progcons.com> 341Title: account.c is broken 342 The code uses a structure before defined if accounting is enabled. 343 Update s-sys5-4.h. m-i80386.h needs work. 344 345From: Peter Wemm <peter@DIALix.oz.au> 346Title: improve m-i80386.h 347 m-i80386.h didn't deal with network byte order very well. As far as 348 I can tell, if you want network byte order, then you probably want it 349 for NFS sharing. In which case you have TCP/IP. Use it's routines. 350 351From: Steven Grimm <koreth@hyperion.com> 352Title: Patch for preview mode display glitch 353 the 'xn' termcap flag is conflicting with clrline() and long 354 subject lines, causing the following line to be cleared. 355 356From: Peter Wemm <peter@DIALix.oz.au> 357Title: fix aux.sh 358 aux.sh was failing when using the nn_elm script from the FAQ. 359 In the process of reformatting it, I found the bug, a missing 360 semicolon. It would fail whenever mailer-pipe-input was unset. 361 362From: Peter Wemm <peter@DIALix.oz.au> 363Title: The Biggie! Merge in NOV code! 364 Well, this is the big one. #ifdef NOV in config.h, and it even works 365 over NNTP. The some of the hacks to get NN to use the libnov code is 366 pretty baroque, but necessary in order to maintain compatability. 367 NN's internals are *very* dependant on the nnmaster database and it 368 does not particularly appreciate the change. If you enable NOV, all 369 bets are off. :-) Most of the code is from Geoff Collyer 370 <geoff@world.std.com> although the NNTP mods and other tweaks are 371 from others whom I cannot identify.. 372 This is very much a Stop-Gap solution. 373 374From: Michael Masterson <mjm@thumper.progcons.com> 375Title: auto-select-rw and maybe ignore-re kill folder mode 376 NN takes a SEGV while reading a folder with these enabled. 377 [Peter: I'm unable to reproduce the bug....] 378 379From: Steven Grimm <koreth@hyperion.com> 380Title: Patch for ignore-re and -x 381 Here's a tiny patch so that ignore-re is turned off when you're 382 using the -x flag, or when you go into a group to read old 383 articles. I think this is better than the way it currently works; 384 currently, if you have ignore-re set on a group, there's no way 385 (that I can discover) to get nn to show you an article whose 386 thread's initial article has expired. 387 388From: Peter Wemm <peter@DIALix.oz.au> 389Title: folder move vs. auto-select-rw 390 Basically, patch up Michael Masterson's problem. Although I 391 couldn't core dump (lucky me..), I did notice that if I was 392 reading a folder, the articles I read were added to the kill file. 393 I've added a temporary override to auto-select-rw that makes it 394 ignored in a folder. This should fix Michael's problem. 395 396From: Peter Wemm <peter@DIALix.oz.au> 397Title: 6.4.19 is no more. 398 So as to distinguish out butchering from Kim's work, we've decided 399 on a revision "bump". It's NN-6.5 now. 400 401From: Peter Wemm <peter@DIALix.oz.au> 402Title: gcc -Wall cleanup 403 It now compiles a lot more cleanly under -Wall, but not 100% 404 clean.. :-) I found some sneaky bugs in the menu code and the 405 init file parser in the process. 406 407From: Peter Wemm <peter@DIALix.oz.au> 408Title: Merge in Article search from Andy Burgess <aab@cichlid.com> 409 There are numerous collisions with the NOV code, and it does not 410 work with digests, etc, etc. It is *not* enabled in config.h by 411 default, as It has not revieved enough testing by the group. 412 It's also not documented in the manual... (Volunteers?) 413 Also, don't enable this if you read via nntp... Your server will 414 hate you... 415 416-- 6.5.0 beta goes out Aug 31 1993 -- 417 418From: Paul Eggert <eggert@twinsun.com> 419Title: The code still says 6.4 in various files, including the documentation. 420 421From: Paul Eggert <eggert@twinsun.com> 422Title: Bogus prototypes for syslog() functions are wrong. 423 424From: Paul Eggert <eggert@twinsun.com> 425Title: NO_X_NEWSREADER should be defined by default. 426 [Peter: Awww! If Tin can do it, why can't I? :-) ] 427 428From: Paul Eggert <eggert@twinsun.com> 429Title: Redeclaration of catch_winch on machines with RESIZING 430 431From: Paul Eggert <eggert@twinsun.com> 432Subject: nn-6.5.0 mishandles time zone changes near short month boundaries 433 NN 6.5.0 mishandles non-GMT time zones near short month 434 boundaries, which causes articles to be incorrectly sorted by 435 date. For example, pack_date("30 Sep 1993 23:00 -0700") and 436 pack_date("1 Oct 1993 06:00 GMT") should yield the same value, but 437 in NN 6.5.0 the former yields 3614760 while the latter yields 438 3616200. This is becoming more and more of a problem as more 439 people use non-GMT time zones in their news. Here is a patch. 440 441From: Toivo Pedaste <toivo@uniwa.uwa.edu.au> 442Title: Missing 'static' on some functions 443 db_init_group(), catch_winch(), grep_article() 444 445From: Steve Davies <srd@peora.sdc.ccur.com> 446Title: signal prototypes misdeclared 447 There were a few places where signal handlers were declared void, 448 but in fact were sig_type. This was OK on machines that had the 449 same definition, but not on those with sig_type == int. 450 Also, truncate() is messed up.. 451 452From: Steve Grimm <koreth@hyperion.com> 453Title: tidy up config.h 454 There were a few typos and errors in config.h-dist 455 456From: Peter Wemm <peter@DIALix.oz.au> 457Title: fix up prototype problems with truncate and gethostname 458 It seems every system has a different idea on what they are supposed 459 to be. I give up... NN has not got internal functions called 460 nn_truncate and nn_gethostname, which are prototyped, and will call 461 "the real thing" if it's available. It seems the safest way. 462 463From: Steve Davies <srd@peora.sdc.ccur.com> 464Title: 6.5 NOV + NNTP cleanups 465 NN + NNTP + NOV : db.c, global.h, nn.c, nntp.c 466 - This patch fixes problems when NN is run on the same machine as 467 the NNTP server. 468 - When compiled with NNTP and NOV, a temp file was being created 469 when the active, active.times, and .overview files were read from 470 the NNTP server. This patch eliminates the creation of that 471 temporary file. In db.c (readactfile, readtimfile, 472 db_read_group), instead of reading from the temp file, the data is 473 read directly from the NNTP server. The idea for this comes from 474 Peter Orbaek (poe@daimi.aau.dk). This requires a simple patch to 475 the nov sources (which probably should be applied by hand rather 476 than running this file through patch): 477 [NOTE: NN-6.5 contains a 'nov-lite' distribution with this patch 478 applied and everything non-essential for NN removed. -Peter ] 479 480--- nov/libnov/oldlib.c Tue Sep 7 13:47:31 1993 481+++ nov/libnov/lib.c Tue Sep 7 13:47:42 1993 482@@ -213,6 +213,7 @@ 483 } 484 485 while ((line = fgetms(gp->g_stream)) != NULL) { 486+ if (line[0] == '.') break; 487 art = (struct novart *)malloc(sizeof *art); 488 if (art == NULL || 489 (prssts = prsovline(line, gp, art, prevart)) == PRSFAIL) { 490 491 492 - The code in db_read_art for handling crossposts when NOV is 493 defined does not handle the (hypothetical?) case where there are 494 extra headers besides Xref: in the .overview file. It also 495 doesn't work when cnews is used to maintain the .overview files, 496 since cnews appears to use xref: instead of Xref: in the .overview 497 file. A new routine "db_fixup_cross_postings" is added to db.c . 498 This routine is derived from code by Charles Henrich and 499 rmtodd@servalan.servalan.com. 500 501From: Peter Wemm <peter@DIALix.oz.au> 502Title: include a minimal nov distribution. 503 Because we need a trivial patch to the libnov code which Geoff 504 Collyer doesn't want to put in his code (and quite understandably 505 because it's a "hack"), and I dont want 10,000 email messages 506 saying "NOV + NNTP doesn't work" when they build it without 507 patching because they never read intructions that tell them to 508 patch it. :-) If anybody does this to me now because they've 509 removed the patched version, I'll reply with a uuencoded 16 Meg 510 kernel crash dump. (ie: a bunny dropping..) :-) 511 512From: Peter Wemm <peter@DIALix.oz.au> 513Title: Fix "Out of memory: cannot allocate -4 bytes" 514 If NN could not determine the window size from the termcap or the 515 TIOCGWINSZ ioctl's, it would try and malloc a negative buffer.. 516 517From: Peter Wemm <peter@DIALix.oz.au> 518Title: typo in nn_gethostname() 519 hostname.c would not compile if HAVE_HOSTNAME defined. 520 521From: Jim Jagielski <jim@jagubox.gsfc.nasa.gov> 522Title: nn -a0 botches display on some hosts 523 There is some interaction between curses, tprintf, and terminfo. 524 525From: Jim Jagielski <jim@jagubox.gsfc.nasa.gov> 526Title: Add new conf/s-aux3-0.h 527 528From: Robert.Stockton@ELROND.GANDALF.CS.CMU.EDU 529Title: nov code overrunning '.' on XOVER data when active is incorrect. 530 If the active file says there is articles, but the .overview is 531 empty, the novall() and novnext() miss the "eof" dot and hang 532 forever. 533 534From: Toivo Pedaste <toivo@uniwa.uwa.edu.au> 535Title: Explicit support for mouse under xterm 536 This code can detect if running under xterm, and will enable mouse 537 events if so. The mouse clicks are used for menu selection, 538 reading, previewing and so on. It makes NN "almost" hands-free 539 for most news reading tasks. 540 541From: Peter Wemm <peter@DIALix.oz.au> 542Title: Minor tweaks for mouse-code 543 Because of a "feature" apparently in the SYSV termio driver and 544 possibly the POSIX termios, NN was disabling raw mode during 545 screen updates in order to enable xon-xoff flow control to work. 546 This was causing character echoing of mouse events during screen 547 redraws and looked rather ugly. This problem never occurs on 548 older BSD systems with the sgtty style driver. This tweak causes 549 NN to not switch off raw mode on Xterms when mouse is enabled.. 550 Heck, if xterm needs flow control, something is very wrong... 551 552From: Tom Limoncelli <tal@warren.mentorg.com> 553Title: Possible memory leaks 554 Purify is not at all happy with some of the things that the NOV 555 code does. I dont think there is much of a problem, although I 556 agree that it's messy. Basically, NN is allocating buffers for 557 the hash tables and not freeing them at exit. As these are reused 558 while NN travels from group to group, this is mostly static. 559 NN does not free() these on exit and Purify complains. The 560 invitation to fix this is open to all.. :-) 561 562From: Peter Wemm <peter@DIALix.oz.au> 563Title: major reorganisation of NOV code. 564 Shift files from nov/* to the top level directory for radical 565 simplification of compilation. Rationalise include files. 566 Substitute some simple routines to reduce extra files needed from 567 the NOV distibution. 568 569From: Peter Wemm <peter@DIALix.oz.au> 570Title: major repairs of above reorganisation. 571 Well, the "simple" reorganisation turned out to be rather dramatic 572 in it's affects. This cost weeks of debugging and nearly caused 573 me to give up and reverse the last upmteen patches (I had 574 neglected to do a CVS checkin before starting). It turned out to 575 me a malloc problem. I *suspect* there is still a memory leak 576 there, but it doesn't seem to show up on the process size after a 577 lengthy reading session.. 578 579From: Ian Nicholls <iann@storesys.coles.oz.au> 580Title: NCR tower 32 tweaks. 581 nn-6.5.0.b2 had troubles with include files. This helps a bit. 582 583From: Marc Boucher <marc@cam.org> 584Title: additions to conf/s-sys5-4.h 585 This adds some omissions to the config file. 586 587From: Marc Boucher <marc@cam.org> 588Title: bzero/memset are out, clearobj is in... 589 Following Kim's coding conventions, change a couple of memset()'s 590 that I introduced to clearobj's. This looks a lot nicer now.. 591 592From: Michael Pins <mtpins@isca.uiowa.edu> 593Title: b4 patch 594 An nntp speedup, along with changing a number of chars back to ints so 595 nn will actually work properly on a number of systems. 596 597nn-6.5.0.b5: 598From: mcm@gorgon.unx.sas.com (Mike Mitchell) 599Subject: Re: Problems with splitting digests 600 Digests are now split under NOV. 601 602From: romain@sword.eng.pyramid.com (Romain Kang) 603Subject: NN hacks for NOV 604 Major NOV hacks. Things should be rather faster now. 605 Also, the addition of nnview for reading folders. 606 607From: Michael Pins <mtpins@isca.uiowa.edu> 608 General bug-fixing and patch cleanup. 609 610nn-6.5.0.b6: 611From lar3ry@zk3.dec.com 612Subject: OSF/1 (err, Digital Unix) patches 613 Fixes for the Alpha and OSF/1. 614 615From: David Dyer-Bennet <ddb@gw.ddb.com> 616Subject: s-linux.h 617 config file for linux 618 619From: jpc@hammer.msfc.nasa.gov (J. Porter Clark) 620Subject: libnov.c bug 621 There's a bug in the code that causes .newsrc to be updated even though 622 the corresponding articles aren't read. Swapped the order of the test. 623 624From: Jacob Martin Bohn Lorensen <jablo@diku.dk> 625Subject: timeout problem with NOV and nntp 626 When the server unexpectedly breaks the connection nn does not always 627 discover this. This can be fixed by just re-opening the connection. 628 629From: Michael Pins <mtpins@isca.uiowa.edu> 630 General bug-fixing, patch cleanup and code cleanup. 631 632nn-6.5.0.b7: 633From: <root@oit.gatech.edu> 634 config files for AIX/RS6000, and a patch for AIX brain damage. 635 636From: Michael Pins <mtpins@isca.uiowa.edu> 637 Put accounting logs under LIB instead of DB, so DB isn't needed 638 unless nnmaster is used. Pulled accounting #def's out of account.c 639 into the new account.h so it's no longer necessary to edit a .c file. 640 641 Fixed line counting in digests (used to claim one too many). 642 643 Included the inst.sh file for the Apollo I've been using for years. 644 Anyone running Domain/OS will want to use this instead of inst.sh. 645 646 Yet more general clean-up. 647 648nn-6.5.0: 649 Added account.h to account dependencies. 650 651 Added freebsd config file. 652 653 Hacked together some docs for nnview. 654 655 Yet even more general clean-up. 656 657nn-6.5.1: 658 Added netbsd config file. 659 660 Fixed up s-template.h. 661 662 SunOS doesn't have memmove (and likely neither do most older BSD 663 derived systems). Early versions of Solaris don't have bcopy.... 664 Added NO_MEMMOVE and used it to ifdef around these. Most (all?) 665 can probably be memcpy, but until I have time to insure that the 666 args don't overlap, I'll be paranoid. 667 668nn-6.5.2: 669 Changed group_number to int32 so > 32767 groups works 670 671 Fixed problem with long References: lines 672 673 Fixed problem with large headers (seen mainly in nnview) 674 675 Hacked in a workaround for sites that don't support 676 LIST active.times 677 nn will now see new groups properly 678 679 More general cleanup of the code and minor bug fixes 680 681From: mtranle@theoden.intellicorp.com (Minh Tran-Le) 682 nn was eating the first number from a LIST response 683 workaround for M$ newsserver braindamage 684 685From: Nick.Holloway@alfie.demon.co.uk (Nick Holloway) 686 fix for INN return message re: mail to moderator 687 688nn-6.5.3: 689 Ripped out "LIST active.times" workaround. Typhoon will support it 690 soon (likely before anyone reads this), and the workaround had bugs. 691 692 Changed accounting so that time used outputs HH:MM instead of HH.MM 693 (does anyone still use accounting?) 694 695 Added a compile time option to turn on more headers to make the 696 GNKSA people happy. The default behavior hasn't changed. 697 698From: Bruce Fisher <bruce@smtl.co.uk> 699 The news spool directory was hardcoded in libnov.c, conflicting 700 with NEWS_DIRECTORY from config.h 701 702From: Dale Talcott <aeh@quest.cc.purdue.edu> 703 A new variable "cross_post_limit" to kill articles posted to more 704 than cross_post_limit newsgroups. Default=0 (no limit). 705 706From: Dale Talcott <aeh@quest.cc.purdue.edu> 707 bug in pack_date.c--computes leap year incorrectly 708 709From: Tom Kartes <tom@zeus.Materna.DE> 710 Killing the current article sometimes kicks you out of the newsgroup. 711 712From: Tom Kartes <tom@zeus.Materna.DE> 713 Saving an article into a file will escape some headerlines if 714 you're using NOV and did not previously view this article. 715 716From: 5ar4ro$ri7@bogon.com (John Henders) 717 a patch that adds S)ign [pgp] as an option to the posting menu 718 719From: "Maarten Carels" <maarten@ic.uva.nl> 720 new config.h option to have a different filename for the 721 overview files 722 723nn-6.5.4: 724 Changed defaults in config.h to more sane ones. 725 Fixed a bug in libnov.c so NOV_FILENAME actually works. 726 Fixed a nasty bug that would occasionally cause nn to go ballistic 727 when ripping a digest if the overview data wasn't quite right. 728 nn now compiles cleanly on FreeBSD. 729 730nn-6.5.5: 731 Changed answer.c to use 4-digit year (making it RFC-822 732 non-compliant, but son-of-rfc-1036 compliant). 733 Back-ported pack-date.c from nn7. It fixes a number of minor bugs, 734 and deals with Y2K brain-damage from other software. 735 736nn-6.5.6: 737 Fixed two new date sorting bugs. Oops. 738 Workaround for new brain-damage in Solaris7 include files. 739 740nn-6.6: 741 Fixed large uid/gid problems 742 Fixed Cc (it actually works now) 743 changed X-Newsreader to User-Agent 744 changed -ltermcap to -lncurses in s-linux.h, as they broke it again 745 the date parser now translates 101 -> 2001 (yes, I'm still seeing this) 746 backported some changes to decode from nn7 747 748From: Dave Olson <olson@bengaltech.com> 749 added NNTP_AUTH support 750 added ssl support 751 nn now posts internally via NNTP (NNTP only) 752 all functions of aux now internal (NNTP only) 753 added support for multiple nntp-server/newsrc/nn-directory 754 755From: Charles Lindsey <chl@clw.cs.man.ac.uk> 756 aliased groups didn't work correctly with NOV/no-NNTP 757 758From: jpc@suespammers.org (J. Porter Clark) 759 added E = save-header-only function 760 761nn-6.6.1: 762 online help files will now get installed properly. 763 now compiles if nntp isn't defined. 764 dumped core on posting if news-record had a bad path. Fixed. 765 no longer segfaults if user over quota. 766 no longer segfaults if NNTPSERVER is invalid. 767 will now search path for executables. 768 passwords longer than 8 chars now work from Solaris. 769 nn will silently ignore any non-digests with [D,d]igest in the 770 subject, and apparently has for quite some time. 771 set the default for also_full_digest to true as workaround. 772 nn has long gotten all the articles in a group when any number of 773 extra articles was requested. Fixed. 774 will now reconnect if it times out during a post. 775 776nn-6.6.2: 777 nn wouldn't compile if NNTP was defined and NOV wasn't. Fixed. 778 779nn-6.6.3: 780 cpp is broken on gcc-3.x. Worked around. 781 ripped out SSL support, as it was apparently fatally broken. 782 removed nnmail and HAVE_ROUTING. No longer needed. 783 converted from varargs to stdargs. 784 added MacOS X support 785 nn will now attempt to discover the domain if not defined. 786 will now reconnect, even to broken servers that don't send 503 787 788From: Bob Hain <bob@me.umn.edu> 789 use full_name instead of pw_gecos in gen_frompath 790 791nn-6.6.4: 792 added iso-8859-15 to chset 793 increased END_OF_LIST (some servers have very large article numbers) 794 changed ospeed to nnspeed to avoid namespace conflicts 795 always rebuild binary killfile when starting nn(NOV) 796 if the ORGANIZATION environment variable is set, nn will use it 797 the editor again accepts arguments, and starts after the headers 798 added new variable use-editor-line, for the above 799 800From: Jacques A. Vidrine <nectar@FreeBSD.org> 801 nn has a remote security hole. Fixed. 802 803From: Felicia Neff (fn@panix.com) 804 the variables for nn_exitmsg when nn is already running were swapped 805 806nn-6.6.5: 807 added -lcurses to macosx TERMLIB 808 nnview doesn't need to talk to the NNTPSERVER 809 removed last vestige of sys_errlist 810 811From: J.K.Wight@newcastle.ac.uk 812 cp is undefined if DOMAIN is defined and HIDDENNET isn't. Fixed. 813 814From: Neil W Rickert <rickert@cs.niu.edu> 815 nn segfaults if it reconnects during initial authentication. Fixed. 816 817From: Jeffery Small <jeff@cjsa.com> 818 nn sends a bogus [post|follow] failed message when using cnews and 819 !nntp. Fixed. (re-added two options to config.h) 820 821 822nn-6.7alpha: 823 Fixed some portability problems with old OSs. 824 First pass at starting to clean up the code. Ick, what a mess. 825 nnmaster and nnacct appear to have been broken during nn-6.6, fixed. 826 The name length displayed is now based upon the width of the screen. 827 nn will now correctly parse From: lines that contain () 828 For example: 829 From: "Jesse James (home)" <jjames@example.com> 830 used to display as "home", the new behavior will display 831 it as "Jesse James". However, this may affect killfiles 832 that depend upon the old behavior. 833 To use the old behavior, set the new run-time variable 834 'old-packname' to true. 835 The hold command from the post/mail menu failed if the users 836 nn_directory wasn't a local filesystem. Fixed. 837 nnpost didn't check to see if there was a held message. Fixed. 838 Posting from nnview failed due to the nntp connection not being 839 setup. Fixed. 840 When replying with a held message, nn would use the post menu. Fixed. 841 Header check macro is now case insensitive for both news and folders. 842 Added GPG support, which comes with a new run-time variable 843 'sign-type', and a compile-time variable SIGN_TYPE to set 844 the default to either pgp or gpg. 845 nn will now print the version before connecting to the newsserver 846 internalized :bug information for more accurate reporting 847 nn will now look in /etc/resolv.conf for a domain if DOMAIN isn't 848 defined and hostname isn't a FQDN. 849 new command ":show config" will show the compile-time definitions. 850 851From: Henning Schmiedehausen <hps@intermeta.de> 852 If query-signature wasn't set, nn ignored signature-append-post. 853 Fixed. 854 855From: Tim <tim@sleepy.wojomedia.com> 856 nntp-user and nntp-password should be settable in the init file. 857 Fixed. 858 859From: Steve Rumble <rumble@ephemeral.org> 860 nn segfaults if a client is unable to resolve it's own name. 861 Fixed. 862 863From: flaps@dgp.toronto.edu (Alan J Rosenthal) 864 nn should trim long References: lines. Added patch to do so. 865 866nn-6.7beta: 867 Cleaned up some old (unused) variables. 868 More general code cleanup. 869 Updated the manpages to include new variables. 870 Updated INSTALLATION and other docs so they're (roughly) current. 871 If new-group-action is set to 0 in init, nn will no longer get 872 active.times. (Yes, this is what the docs have claimed was 873 always true, it actually works now.) This can cut startup 874 time in half on slower links. 875 876nn-6.7.0: 877 And yet even more general code cleanup. 878 rearranged startup code so macros and keymaps work in the init file. 879 only query-signature if .signature exists 880 fixed some portability problems with NeXTstep-3.3 and NetBSD 881 882nn-6.7.1: 883 declared main to be int in usercheck.c and cvt-help.c 884 changed a few refs to nn-6.5 and nn-6.6 to nn-6.7 885 clear display before printing "No News" so as to not overwrite 886 887nn-6.7.2: 888 changed TERMIO to TERMIOS in s-linux.h 889 execute needs to exit, not return, if exec fails 890 callable programs (i.e pager) should accept arguments 891 892nn-6.7.3: 893 Fixed the Makefile so recent versions of gcc no longer choke. 894 The -N option appears to have been broken since 6.6.0. Fixed. 895 896