BBS in a Box 3

home *** CD-ROM | disk | FTP | other *** search

/ BBS in a Box 3 / BBS in a box - Trilogy III.iso / Files / Tele / C / Comet2.1.3 Folder / Comet / !changes next >

Wrap

Text File | 1991-11-07 | 41.7 KB | 1,489 lines | [TEXT/MPS ]

date Friday, August 31, 1990 11:05:01 AM menu.c: put getcontext() before savetext() & printtext() added CONN_SERD case to resetsendbuf() main.c: changed opn_usr to tcpopn_usr in tcp_init() call date Tuesday, September 4, 1990 2:03:11 PM main.c removed maintask var commented out tcpemdp = NULL statement in releaseconn() config.c changed ReleaseResource(reshand) to DisposHandle(reshand) in copyresource(); could cause major problems added = NULL when doing releaseconfig(), not a big deal added twp->resfid = 0 when == appresfid, not a big deal date Thursday, September 6, 1990 2:05:30 PM mactcp.c: emdp->rcvbuf = StripAddress(emdp->rcvbuf); emdp->mtstream->csParam.create.rcvBuff = emdp->rcvbuf; date Friday, September 7, 1990 3:19:25 PM config.c: added UseResFile(newresfid); after copyresource... date Monday, September 10, 1990 4:14:29 PM token.c/3270token.c: moved SYS_LOCK test to lock out all input --changed nosessionstr to char * from array to avoid Aztec arg handling problem File "::emlib:h19key.c"; Line 457 # pr25(0, nosessionstr); File "::Comet:main.c"; Line 307 # char * nosessionstr = "No Active Session"; File "::Comet:main.c"; Line 558 # pr25(0, nosessionstr); File "::Comet:scrinit.c"; Line 899 # pr25(0, nosessionstr); /* no session exists */ date Tuesday, September 11, 1990 5:50:29 PM File "::emlib:h19.c"; Line 55 # resumetokens(); File "::emlib:key.c"; Line 1846 # resumetokens(); /* abort any DELAY tokens */ File "::emlib:vt100.c"; Line 67 # resumetokens(); File "::Comet:3270com.c"; Line 197 # resumetokens(); File "::Comet:token.c"; Line 63 # resumetokens(twp) File "::Comet:token.c"; Line 98 # tm_set((int) tkptr->entry, resumetokens, emdp, emdp->token_tm); Obviously, resumetokens needs an argument! The fix for the zeroed 3270 asctoebc[] entry problem... date Thursday, September 13, 1990 4:17:52 PM config.c backed out ReleaseResource(reshand) change made Tuesday, September 4, 1990 2:03:11 PM mactcp.c added tcp_read.csCode = fsRdWrPerm; /* ioPermssn field */ before PBOpen to try to fix driver opening problems... windh = NULL; changed to iconh = NULL; in savwindpos date Friday, September 14, 1990 3:10:34 PM File "::Comet:config.c"; Line 354 # if ((newresfid = OpenRFPerm(&emdp->appfile.fName, (short) 0, (Byte) fsRdWrPerm)) == -1) { File "::Comet:config.c"; Line 376 # if ((newresfid = OpenRFPerm(&emdp->appfile.fName, 0, (Byte) fsRdWrPerm)) == -1) { File "::Comet:config.c"; Line 411 # if ( (newresfid = OpenRFPerm(&emdp->appfile.fName, emdp->appfile.vRefNum, (Byte) fsRdWrPerm) ) == -1) { File "::Comet:config.c"; Line 585 # if ( (newresfid = OpenRFPerm(&reply.fName, reply.vRefNum, (Byte) fsRdWrPerm)) == NEG) { File "::Comet:config.c"; Line 597 # if ( (newresfid = OpenRFPerm(&reply.fName, reply.vRefNum, (Byte) fsRdWrPerm)) == NEG) { File "::Comet:config.c"; Line 750 # newresfid = OpenRFPerm(&emdp->appfile.fName, 0, (Byte) fsRdWrPerm); File "::Comet:config.c"; Line 972 # if ((thedp->resfid = OpenRFPerm( File "::Comet:config.c"; Line 1071 # if ((setresfid = OpenRFPerm(setfile, (short) 0, (Byte) fsRdWrPerm)) == -1) { File "::Comet:menu.c"; Line 648 # if ((newresfid = OpenRFPerm(&openreply.fName, (short) 0, fsRdWrPerm)) == -1) { changed calls to openrfperm to use the document vrefnum rather than 0... date Monday, September 17, 1990 3:06:06 PM config.c: reordered so this does not occur in openlaunchres... appvref = emdp->downvref = WDPB.ioWDVRefNum; appvol = emdp->downvol = WDPB.ioVRefNum; appdir = emdp->downdir = WDPB.ioWDDirID; /* we always expect a connection structure to exist */ makeconn(); ICK! emdp is not valid until AFTER makeconn! ... This fixes the nasty crashing problem on Steve's machine 2.0D21 date Wednesday, September 19, 1990 1:16:20 PM screen.c: modified so TextSize sets font in screen_upd_qd config.c: switched ioWDVRefNum and ioVRefNum, later backed out (vol 4 file mgr documentation is rather confusing on the issue of what they are!...) 2.0D22 date Friday, September 21, 1990 1:45:05 PM scrinit.c: if ( (!environs.hasColorQD && screenBits.bounds.right == 512) || !indisplay(emwindow) ) { /* if the window is not in the display, move if it necessary */ added test so Mac+ window will be centered properly date Monday, September 24, 1990 3:48:46 PM atddp:at_init.c omni:arp.c changed subnet routine so it returns an int rather than an unsigned char config.c: switched WDPB vrefnum/refnum so they are correct event.c: fixed clickdelay so it is set on all iconwind mousedowns date Wednesday, September 26, 1990 3:56:17 PM tftp_srv.c: tftp_util.c: fixed MacTCP tftp routines. date Thursday, September 27, 1990 10:41:35 AM zap.asm: added fastscroll() & fastindex() routines emlib:vt100.c, h19.c, asciidraw.c: modified to use paramterized fastscroll() & fastindex() routines screen.c: fixed for vt100 so zapinvert not reset in chkinvert() vt100.c: fixed to display inverted bold correctly in ds mode. mactcp.c: enabled async sends date Thursday, September 27, 1990 5:54:55 PM mactcp.c: fixed send completion to use UserDataPtr mactcp.h: increased PBCOUNT from 1 to 8 to use multiple async sends asciitoken.c: added putflush() to ED_PASTE to hasten pastes event.c: fixed invalid handle in releasescrap(). "scraptemp == NULL"; changed to an assignment date Friday, September 28, 1990 12:11:22 PM File "::emlib:key.c"; Line 87 # flushactions(keydp); File "::emlib:key.c"; Line 1836 # flushactions(twp) File "::Comet:3270token.c"; Line 42 # flushactions(emdp); /* kill all actions remaining */ File "::Comet:3270token.c"; Line 369 # flushactions(emdp); /* kill all actions on KYBD_LOCK */ File "::Comet:3270token.c"; Line 385 # flushactions(emdp); /* kill all actions on KYBD_LOCK */ File "::Comet:3270token.c"; Line 460 # flushactions(emdp); /* kill all actions on KYBD_LOCK */ File "::Comet:3270token.c"; Line 513 # flushactions(emdp); /* kill all actions on SYS_LOCK if typeahead off */ File "::Comet:3270token.c"; Line 577 # flushactions(emdp); /* kill all actions on KYBD_LOCK */ File "::Comet:3270token.c"; Line 700 # flushactions(emdp); /* kill all actions on KYBD_LOCK */ File "::Comet:event.c"; Line 161 # flushactions(keydp); File "::Comet:main.c"; Line 555 # flushactions(emdp); parameterized flushactions() calls to use a winds * emlib:key.c: added getcontext(keydp); so macro flag shows in correct window date Monday, October 1, 1990 10:20:37 AM h19.c: added emend() when switching modes between vt52/vt100 to keep cursor alive vt100.c: added <ESC>[L (insert line) & <ESC>[M (delete line) capabilites from h19.c added emend() when switching modes between vt52/vt100 to keep cursor alive added charsetS0, decom, & wrapped flags to save/restore cursor state event.c: fixed bug in copyscreen() where a single char at beginning of line would dropped asciidraw.c: modified ins_lin & del_lin to check parameters & handle 23d line call OK main.c: em.h: eminclude.h: key.c: added vtkeyoverride so user can override host application keypad mode main.c em.h: eminclude.h: mactcp.c: menu.c: config.c: config.h: added mtcpsendasync global to control MacTCP async send menu.c: modified so color border cleared when switched from color to !color zap.asm: put 16 pt. underline one scan line higher to coincide w/ quickdraw fixed clr11 so no junk: move.>l< $3f, d1 event.c: removed SetPort(keywindow) called after GetNextEvent on suspicion that crashes where other apps crash after open when MultiFinder running and A5 is wrong for comet, but other app (MPW) seems to have had its port set to the same port as comet... date Tuesday, October 2, 1990 10:29:42 AM vt100.c: fixed so LF rather than CR does newline trick screen.c scrinit.c: added windtightfit() so that upper left corner is "home screen" and windows are forced to fit on all screens which are "tight" (512 & 9 pt., 640 with 16 pt.). This makes it easier to get fast drawing... File "::Comet:main.c"; Line 1507 # else if (devt->what == activateEvt) { File "::Comet:util.c"; Line 49 # else if (devt->what == activateEvt && devt->message == dptr) { added activeflag test to avoid crash when the routine sees another window's active event and tries to use the invalid dialog window ptr! key.c: fixed so Apple-. doesn't send a '.' after cancelling macros. fixed so 255 loopcount continues indefinitely config.c: main.c: added setwindvols() and other code so default save will occur in app directory date Wednesday, October 3, 1990 11:35:08 AM mactcp.c: added traps to debug MacTCP localtalk problems... using only 1 PB for sync send, since multiple SYNC packets wind up with iocompletion set in some pb--somehow! date Thursday, October 4, 1990 11:08:33 AM event.c: fixed bugs in table mode screencopy where extra char appeared at end of line 2.0D23 NOTE: versions after this date have the year set to 1991 because the Mac came back from an upgrade set this way and I failed to check it until it was rather late! date Thursday, October 10, 1991 9:59:33 AM h19.c: display glitch induced by vt100 mods--changed occurrences of setascattr_ds(attrib); to setascattr_ds(attrib, emdp->inselection); vt100.c: deleted tryzapline() which is obsolete... serial.c: trigger setcursor() on keycommand to fix Steve's response problems date Friday, October 11, 1991 11:46:20 AM prettywind.c: mactcp.c: em.h eminclude.h added resend count to window struct and drawinvcount() to highlight resends date Tuesday, October 15, 1991 11:26:00 AM util.c: added inerror flag to prevent nested error() calls 2.0D24 date Wednesday, October 16, 1991 7:53:45 PM event.c: serial.c: added autokey to event.c and used for setcursor() test in serial gethost.c: added bkrd_service() in DNR wait loop date Tuesday, October 22, 1991 4:48:00 PM key.c: File "::emlib:key.c"; Line 1732 # ungetaction(tokep) File "::emlib:key.c"; Line 1813 # ungetaction(); File "::emlib:key.c"; Line 1842 # ungetaction(); *** another big whoops here... screen.c: added showcursor() in endibmdraw_qd() config.c: fixed so Save global configuration saves current settings properly scrinit.c: screen.c: event.c: em.h: eminclude.h: added textwindow support event.c: changed scrapscreen() & scrapline() to use appendtextw() date Thursday, October 24, 1991 9:20:06 AM config.c: used WDPB.ioWDVRefNum rather than ioVRefNum for all SFGet/Put "resvol" references mactcp.c: h19key.c: added prerrint25() calls to fix error reporting, change close reporting to use 25th line gethost.c: changed domain name resolver to report errors added fillwind("Resolving host name ") message 2.1D1 date Monday, October 28, 1991 11:16:35 AM omni_comp.c: added cseg declaration so aGlobal stored in C code segment rather than data bad function parameters fixed... File "::emlib:macro.c"; Line 877 # if (! memtest(length), "to dump key macros") if (! memtest(length, "to dump key macros")) File "::Comet:main.c"; Line 746 # setcontext(&conns[0]); setcontext(conns[0]); File "::Comet:config.c"; Line 511 # setmsession(FALSE); unneeded, so deleted.... should have been setmsession(emdp); File "::Comet:menu.c"; Line 1180 # setmecho(); setmecho(keydp->ucb.u_echom); File "::Comet:tftp_srv.c"; Line 89 # oldA5 = getmyA5(); added declaration for unsigned long getmyA5(); File "::Comet:3270token.c"; Line 715 # entryerr(code); entryerr(tkptr); File "::Comet:screen.c"; Line 1602 # useqd(); if buildfont() || setlinestarts() ... useqd(twp) File "::Comet:token.c"; Line 226 # selreset(); *** very nasty!!! in SELWORD token routine... selreset(emdp); File "::emlib:key.c"; Line 106 # if (ibmokpnorm(thechar, keycode)) if (ibmokpnorm(keycode)) File "::emlib:key.c"; Line 117 # if (ibmokp3(thechar, keycode)) if (ibmokp3(keycode)) File "::emlib:macinit.c"; Line 110 # pr25("You're VERY short of memory!"); unused, this is probably why the growzone routine crashed... pr25(0, "You're VERY short of memory!"); File "::emlib:printf.c"; Line 40 # (*vt100)(thechar); unused... vt100(thechar); File "::Comet:event.c"; Line 989 # scraptemp = NewHandle(0); File "::Comet:event.c"; Line 993 # scraptemp = NewHandle(0); changed 0 to 0L main.c event.c emcursor.c fixed so keydp->emwindow used rather than keywindow, which could be keydp->textwindow date Tuesday, October 29, 1991 9:31:16 AM main.c ft.c added ibm_free(twp) to free ibm memory tftp_util.c added mtcp_tfdoerr() so no crash will occur on foreign host error mess gethost.c setA5(oldA5) before returning from resolved() date Friday, November 1, 1991 9:41:30 AM util.c: changed so when window is left up modaldialog = modalstate; rather than modaldialog = FALSE; to fix crashes when called within namewindow()/resolve_name()/mtdnresolve() BUT it still doesn't work... error() modified to use SetItem rather than ParamText() prettywind.c: moved macroflag to top and made showmacroflag() invert when tokens delayed token.c: added code in token_service() to invalidate macrorect when tokens in q event.c menu.c: added seltext/changed print/save routines to serve textedit window config.c: made KEYPAD3 default as per Stewart Lynn's request telnet.c: added prerr25("Executing logon macro... press Command-. to interrupt"); key.c: added code to flushactions() to cancel logon macro highlight/prerr25() serial.c: moved hostconncomplete() to main() to fix logon macro prerr25() emlib:basic.rsrc updated help text for TEXT windows menu.c: added queryconnclose() for serial session dropped Abort... and replaced it with Re-open session emcursor.c: changed so selmenuitems(keydp); * event.c: fixed menucheck routines so they all use twp args rather than keydp setfontsizemenu(twp->fontsize); /* font size */ keymchange = keymenus(keydp); setmchange = setmsession(keydp); setconnchange = setmenuconn(twp); date Sunday, November 3, 1991 12:11:40 PM main.c: changed if (emwindow == keydp->emwindow) { to if (keydp == NULL || emwindow == keydp->emwindow) { to make menus be set up correctly when no key window yet set... date Monday, November 4, 1991 12:15:27 PM config.c: changed reference to 'comet settings' to 'Comet Default' macro.c: menu.c menudefs.h added setmacro() date Tuesday, November 5, 1991 2:59:41 PM token.c: added intoken_service to avoid error() cascades... main.c: added inio_service to avoid error() cascades... screen.c: added inscreen_service to avoid (unlikely) error() cascades... date Wednesday, November 6, 1991 10:16:18 AM ft.c: textedit.c: main.c: added textupedit() and textdownedit() key.c vt100.c: menu.c: added vtkeyset() and support for user setting keypad 2.1D2 date Friday, November 8, 1991 11:13:24 AM event.c menu.c textedit.c #ifdef'ed textwindow additions to produce 2.0D25 font.dam: restored old 9pt bold font as per IR retraction of request... menu.c util.c: changed textwindow() to helpwindow() for clarity h19.c: used ypos/modflg = SCRALLMOD rather than ypos + 1 in 'J'/SUB logic: for (count = ypos + 1; count <= LASTROW; count++) vt100.c: removed emdp->color tests where clearrect() called date Monday, November 11, 1991 11:11:34 AM File "::emlib:h19key.c"; Line 276 # line25on() File "::emlib:prettywind.c"; Line 88 # if ( ! line25on() && (emdp->controlschanged || !emdp->dsdraw)) { File "::emlib:prettywind.c"; Line 593 # if (line25on()) { File "::Comet:emcursor.c"; Line 253 # if (!line25on()) File "::Comet:main.c"; Line 560 # /* if (!line25on()) File "::Comet:screen.c"; Line 208 # if (line25on()) File "::Comet:screen.c"; Line 850 # if (line25on()) added twp parameter to line25on() macro.c: added emdp->keyxhead = NULL; in keyxrelease() to be safer... menu.c: added querynewconf(); h19.c: added IBM keymode ESC sequence: case '+': { /* go into IBM keymode */ emdp->ibm_keymode = TRUE; setibm_keymode(emdp->ibm_keymode); break; } case '-': { /* exit IBM keymode */ emdp->ibm_keymode = FALSE; setibm_keymode(emdp->ibm_keymode); break; } event.c: added SetCursor(&arrow); to emwdeactivate() to avoid cursor disappearance asciidraw.c: added if (autokey) setcursor(); to emend() so cursor displays fast 2.0D25 2.1D3 screen.c: added SetPort() & TextFont() etc. in updatewind_qd() to fix font size prob date Tuesday, November 26, 1991 9:13:13 AM File "::Comet:menu.c"; Line 3517 # nvtaltcursor = '|'; changed to '[' so setting key pad with menu will work correctly. date Wednesday, December 4, 1991 11:46:24 AM mactcp.c: fix MacTCP crash on connect failure when autoclosewind: changed mactcp_wrapup() so mtreset() comes *before* cls_usr() not *after* mtreset(); cls_usr(); main.c: deleted unused global short mfwait; /* per window wait for MF updates */ asciidraw.c: emend(): changed so no longer done when fast drawing in use if (!mfbackground) ObscureCursor(); /* hide until next motion */ date Monday, December 9, 1991 11:19:15 AM config.c: code produced erroneous error message when "comet Default" created successfully if ((newresfid = makeconffile()) != -1) { /* we made the new file & set resfid, break */ if ((newresfid = myopenresfile(&setfile[0], setresvref, (Byte) fsRdWrPerm)) == -1) { /* no config, try make a config file in the system folder */ nosysconf = FALSE; error("Can't make 'Comet Default' document in System Folder'"); } date Tuesday, December 10, 1991 10:58:19 AM menu.c: added before doing anything in OPENCONN (not really necessary but parallel) getcontext(keydp); prettywind.c h19key.c: added controlschanged = TRUE where InvalRect(&emdp->toprect); called so q1 buttons redrawn OK when autologins complete date Wednesday, December 11, 1991 10:16:14 AM h19key.c: added to pr25() so that the macro-executing prompt won't get wiped out... /* guarantee it won't get cleared by a timer entry */ tm_clear(windp->timer25); /* changes to make the double click/select work properly */ token.c: added bkrd_token_service(); telnet.c: called bkrd_token_service rather than save context locally emcursor.c: now doubleclick tested before mouse is tracked and bkrd_token_service() date Thursday, December 12, 1991 11:43:10 AM asciitoken.c: ft.c: moved cft() & cft2() into ft.c where they belong token.c ft.c: moved ft_header(), ft_footer(), & send_block() into ft.c where they belong date Friday, December 13, 1991 9:36:56 AM asciitoken.c: switched showinsert() & clrinsert() in doinsert() cases so appearance correct date Tuesday, December 17, 1991 9:17:44 AM xlate.h: fixed so host 0x3b mapped to double dagger... scrinit.h: textedit.c: fixed so text window same size as emulator so paging works well, modified paging to page exactly one screen rather than the screenlines - 1 norm. em.h eminclude.h: event.c menu.c added hidetextwindow code so edit window won't resurface if you kill it 2.0D26 release for library: known bug with MacTCP close menu (menus not reset) ********************************* date Thursday, December 19, 1991 9:43:10 AM main.c mactcp.c: parameterized tcp_release() call added mtcpconnopen and other logic in another attempt to fix MacTCP auto-closing crash (last changes left session looking as if it was on when user used a File Close session command). Replaced calls to mactcp_wrapup with closeflag TRUE so cls_usr always handled asynchronously. icon.rsrc: reduced minimum size from 550 to 500 (recommended 716800) prettywind.c: reverted right hand MACRO to small format and previous location macro.c: key.c: added ALT key support so ALA character set can be handled like TinCan token.c: added RSLT_WIND to select window (menuwind()) menu.c: added some protection to menuwind() if (windnum >= menuwindowcount) return(-1); thewind = menuwindows[windnum]; if (thewind == NULL) return(-1); main.c: modified namewindow() dialog so TCP items disabled when not relevant TODO does not work properly date Friday, December 20, 1991 9:11:35 AM serial.c: removed connopen = x since this var belongs to opn_usr() & cls_usr() prettywind.c: added showaltmacro() & clraltmacro(), replacing pfshowdubshift... mactcp.c: TODO? mactcp_status() could show last status when no connection exists... but MacTCP doesn't use valid params; set statblk in mactcp_wrapup? menu.c: added static char ibm_keymode tests in setmsession() so IBM-Action/ASCII menus enabled correctly... added setemdpresvol(); to fix MF document location problems... added disposeonclose test to WMCLOSEWIND to handle async window destruction date Thursday, December 26, 1991 10:38:01 AM main.c: added releaseconfig() in newconn() so macros/hostaddr are not inherited. ************************** we return now to the correct date -- this is really 1991! date Wednesday, January 2, 1991 12:52:47 PM key.c: added SPACE -> NULL in optkeymap date Monday, January 7, 1991 11:18:12 AM screen.c: saved & restored font size in ibm_ch_draw_qd() so size always correct... scrinit.c: InitCursor(); /* guarantee that the cursor is a visible arrow */ date Wednesday, January 16, 1991 10:55:25 AM screen.c: SetPort(oldport); date Thursday, January 17, 1991 10:40:55 AM screen.c: Koehler/Hornig performance bug report--don't redraw 3270 screen promptly because color performance goes to hell! /* (*emdp->screen_upd)(); delay to enhance performance when doing LIST etc which send screen as up to 10 separate packets....*/ config.c: /* make QuickDraw the default drawing method, since dsdraw may fail catastrophically on some configurations (IIci w/2 page display) */ trydsdraw = FALSE; main.c: scrinit.c: added code to bring 16 pt window down on Portrait display... date Tuesday, January 22, 1991 12:35:42 PM zap.asm modified so library diacritics are drawn and erased correctly textedit.c: fixed so diacritics appear correctly 2.1D4 date Monday, January 28, 1991 10:05:28 AM asciidraw.c: bug found in emacs testing for CS108 by Brad Glade: scrolling in region failed to pick up last line; also deleted hazardous "optimization" changed del_lin() and ins_lin QuickDraw case; was: emdp->bigrect.bottom = emdp->scrollbottom * emdp->lineheight; now: emdp->bigrect.bottom = (emdp->scrollbottom + 1) * emdp->lineheight; date Wednesday, February 6, 1991 8:04:41 PM screen.c: added to updatewind_ds() so no refresh done unless needed... if (SectRect(&emdp->bigrect, &(*updwind->visRgn)->rgnBBox, &intersect)) also added endibmdraw_ds() at end of routine so cursor obscured... zap.asm asciidraw.c h19.c vt100.c screen.c: added SwapMMUMode calls so 32-bit mode used when zapping screen buffer... not included in zapchar() for efficiency considerations... but it doesn't work!] scrinit.c: try StripAddress() on screenbase instead... tcp/tcp.c: set resend timer to fix hanging problem reported by Bill Young... if (resend) { tm_tset(retry_time, tmhnd, NULL, tcptm); /* set resend timer */ break; } asciidraw.c: made userxoff work again; but is it OK to hang until a new event arrives? date Thursday, February 7, 1991 8:36:31 AM asciidraw.c h19.c vt100.c 3270com.c menu.c config.c: changed saveclear & savescroll to logsession per Tom Young's request fixed 3270 so PF key triggers save to make logsession work as expected saved setting in config asciitoken.c: fixed so IBM PF keys do logsession menu.c made Command-A "Select All" to be more standard and Command-S "Append Screen" so append is done when you save a screen, which seems to be the most useful default... made goaway click call menuwind(WMCLOSEWIND) rather than queryclosewind() so interface identical (user gets dialog to close session) fixed selmenuitem to reflect selection for edit window appends event.c scrinit.c added code so textwindow stays hidden event.c: in mfswitch changed InvalRect(&keydp->dubshiftrect); to InvalRect(&twp->dubshiftrect); so multifinder switches would definitely do update & enable cursor... asciidraw.c: fixed del_char(), which was blanking the wrong character... key.c: added test so eternal loops not decremented! if (emdp->loopcount != -1) --emdp->loopcount; date Tuesday, February 12, 1991 10:12:51 AM menu.c: window dialog had ba_bs sense reversed... date Wednesday, February 13, 1991 9:09:40 AM asciidraw.c screen.c: tried ShieldCursor() instead of HideCursor() to fix Steve's cursor flicker complaint... WHUPS! wrong complaint but works OK... textedit.c: modified to cut first page rather than choke when >32000 chars parameterized ScrollToLine() so cursor keys will scroll OK menu.c: added GOTOEDITEND vt100.c: cut beep() when XOFF received screen.c: changed update routines so that the cursor is reasserted if it was visible when update started... prettywind.c: mactcp.c: screen.c serial.c modified drawincount() so serial count always drawn as kchars.... made drawoutcount() & drawinvoutcount() parameterless also for consistency... event.c: menu.c token.c added selectwindow() bottleneck so emwindow deactivated before other window is activated (mousecurs turds...) event.c: called expandwindow() rather than inline code when iconwindow has mousedown incontent changed expandwindow() so HideWindow(iconwindow); replaced to make sure it will be hidden... HideWindow(twp->iconwindow); date Thursday, February 14, 1991 1:33:57 PM token.c config.c added edkillselect option so emulator selection automatically reset token.c: added selreset token to reset selection asciitoken.c: recompiled so IBM INSERT mode worked OK with key.c... date Friday, February 15, 1991 1:30:58 PM menu.c: added EDRESETSELECT global option added ibm_keymode option to control window dialog main.c: added AUTOMATIC to namewindow terminal selection routines textedit.c: fixed so automatic scroll occurs if selection at end of text; this way user selections are preserved, auto scrolling is automagic key.c: added shift-Tab to do field-backtab for Tom Young font.dam: minor changes in fonts--aligning +, -, etc. added top dot to '$'! date Saturday, February 16, 1991 1:04:00 PM scrinit.c: changed centerwind() & centerpoint() so window will center over screen rather than top window dialog.c menu.c main.c util.c moved most of the dialogs into the newly created dialog.c dialog.c: modified dialogs to show session name serial.c: fixed so modem & printer ports have separate SERD buffers to run at the same time... mactcp.c: added watch cursor for Synchronous Opens... screen.c scrinit.c event.c textedit.c: made .text window sizeable, added dv to setwsize(dh, dv) date Sunday, February 17, 1991 12:19:41 PM event.c: fixed screencopy so passwords won't be copied in new logging modes! date Monday, February 18, 1991 10:53:05 AM event.c: added textwindow case to emwdeactivate so scrap always coerced properly menu.c: fixed copy & append edit items so they will work if no connection date Tuesday, February 19, 1991 6:33:39 PM event.c: added to make keyboard more responsive to cursor motion... if (autokey) /* break out of event loop so other stuff gets serviced */ return(0); font.dam: modified 9 pt font so all the bits inside 6 bit wide boundary date Wednesday, February 20, 1991 8:44:05 AM main.c scrinit.c textedit.c screen.c: added isdiacritic[] to manage 3270 & text drawing better when diacritics are present event.c: Dick's crashing problem--HUnlock scraptemp in puttextscrap() only when necessary dialog.c: added when trydsdraw mode changed in setglobal() InvalRect(&twp->bigrect); date Thursday, February 21, 1991 4:36:01 PM setscript.asm: added to try to keep 512KEs running Sys 5 from crashing tst.w _keytron bne.s abort date Friday, February 22, 1991 11:24:46 AM menu.c: added to try to keep 512KEs running Sys 5 from crashing if (environs.systemVersion == 0) /* can't handle the nested menu */ return; 512KE now launches Comet successfully, needs more testing... date Monday, February 25, 1991 12:11:21 PM dialog.c: fixed so session name appears, and "press key" prompt is more prominent asciidraw.c vt100.c h19.c fixed so QuickDraw clears handled through modflg & del_char works... basic.rsrc: updated help text date Wednesday, February 27, 1991 12:34:22 PM asciidraw.c vt100.c fixed clrflg on del_lin() and ins_line vt100.c fixed so 1J erases current character too! date Wednesday, February 27, 1991 11:50:06 PM dialog.c fixed so setwindconf() ibm_keymode calls setibm_keymode() basic.rsrc: rearranged IBM menu to fix ResEdit re-arrangement & improved layout date Friday, March 1, 1991 10:36:01 AM dialog.c: fixed bug where windconf() dialog failed to set ASCII wrap_around... vt100.c: fixed vtnewline mode Saturday event.c: hacked in Quickdraw print routines date Wednesday, March 6, 1991 1:31:48 PM event.c: fixed getselection logic so invisible works properly token.c 3270token.c fixed so autologon sleep will work correctly (token processing must pass stdtoken() events even when SYS_LOCK is true) date Thursday, March 14, 1991 12:52:58 PM menu.c menudefs.h fixed so EraseInput & EraseEOF not switched! asciidraw.c: added emobscured so scrolling uses screen refresh rather than ScrollRect when window is obscured date Thursday, March 28, 1991 10:29:32 AM telnet.c added OPTTM so WILL TIMING MARK is returned to host... screen.c: added &= ~SCREEN_EVENT to updatewind_ds() to try to fix double-draw observed by Dick... dialog.c: fixed fillwindow so it handles large fonts without clipping the top... date Monday, April 1, 1991 9:26:22 AM event.c: fixed so save of text window uses selection added dodialog option to QDprinttext() vt100.c: added print screen feature, fixed emtestvis() date Tuesday, April 2, 1991 7:07:01 PM bigfont.dam: switched alif & ayn so they were in the right places... event.c: modified TE scrap routines so they use TEFromScrap() & TEToScrap() (sometimes copy selection/paste into TE failed!!!???) config.c: added FlushVol() calls to makeconffile() & saveconfig() to try to make more robust in case of crash (esp. makeconffile, which often gets locked on first-launch crashes...) gethost.c: found mactcp 1.0.2 crashing bug in DNR; added code to abort DNR when Command-. depressed, since it hangs when no DRN server configured! h19.c: changed newcurs() to newcursor() in h19reset() to guarantee cursor OK 3270com.c: added newibmcursor() wherever cursor_ptr is set to guarantee cursor OK screen.c: update cursor immediately if twp->cursorwait date Monday, April 8, 1991 10:39:12 AM serial.c fixed so you can reconfigure without being connected menu.c config.c serial.c added autohangup configuration option main.c mactcp.c menu.c changed windmenumark() to try to ensure mark is always set (int vs. char) h19.c added emdp->clrflg = 0 in h19reset date Friday, April 19, 1991 12:15:22 PM event.c: qdprint mods for 48-pt Laser font textedit.c: changed so textwcut() used twp arg rather than keydp! scrinit.c: modified indisplay() so iconwind won't be inappropriately forced off top of display... main.c: added tftp_direction so tntfdn doesn't do textdownedit if not download... date Tuesday, April 23, 1991 9:57:33 AM main.c: placed addwindmenu() after openconn() in newconn() so mark appears... h19.c vt100.c fixed so CFTMODE works for ASCII & also vt100... 2.1D6 date Tuesday, April 30, 1991 4:24:50 PM menu.c: added getcontext() before showtape() so tape PICT will draw OK date Wednesday, May 1, 1991 3:02:00 PM mactcp.h upped PBCOUNT to 48 from 8 so key-lossage less likely? somehow I doubt this will help... date Monday, May 6, 1991 1:39:38 PM tftp_srv.c: fixed so duplicate requests are discarded & ntftps-- when TIMEOUT asciidraw.c: fixed emtestvis() so 16 pt font works right on 640 X 480 screen added emupdwait flag so cursor crud less likely event.c: tried RectInRgn() rather than SectRect for updatewait() test dialog.c: added DlgCopy/Paste() to nmfilter() & alfilter() default dialog filters date Friday, May 10, 1991 4:57:31 PM rearranged link order so zapchar() is within segment... 2.1B1 date Monday, May 13, 1991 5:25:53 PM menu.c: added SystemEdit() so Bug report will get copied under multifinder 2.1B1 date Monday, May 20, 1991 3:39:22 PM menu.c: added SystemEdit so .edit window cut/copy gets picked up under Multifinder date Tuesday, May 28, 1991 2:46:27 PM macro.c: changed so key macro dialog help shows all codes... config.c: added wrap_around = emdp->wrap_around so wrap set correctly in default..., h19.c: fixed so downloads will work! (was reset to NORMMODE...) scrinit.c: in testds() replaced useqd(emdp) with useqd(twp) so changing to quickdraw mode in global dialog looks right... date Wednesday, May 29, 1991 2:36:58 PM textedit.c: menu.c: fixed so Edit Cut disabled when not edit window added TEClear date Thursday, May 30, 1991 3:35:17 PM main.c: added in releaseconn() 0 window case after keydp == NULL to prevent filltextwind() crash when no window up: keywindow = NULL; date Friday, May 31, 1991 1:12:17 PM scrinit.c: added code to fix crashing with IIci/2-pg display config.c: set trydsdraw = TRUE; by default since major fast-draw bug fixed!!!!! main.c: fixed fillwindow() crash by making sure valid font/size passed! 2.1B2 telnet.c: fixed so telnet connection has both LOCAL & NEWLINE true by default scrinit.c: tried dropping StripAddress() for screenbase to fix for E-Machines monitor asciitoken.c: made LOCAL mode do CR/LF, and added LF when CR received macro.c: no longer automatically interpret CR as ENTER in IBM mode. fixed so Help button enabled in setmacro() dialog bigfont.dam: closed size of 48 pt font down to 18 from 21 pixels to see if helps character spacing of QDprinttext() (QD->PostScript interpretation sux, adds variable spacing per char/space and draws it too wide!) date Thursday, June 13, 1991 3:29:44 PM config.c: went back to trydsdraw FALSE by default since E-Machines 2-page display still crashes... dialog.c: vt100.c added vt102, vt220 options to vt100 2.1B3 date Monday, June 17, 1991 1:22:55 PM dialog.c: testdsall() done AFTER DisposDialog() so fast-draw mode set OK vt100.c: fixed so the enabled terminal will be properly reported by ESC>[0c over telnet connections dialog.c: added option to switch Standard keyboard +/- keypad screwups... date Monday, June 24, 1991 1:58:26 PM emcursor.c: parameterized LASTCOL and LASTROW redefinded SCRALLMOD to be 0xFFFFFFFF rather than 0xFFFFFF to support up to 32 rows in the emulator display... screen.c: fixed so updatewait() in screen_service emcursor.c: fixed to use cupdatewait(keydp) to get correct result! date Friday, July 5, 1991 1:27:09 PM dialog.c setcolormap() now uses overwind() rather than centerwind() screen.c tickcursor = cticks + 1; rather than + 2 for better cursor tracking with color date Monday, July 8, 1991 1:37:11 PM config.c: em.h eminclude.h added colormap with approximate color matches of user-selected colors to speed performance on IIsi w/no FPU; ALAS, the problem is not FPU-less-ness but System 7.0, and this does not help! config.c: config.h: main.c: made keypadswitch a global in global conf rather than per-session date Friday, July 12, 1991 10:36:05 AM token.c: added showcursor() when EM_HIDE is done to make sure cursor visible config.c: added copyresource('CONN', 0, twp->resfid, newresfid); so the connection resource is copied to new documents main.c: removed from cls_usr() /* prettywind(TRUE); unnecessary, results in cleared area when color is on */ date Wednesday, July 17, 1991 1:09:02 PM dialog.c: added to setwindowconf escmap case so correct item is highlighted... GetDItem(wcdial, item_id, &type, &hitem, &rect); event.c: added page numbering to QDprinttext date Thursday, July 18, 1991 12:15:10 PM token.c: added so eternal loop works properly... case RSLT_LOOP: { /* emulator action key */ if (tkptr->entry == 255) /* loop forever */ tokenloop(-1); else 2.1B5 Wednesday, July 24, 1991 12:22:41 PM dialog.c:added in case GC_SAVE: so ASYNC config saved OK! mtcpsendasync = omtcpsendasync; date Thursday, July 25, 1991 2:50:52 PM mactcp.c: added to drawcount so status always updated OK... if (statblk.csParam.status.amtUnackedData == 0) twp->resending = FALSE; config.c: int tquit = quitonclose; event.c: added to screencopy(), so the end of a copy range is always NULL terminated *destp = '\000'; /* put a null at the end for C-safety's sake */ date Friday, July 26, 1991 10:34:38 AM 3270token.c: added code to NEW_LINE case so pastemode will abort properly comet2.1 event.c: screencopy() added to fix so first line of copy not duplicated if (firsttime) charcount = emdp->linelength - emdp->selxstart; mactcp.c: closing problems could be due to mtcpconnopen being prematurely reset? emdp->mtcpconnopen = FALSE; mactcp.c: changed so send pb is statically allocated so that the window struct * can be used consistently as the UserDataPtr-- the mactcp_event ASR routine may be called as a result of a timeout with a pb pointer as UserDataPtr rather than the conp window ptr... h19key.c: removed from reset25() so error messages will appear correctly on connection failure when connopen FALSE... else { pr25(0, nosessionstr); } comet2.2 date Thursday, August 1, 1991 10:45:57 AM token.c: added ShowWindow() to EM_EDIT so text window made visible if not already mactcp.c: the source of problems with MacTCP resends being bad! waitcount == 0 does not imply fillcount == 0! emdp->sendp = emdp->fillp = emdp->sendbuf; /* avoid MacTCP wrap double-send behavior */ "real" Comet 2.1 date Monday, August 12, 1991 1:58:57 PM mactcp.c: investigating problems with 3270 multiple-packet async sends, where resends are never sent by mactcp. changed ulptimeoutAction = 1 in close; contrary to the documentation, 1 is abort, 0 is report! date Wednesday, August 14, 1991 10:47:10 AM mactcp.c: always decremented when mts_done, otherwise failed calls due to broken connection lead to starvation --send_pend; date Friday, August 16, 1991 11:55:54 AM mactcp.c: added for safety's sake (shouldn't be necessary) abortPB.csParam.abort.userDataPtr = emdp; relPB.csParam.release.userDataPtr = twp; date Wednesday, August 21, 1991 12:53:07 PM event.c: token.c: re-enabled Print Draft, added token to print draft so IR folks can get properly-justified screen dumps... case RF_PRINTDRAFT: { printtext(tkptr->entry, FALSE); vt100.c: fixed send{Up,Down,Left,Right}() so alternate cursor char used w/o respect to alternate keypad mode... appkeypadmode = TRUE; hd:MPW:AzInclude:TCPPB.h: added TCP_Release data structure, not defined in MacTCP defs?!... date Thursday, August 22, 1991 10:30:14 AM event.c: removed from QDprinttext since it didn't do the job and might have side effects on non-color machines... /* emdp->color = TRUE; color on to get inverse text */ 3270com.c: whups! one of Peter's constants that I missed in read_mod()... Should fix Tom Dimock's reported Model-5 (27 X 132) bug... ptr = look_attr(scr_map + 1919, FWD); becomes ptr = look_attr(scr_map + emdp->screensize - 1, FWD); date Wednesday, September 4, 1991 10:17:22 AM FAL2 has been configured with dinky 2K windows, causing Comet 2.0D25 to break (packets over window won't be resent properly) tcp/tcp1.c: try to fix TCP failure to resend/failure to send packets over host window limit problem--added code to tcp_ex() newsend = TRUE; return; tcp/tcp.c: moved tcp_ex() into tcp.c, added windwait flag for prompt send Comet 2.1.2 for Omninet date Wednesday, September 25, 1991 1:03:15 PM telnet.c: fix for Multinet negotiation problem: Multinet sends data before completing Terminal Type negotiation if (!emdp->termset) { /* Multinet: ignore negotiation attempts if type fixed */ /* prevent attempts to open a session with the broadcast address -1 ... */ gethost.c: if (hsockp->u.addr == -1) return(-1); main.c: added for connopen MACTCP case: if (twp->fhost.u.addr != -1L && twp->fhost.u.addr != 0L) { /* avoid opening sessions to broadcast address, can induce Cornell AT-GW to produce telnet storms... */ error("Can't open Telnet session to the broadcast address; please reconfigure the session address"); } date Tuesday, October 1, 1991 1:58:08 PM serial.c: menudefs.h: menu.c: added 38.4K serial option using CTS handshake. date Thursday, October 17, 1991 5:22:10 PM emcursor.c em.h eminclude.h added notrackcurs macro to turn off cursor positioning with mouse-- some host apps can't handle cursor sequences! token.c rcodes.h added notrackcurs macro and Telnet command macros... Bill Garrison: macro to turn off cursor tracking for a session menu.c: fixed so Telnet commands Erase Char and Erase Line will work with ASCII hosts! date Wednesday, October 30, 1991 3:54:00 PM token.c: Steve Worona: fixed Select Word so it selects the portion of a word only up to beginning of line date Thursday, November 7, 1991 2:31:18 PM config.c config.h event.c added courier printing option to QuickDraw printing so 1) columns come out aligned OK 2) and you don't take time building high-res Comet font