mhchanges

早期freebsd实现

Tue May 21 22:33:58 1985  /mtr <mrose@nrtc-isc>

	uip/scansbr.c: when copying characters to the scan listing,
	    know about underlining conventions


Wed May 22 10:16:17 1985  /mtr <mrose@nrtc-isc>

	sbr/m_getfld.c: Eom() was botching gathering the UNIX from line
	    (fix from Matt Crawford <oddjob!matt@lbl-csam>)
	sendmts/smail.c: sm_end(NOTOK) was clobbering sm_reply


Wed May 22 15:29:25 1985  /mtr <mrose@nrtc-isc>

	popsupport/popser.c: if maildrop doesn't exist, note it and
	    continue
	conf/doc/bbc.rf, uip/bbc.c: add -user switch to help POP debugging
	bboards/getbbent.c: have getbberr() return useful info for
	    setbbfile (), setbbinfo (), and setbbent ()
	popsupport/{popaka,popser,popwrd}.c, uip/{bbc,spop}.c: take advantage
	    of getbberr()


Wed May 22 19:49:43 1985  /mtr <mrose@nrtc-isc>

	Makefile, config/mh.h, conf/makefiles/*, bboards/getbbent.c,
	    */llib-l*, tws/{dtime,dtimep}.c: start the linting process,
	    lint everything except the uip/ directory
	sendmts/smail.c, uip/trmsbr.c: move discard to sbr/ library
	mf/mf.c: lint fixes
	sbr/{getans,m_{gmsg,setcur},refile,showfile}.c: lint fixes
	sbr/{putenv,sprintb}.c: #include mh.h
	sbr/*: VOID -> (void)
	bbsupport/bb{aka,exp,tar}.c: lint fixes


Thu May 23 09:06:32 1985  /mtr <mrose@nrtc-isc>

	conf/makefiles/{mts,uip}, uip/msgchk.c: moved from mts/

Thu May 23 22:06:25 1985  /mtr <mrose@nrtc-isc>

	conf/makfiles/uip,uip/*.c: next step in the lint process
	config/mh.h, sbr/llib-lmh, uip/rmf.c: move remdir() to sbr/makedir.c


Mon May 27 11:33:32 1985  /mtr <mrose@nrtc-isc>

	config/mh.h, sbr/{closefds,m_getdefs}.c: #ifdef OVERHEAD experimental
	    code


Mon May 27 17:05:18 1985  /mtr <mrose@nrtc-isc>

	tws/dp.c: move to uip/
	uip/formatsbr.c: new module
	config/scansbr.h: define stuff for formatsbr


Mon May 27 21:45:23 1985  /mtr <mrose@nrtc-isc>

	uip/{ap,addrsbr}.c: upgrade ap to use formatsbr; required a new
	    param to getm(), so modules which call getm() got changed
	    as well
	othersupport/rcvtrip/rcvtrip: upgrade


Tue May 28 19:45:36 1985  /mtr <mrose@nrtc-isc>

	config/scansbr.h,uip/{inc,msh,rcvtty,scan,scansbr}.c: bite the
	    bullet and have scansbr use a format string
	uip/{inc,scan}.c: no more -[no]numdate, -[no]size, -[no]time
	    and use -format string and -width columns instead


Wed May 29 14:53:33 1985  /mtr <mrose@nrtc-isc>

	othersupport/mtrenv/*: update


Wed May 29 16:38:49 1985  /mtr <mrose@nrtc-isc>

	Makefile, conf/makefiles/*: support depend.  N.B.: This is for
	    DEVELOPMENT WORK ONLY. "make depend" runs unifdef because some
	    #include files are configuration specific.  Sadly, unifdef
	    isn't on all systems (it originated at Rand and migrated to
	    4.2BSD).  As a result, unless you have unifdef, don't run make
	    depend.  The dependencies listed in the conf/makefiles/* files
	    are "generic".


Fri May 31 22:19:15 1985  /mtr <mrose@nrtc-isc>

	uip/replsbr.c: bite the bullet and have repl use the format string
	    routines as well


Fri May 31 23:33:59 1985  /mtr <mrose@nrtc-isc>

	conf/doc/*.rf: more doc fixes; document WHATNOW, OVERHEAD
	sbr/m_getdefs.c: more OVERHEAD code to optimize MH subforks


Sat Jun  1 01:35:06 1985  /mtr <mrose@nrtc-isc>

	conf/doc/mhook.rf, uip/rcv*.c: account for MMDF-II change regarding
	    rcvmail hooks and maildelivery(5)


Sat Jun  1 11:18:06 1985  /mtr <mrose@nrtc-isc>

	mf/mf.c, uip/*sbr.c: more register declarations


Sat Jun  1 14:45:25 1985  /mtr <mrose@nrtc-isc>

	uip/{pick,picksbr,msh}.c: bite the bullet and have pick allow
	    complex booleans
	conf/doc/{mh-chart,pick}.rf: ditto

Sun Jun  2 14:07:46 1985  /mtr <mrose@nrtc-isc>

	uip/whatnowsbr.c: don't say the draft's left around if it isn't
	    (suggested by Mike O'Brien)
	uip/whatnowsbr.c: return value from non-initial calls to m_edit()
	    not handled right (pointed out by Richard Johnson)
	sbr/m_getfld.c: be a bit smarter when recognizing that idiotic
	    "From " line from silly UNIX-style maildrops


Sun Jun  2 18:36:19 1985  /mtr <mrose@nrtc-isc>

	conf/doc/*.rf: more clean-up
	uip/ali.c: change help messages a bit, and add -[no]normalize
	    switches
	conf/doc/mh-tailor.rf: new file
	config/mh.h,uip/{post,repl}.c: move definition of OUTPUTLINELEN
	    to mh.h so sbr/help.c can use it
	sbr/help.c: smarter output routine for options


Mon Jun  3 11:28:46 1985  /mtr <mrose@nrtc-isc>

	TODO: remove since there's nothing worth left doing in it!


Mon Jun  3 14:54:30 1985  /mtr <mrose@nrtc-isc>

	othersupport/usenix: new directory containing the MH paper
	    presented at the 1985 Summer Usenix Conference


Mon Jun  3 22:15:18 1985  /mtr <mrose@nrtc-isc>

	conf/doc/packd.rf, uip/packd.c: de-implement, essentially useless
	    under 4.2BSD


Tue Jun  4 13:30:27 1985  /mtr <mrose@nrtc-isc>

	uip/trmsbr.c: optimize a bit, the ontty test in clear_screen ()
	    isn't needed


Tue Jun  4 20:00:32 1985  /mtr <mrose@nrtc-isc>

			  -- Suggested by Dave Yost --
	conf/doc/tmac.h: use ".nf" after ".SH" inside various macros
	    (since ".SH" tends to set ".fi")
	sbr/m_{gmsg,sync}.c: some .mh_sequences optimizations for hm
	conf/config/config.c: support #ifdef NOMHSEQ


Thu Jun  6 23:03:09 1985  /mtr <mrose@nrtc-isc>

	uip/{format,mhl}sbr.c: add some definitions for sprintb()
	bboards/bboards.h: define BBITS there instead of uip/bbc.c


Sun Jun  9 14:44:36 1985  /mtr <mrose@nrtc-isc>

	conf/makefiles/uip,uip/show.c: make show/next/prev one program
	uip/{show,next,prev}hdr.c: remove as a result of above
	conf/doc/{next,prev}.rf: update


Sun Jun  9 16:37:40 1985  /mtr <mrose@nrtc-isc>

	uip/dropsbr.c, uip/sbboards.c: .cnt mechanism broken,
	    introduce new routine mbx_openX() in dropsbr.c and use that
	    instead of mbx_open when opening an .cnt file



Sun Jun 16 10:21:33 1985  /mtr <mrose@nrtc-isc>

	uip/{ap,dp,{repl,scan}sbr}.c: the old echo controversy raises it's
	    ugly head: suppose we have a line in a replcomps file which is
	    entirely in a %<...%> conditional and the conditional is false.
	    In that case, we still get a blank line.  This is very BAD if it
	    occurs in the headers.  Two solutions possible:
		1. If a call to FSprintf() returns a empty string, don't
		   put out a \n.
		2. In new_fs, put the \n's in and just return one big
		   string instead of an array of strings.
	    Option [2] was implemented initially because it lets conditionals
	    span multiple lines in the file.  In order to keep the format files
	    from looking too complicated, EOL was ignored in format files,
	    \n's should be embedded where appropriate.  A single format
	    string argument is considered to have an implicit \n.  This is
	    inconsistent, but it keeps the format files readable.  Option [2]
	    was discarded though since it required a \n at the end of each
	    line in the format file if there was no conditional stuff
	    there.  So, option [1] got implemented even though it doesn't
	    allow conditionals to be multi-line (which is okay, since you
	    can use \n in a line).


Thu Jun 20 19:55:23 1985  /mtr <mrose@nrtc-isc>

     -- Performance Enhancements from the 4.3BSD folks (by Van Jacobson) --
	config/mh.h: define vfork for systems without it
	sbr/{refile,showfile}.c,uip/{replsbr,send,whatnowsbr}.c:  replace
	    fork() with vfork()
	config/mh.h: move Van's getfld.h definitions into here
	sbr/m_getfld.c: major performance tuning!  (use -DRPATHS to get
	    Return-Path: info from UNIX from: line)
	uip/{inc,msh}.c: m_unknown parameter added
	uip/trmsbr.c: support for TIOCGWINSZ in 4.3BSD(!!)
	uip/scan.c: -[no]reverse under #ifdef BERK (I really HATE this)
	    also, update context prior to scan, again under #ifdef BERK
	uip/scansbr.c: performance enhancements plus inline copy of message
	    body to output file for inc
	uip/mhlsbr.c: under #ifdef BERK pipe output to /usr/ucb/more if
	    stdoutput is a terminal.
	uip/post.c: #ifdef DUMB a bit dumber for Berkeley addrsbr (actually
	    these changes were made earlier for other reasons)
	uip/addrsbr.c: use prescan-like routine in SendMail under #ifdef
	    BERK, also enable wild-carding.  The former probably breaks
	    the stuff in ap, et. al.


Fri Jun 21 13:09:29 1985  /mtr <mrose@nrtc-isc>

	uip/addrsbr.c: liked the wild-carding, so even for #ifndef BERK you
	    get it.  Of course in this case, you can wildcard on both the
	    mbox and the host.  It turns out that this is really useful for
	    people who have many mailboxes and get a lot of munged mail.


Mon Jun 24 14:58:25 1985  /mtr <mrose@nrtc-isc>

	conf/doc/*.rf: Numerous documentation fixes suggested by Jerry
	    Sweet.
	conf/doc/ADMIN.rf: New document: Administrator's Guide


Mon Jun 24 23:31:42 1985  /mtr <mrose@nrtc-isc>

	uip/send.c: try to prevent confused annotations.  The algorithm
	    works for dist and repl, but not forw


Wed Jun 26 12:36:38 1985  /mtr <mrose@nrtc-isc>

	MH #5.457 becomes MH #6.1 in beta


Mon Jul  1 06:26:38 1985  /mtr <mrose@nrtc-isc>

		       -- Suggestions from Bob Desinger --
	config/mh.h, sbr/m_backup.c: locate SBACKUP definition in mh.h
	sbr/m_gmsg.c, uip/rmf.c: handle SBACKUP prefix files appropriately,
	    also, slight optimizations


Mon Jul  1 09:18:42 1985  /mtr <mrose@nrtc-isc>

	sbr/m_{file,gmsg}.c, uip/{msh,refile,rmm,send}.c: #ifdef notdef the
	    stuff marking messages as DELETED, since it's not used


Mon Jul  1 22:22:49 1985  /mtr <mrose@nrtc-isc>

	conf/doc/mhl.rf, uip/mhlsbr.c: remove the #ifdef BERK stuff and
	    generalize it.  If the MH profile entry mhlproc is defined, then
	    you get the BERK behavior using the mhlproc as the output filter.
	    Otherwise, you get the old behavior.  If this works nice, I'll
	    make "more" the default mhlproc and set-up config.c, mh.h, and
	    m_getdefs.c as appropriate (with the documentation mh-profile.rf)
	    For the moment, I'm using "less" as my mhlproc and it works
	    fine.


Thu Jul  4 08:29:11 1985  /mtr <mrose@nrtc-gremlin>

	conf/doc/ADMIN.rf: put in a word or two about running conflict
	conf/makefiles/dist: add rules for .imp files


Thu Jul  4 08:38:01 1985  /mtr <mrose@nrtc-gremlin>

	tws/{dtime,dtimep}.c: know about J{D,S}T, also under #ifdef HUJI
	    generate it if appropriate.  (from Danny Braniss)


Thu Jul  4 16:00:38 1985  /mtr <mrose@nrtc-gremlin>

	othersupport/mtrenv/mhbox/dp.debug: new file


Thu Jul  4 17:36:34 1985  /mtr <mrose@nrtc-gremlin>

	uip/msh.c: "pack" should be called "packf".


Sun Jul  7 13:26:28 1985  /mtr <mrose@nrtc-gremlin>

	uip/{addsbr,post,rcvdist,replsbr,sbboards}.c: support #ifdef BANG
	conf/examples/READ-ME: ditto


Sun Jul  7 13:31:01 1985  /mtr <mrose@nrtc-gremlin>

	othersupport/scripts/: new directory for scripts.  First is Bob
	    Designer's modifications to JLR's append script.
	othersupport/mtrenv/bin/append: remove


Tue Jul  9 18:59:21 1985  /mtr <mrose@nrtc-gremlin>

	 -- Suggested by Thomas Scott Christiansen <[e-mail]> --
	uip/msh.c: move big Msgs[] structure from auto to static
	othersupport/netnews/{.,mhbox}/: new directories
	othersupport/netnews/{mh_profile,mhbox/flupcomps} new files


Wed Jul 10 09:04:04 1985  /mtr <mrose@nrtc-gremlin>

	uip/send.c: typo defining LISTDSW


Wed Jul 10 12:07:42 1985  /mtr <mrose@nrtc-gremlin>

	uip/addrsbr.c: another de-reference of NULL fixed


Wed Jul 10 14:08:39 1985  /mtr <mrose@nrtc-gremlin>

	conf/makefiles/popsupport,popsupport/popd.{c,8c}: a bit of clean-up


Wed Jul 10 14:48:05 1985  /mtr <mrose@nrtc-gremlin>

	uip/mhlsbr.c: don't SIG_DFL signals on start-up; if run from msh,
	    this loses big, if not run from msh, then signals are SIG_DFL
	    anyway...


Wed Jul 10 20:40:53 1985  /mtr <mrose@nrtc-gremlin>

	conf/config/config.c,config/mh.h,sbr/m_getdefs.c,uip/mhlsbr.c:
	    mhl's choice of mhlproc was confusing to forw.  Add new variable
	    moreproc to fix the problem.  Note as a result, the old BERK
	    behavior is the default, you have to turn off moreproc in your
	    .mh_profile if you don't want mhl to use it


Thu Jul 11 12:34:15 1985  /mtr <mrose@nrtc-gremlin>

	conf/makefiles/uip: oops, forgot a @END: SENDMTS inside an @BEGIN: POP


Fri Jul 12 10:32:40 1985  /mtr <mrose@nrtc-gremlin>

	config/mh.h,sbr/uleq.c: change strindex() -> stringdex () to avoid
	    MMDF name conflict
	uip/{addrsbr,dropsbr,msh,rcvpack,sbboards}.c: ditto


Sun Jul 14 12:54:55 1985  /mtr <mrose@nrtc-gremlin>

		     -- Enhancements from John L. Romine --
	uip/mhlsbr.c: minor cleanup


Sun Jul 14 13:23:49 1985  /mtr <mrose@nrtc-gremlin>

	tws/dtimep.c: have "BST" mean British Summer Time not
	    Bering Standard Time


Sun Jul 14 13:44:13 1985  /mtr <mrose@nrtc-gremlin>

	othersupport/bboards: version #2.1 as suggested by Jerry Sweet
	conf/doc/{ADMIN,mf,mh-{mts,tailor},bb{exp,oards5,tar},pop8}.rf: fixes
	    suggested by Jerry Sweet


Sun Jul 14 18:51:22 1985  /mtr <mrose@nrtc-gremlin>

	uip/formatsbr.c: handle the blank line bug by hacking it (see the
	    "echo controversy" discussed above).  The solution is to insert
	    a space in truly blank lines...


Tue Jul 16 12:41:54 1985  /mtr <mrose@nrtc-gremlin>

	uip/ali.c: forgot to call mts_init() to get "everyone", and "noshell"


Wed Jul 17 09:47:19 1985  /mtr <mrose@nrtc-gremlin>

	bboards/mmdf/mmdf.2/bboards/bb_wtmail: change dist_address to
	    ds_address, dist_log to ds_log


Thu Jul 18 15:59:19 1985  /mtr <mrose@nrtc-gremlin>

	conf/mh-gen.8: typo


Sun Jul 21 15:36:24 1985  /mtr <mrose@nrtc-gremlin>

	conf/doc/{send,whom,mh-alias}.rf:  aliasing hints
	conf/doc/{{MH,tmac}.h,template}: support hints section
	conf/doc/mh-mts.rf: minor fix


Mon Jul 22 16:50:54 1985  /mtr <mrose@nrtc-gremlin>

	uip/picksbr.c: logic for handling "--header value" botched


Tue Jul 23 20:30:39 1985  /mtr <mrose@nrtc-gremlin>

	uip/whom.c: pass more switches to post.  Since they're ALL secret,
	    no documentation needed!



Wed Jul 24 14:07:31 1985  /mtr <mrose@nrtc-gremlin>