!Thursday, 03/10/83 20:35:43 *Terry $se #Fixed several problems with displaying messages/help when the #edit buffer was empty. Added code in 'display_message' to handle #things like the "o-" command. !Friday, 04/08/83 18:19:08 *Terry $=ttypes= #Added "nec" as a termnal type, for the spinwriter. !Friday, 04/08/83 20:32:53 *Terry $Reference Manual #Updated all of the build files in =doc=/build to use the #"=fmac=/ev?*" macros instead of the "=fmac=/env?*". Changed the #files in =doc=/man to use the newer macros. Updated #=doc=/help/(bnf general) to use the newer macros also. # #This change might have effect on local documentation on other #systems. # #The reason for these changes are: #1) the newer macros are more flexible; #2) makes possible versions of the Reference Manual for the #diablo, spinwriter, or for typesetting. !Monday, 04/11/83 21:21:51 *Terry $find #Installed new version that has a "-i" option. This option causes #case to be ignored in both the search pattern and the text of the #file(s) to be searched. Reorganized the documentation. !Tuesday, 04/12/83 00:49:21 *Terry $sh, sh_ls.r, lsquit #Fixed the problem of the terminal being left in half duplex when #the sequence "ESC CTRL-P" was typed. Added a call to DUPLX$ in #'lsquit' to restore the value in Lword. !Wednesday, 04/13/83 17:15:20 *Ed Hunt $vcg_opt vcg_load3 vcg_load4 #vcg_opt: fixed 'l_state' variable under LDA_INS case in optimize # so it's set to UNKNOWN when A reg is loaded # #vcg_load3, vcg_load4: added calls to generate test and branch #for # load_l(r)shift(aa) after shift count is loaded - should # not perform shift if shift count is 0 # #rebuilt & installed vcg in =src= and in /share/c_tape !Thursday, 04/21/83 18:29:18 *Terry $mkusr, rmusr #Source and documentation moved to the attic and binaries deleted #from the system. With Primos 19, user account creation and dele- #tion totally changed; and these programs are no longer useful. !Thursday, 04/21/83 18:30:47 *Terry $dprint #Added the "-x" option back (see v8.1_history) to prevent the #initial page feed. Also added a warning about its use in the #manual entry. !Friday, 04/22/83 17:31:58 *Terry $who, ns, mon #Changed to use the new/rearranged data structures of the PRIMOS #19 version of GMETR$. !Wednesday, 04/27/83 20:16:38 *Terry $=incl=/keys.r.i, =incl=/errd.r.i #Updated to primos 19.1 compatible. !Thursday, 05/05/83 19:53:06 *Terry $=src=/lib/math #Installed Gene's new math routines as 'vswtmath.' These routines #are more accurate/precise than the Prime "equivalents". Instal- #led the manual entries and the new math library guide. !Thursday, 05/05/83 19:54:05 *Peter $=doc=/man/s1/guide.d #Informed 'guide' about the new SWT math library guide. !Thursday, 05/12/83 16:05:47 *Terry $build files for all subsystem libraries #Changed to call "lib>libedb" instead of "lib>*libedb". Prime #changed the name at Rev 19. !Thursday, 05/12/83 16:09:09 *Terry $vswtml, dacos, dasin, dbexp, dbsqrt, dflot, drand #Removed the old double precision routines from the 'vswtml' #library because the new Swt math library 'vswtmath' has outdated, #obsoleted, and outperformed them. !Friday, 05/20/83 07:56:28 *jeff $lfo #Installed 'lfo' and associated documentation. !Friday, 05/20/83 08:05:41 *jeff $sema #Replaced 'sema' with the new version that uses 'parscl' and #manipulates semaphores with initial values. Moved the old ver- #sion to the attic. Installed the new documentation. !Friday, 05/20/83 08:10:33 *jeff $pr, sp #Changed the sema command to conform to the new calling sequence. !Friday, 05/20/83 14:52:29 *jeff $ffind #Installed 'ffind' and associated documentation. !Tuesday, 05/24/83 18:37:06 *Terry $dopen$, filtst, follow, mkdir$, remove #Changed to call Primos AT$HOM instead of ATCH$$ with the KHOME #key. !Tuesday, 05/24/83 18:39:28 *Terry $t$clup, tscan$, call$$, sprot$ #Changed to call Primos AT$HOM instead of ATCH$$ with the KHME #key. !Tuesday, 05/24/83 18:40:53 *Terry $=src=/lib/cio/src/unlink.r #Changed to call Primos AT$HOM instead of ATCH$$ with the KHOME #key. !Tuesday, 05/24/83 18:42:03 *Terry $cp, lf, chat, del, passwd, chown #Changed to call Primos AT$HOM instead of ATCH$$ with the KHOME #key. !Tuesday, 05/24/83 18:43:13 *Terry $swtseg #Installed the updated swtseg for Rev 19.1. It is now installed #as 'swtseg.save'. !Saturday, 05/28/83 18:05:32 *jeff $sh, svdel, svdump, svget, svlevl, svmake, svput, svrest, svsave, svscan #Reworked the shell to create it as a set of user callable #subroutines. This was done to allow the user access certain #shell variable routines and the command interpreter routine #(SHELL). Created a new library "vshlib" and installed it in #=lib=. Documentation for the shell variable subroutines was #installed and the new library is loaded in segment 2034. The #initial executable routine is in "=lbin=" so a person can run the #shell by setting "_ci_name" to "=lbin=/sh". !Wednesday, 06/01/83 19:54:47 *jeff $dmpcm$ #Fixed to display the bad password and return labels correctly. #Previously the segment and word sections were backwards. !Thursday, 06/09/83 08:35:12 *jeff $call$$ #Added support for EPF's. Recompiled and installed on C system. !Monday, 06/13/83 00:16:30 *Terry $follow, getto #Installed new versions of these routines that use the new Primos #AT$ file system calls. Updated doc. All old functionality #remains, and these routines now save and restore the value of #Bplabel (so they callable anywhere). !Wednesday, 06/22/83 15:14:40 *jeff $sh #Installed the experimental shell library under =src=/lib/sh. !Friday, 06/24/83 17:30:04 *Terry $follow, getto #Corrected a couple of bugs. 'Follow' did not attach back home if #there was error attaching to the last node of a path. 'Getto' #did not check for an upper length limit while packing the name of #the last node into the output parameter ('name'). !Friday, 06/24/83 23:38:16 *Terry $=src=/ext.r/(batch bmerge ld guess mkcl brefs bugfm).r, comp_sub.r.i #Cleaned the code, removed unused variable declarations and #declared some undeclared functions. !Sunday, 06/26/83 00:28:43 *Terry $=src=/lcl/lib/p4clib.u/p4clib.r #Cleaned up the code some; removed unneeded declarations and fixed #some possible bugs in the following routines : p$exp, p$log, #p$sqt. These routines were not using the REAL type variables on #the stack; changed references to "store" to "storer". !Sunday, 06/26/83 00:51:53 *Terry $as6800, as8080, cal, chown, intel, lib, lk, passwd, ts, size #Cleaned up the code some; removed unneeded declarations and #declared some undeclared functions. !Monday, 06/27/83 22:02:57 *Terry $alarm, ar, cat, cmp, cn, cp, crypt, define, detab #Cleaned up the code some; removed unneeded declarations and #declared some undeclared functions. !Tuesday, 06/28/83 10:07:14 *Terry $diff, dnum, entab, fdmp, ffind, field, help, kwic #Cleaned up the code some; removed unneeded declarations and #declared some undeclared functions. !Tuesday, 06/28/83 10:08:49 *Terry $lf, macro, mkdir, mt, news, os, print, profile, publish #Cleaned up the code some; removed unneeded declarations and #declared some undeclared functions. !Tuesday, 06/28/83 10:10:27 *Terry $rdjoin,rdcat,rdextr,rdmake,rdproj,rdsel,rdsort #Cleaned up the code some; removed unneeded declarations and #declared some undeclared functions. !Tuesday, 06/28/83 10:11:47 *Terry $sema, spell, stats, template, term, yesno #Cleaned up the code some; removed unneeded declarations and #declared some undeclared functions. !Tuesday, 06/28/83 10:22:01 *Terry $=src=/lib/edt/ed.r #Cleaned up the code some; removed unused variable declarations #and declared some undeclared functions. !Tuesday, 06/28/83 10:33:20 *Terry $decode, dtoc, lopen$, ttyp$l #Cleaned up the code some, declared some undeclared functions. !Tuesday, 06/28/83 23:27:48 *Terry $define #Corrected a VERY stupid error in the dynamic storage common block #declaration. An extra variable had been added in front of the #"Mem" array. Thus all array offsets were one off from the "ds?*" #library routines. !Thursday, 06/30/83 14:40:30 *Terry $old_ar #Removed from the system and thrown into the attic. !Thursday, 06/30/83 19:22:47 *Terry $se, clrscreen.r, clear_to_eol.r #Corrected 2 typos in calls to T1OU, they were calls to "tnou". #From Logica. !Friday, 07/01/83 13:42:18 *jeff $cof$, ioinit #Added "Term_buf (Term_cp) = EOS" after "Term_cp = 1" to ensure #upward compatability with the new version of tgetl$ and tcook$. !Friday, 07/01/83 13:55:57 *jeff $block #Removed the grubby blanks from the end of each output line. !Friday, 07/01/83 17:18:13 *Terry $memstat #Removed 'memstat' from the system and moved it into the attic. #It depends on the MEM$ST GaTech gate, and nobody has put it in #yet. !Sunday, 07/03/83 16:58:47 *Terry $sema #Fixed the error messages to be like the rest of the Subsystem. #Updated doc. !Sunday, 07/03/83 20:06:22 *Terry $rp, rp.stacc #Corrected a problem with 'rp' eatting too many semi-colons in #statements like "for (;; i += 1)". The correction involved stop- #ping the next token pre-fetch with the 'stacc' dot (".") #operator. From Logica. !Thursday, 07/14/83 12:57:32 *Terry $who #Installed updated version that handles 32 character login names, #displaying login projects, and the different user types #(terminal, phantom, etc.). Updated doc. !Tuesday, 07/19/83 16:03:17 *Terry $cp #Corrected a problem with copying directories. It was caused by #an interaction with the new 'getto'. I don't quite see how it #worked before, but TA$ must have attached home for it. !Tuesday, 07/19/83 17:51:39 *Ed H. $=bin=/vcg, =lib=/ciolib, =doc=/man/s1/vcg, =doc=/fman/vcg #Installed vcg with object text generator; ciolib with vcg shift #routines; updated vcg doc !Tuesday, 07/19/83 17:53:04 *Ed H $=src=/spc/vcg.u, =src=/lib/cio #Installed source code for vcg & updated ciolib !Monday, 07/25/83 12:13:11 *roy $ring #Installed binaries of the ring commands (broadcast, execute, #setime, and terminate) in =lbin=, binary and shell file for #'ring' in =ebin= and =lbin= respectively, and put the source for #ring and its commands in =src=/lcl/spc/ring.u. !Monday, 07/25/83 12:15:32 *roy $=template= #Added =ring= template. Defined as "/sc/ring". !Monday, 07/25/83 13:18:41 *roy $primos.comi #Changed RING startup to use new procedure in SYSTEM>RING.COMI !Monday, 07/25/83 14:05:40 *Jeff $group #Installed new 'group' command. It allows a user to list his #current login groups, or to test if certain groups are active. #Installed doc. !Monday, 07/25/83 14:07:48 *Jeff $pword #Installed new version of 'pword'. This version uses the new #Primos CHG$PW routine. Totally rewrote doc. !Monday, 07/25/83 14:10:49 *Terry $the old pword command #Removed the old version and it doc from the system and put into #the attic. !Wednesday, 07/27/83 20:07:50 *Terry $se, =src=/spc/se.u/source/doopt.r #'Doopt' was calling 'saynum' with too many parameters. 'Saynum' #only has 1 argument. !Thursday, 07/28/83 14:24:38 *roy $=doc=/man/s3 #Added doc for 'ring' support commands: broadcast, execute, #setime, and terminate. !Friday, 07/29/83 08:21:05 *jeff $sh #Remove the old shell and added the shell library as the standard #shell. While I was at it, I converted everyones shell variables #to the new format. Also removed any references to the unsup- #ported version of the shell library (in =lbin=). !Friday, 07/29/83 08:23:03 *jeff $history #Modified 'history' to use the 'group' command to check for #privilege rather than looking at the user name. Validated users #no longer have to log in as SYSTEM in order to update the history #file correctly. !Friday, 07/29/83 11:42:00 *jeff $tgetl$, tcook$ #Added new 'tcook$' and 'tgetl$'. These improve functionality by #allowing an arbitrary number of characters to be returned, depen- #ding on the size of the input buffer passed as an argument. !Friday, 07/29/83 12:05:58 *jeff $tcook$, ioinit #Fixed to not print a NEWLINE character after writing out the kill #response. 'Ioinit' now initializes the default kill response to #be "\\" followed by a NEWLINE. To make the change invisible, I #changed everyone's "kill_resp" variable to contain a trailing #NEWLINE character. !Friday, 07/29/83 15:13:49 *jeff $shell #Fixed to call 'page' to page gossip. If there are less than 23 #lines then it will not pause unless the shell variable #"pause_gossip" is declared. !Friday, 07/29/83 15:47:36 *Terry $history #Put 'history' back the way I had fixed it. It checks for the #user name SYSTEM or for being logged in with group access of #".swt". !Tuesday, 08/02/83 13:33:54 *roy $=src=/std.r #Added source for 'cset' and installed it. !Tuesday, 08/02/83 13:35:05 *roy $=doc=/man/s1 #Added doc for 'cset'. !Tuesday, 08/02/83 13:46:41 *roy $=src=/std.r #Added 'gtod' and installed it. !Tuesday, 08/02/83 13:47:52 *roy $=doc=/man/s1 #Added doc for 'gtod'. !Tuesday, 08/02/83 14:48:18 *jeff $shell #Reworked the shell library. Changed the names to more meaningful #names since they no longer are compiled with 'sep' and changed #the common block files to conform to subsystem standard. Added a #'clean' procedure to clean up the directory after a build. !Thursday, 08/04/83 14:18:21 *roy $clear #Moved clear.r from =src=/spc/clear.u to =src=/std.r since 'vth' #is now part of the standard library. Also recompiled it so that #it uses the new library. !Thursday, 08/04/83 14:19:52 *roy $mon and sol #Rebuilt and installed versions that use the new 'vth' library. !Thursday, 08/04/83 14:27:40 *roy $history #Fixed 'echo's so extra "@n"s aren't 'tlit'ed off. !Tuesday, 08/09/83 09:48:23 *jeff $bind #Installed 'bind' and associated documentation. !Wednesday, 08/10/83 14:57:39 *Ed Hunt $=src=/spc/vcg.u/otg.u/src/otg$mref.r =bin=/vcg #Funked otg$mref.r to handle indirection on 2-word X- and Y- #register instructions. Installed refunked vcg in =bin=. !Monday, 08/15/83 12:23:48 *roy $ring #Fixed bug which prevented report of status of ring request for a #system that was not in the ring. !Tuesday, 08/16/83 12:32:50 *jeff $vth_com.i vth_def.i vtterm vtilin vtdlin #Included support for the 'vtilin' and 'vtdlin' subroutines which #insert and delete lines immediately to the terminal screen. This #included allocating space in the common block for hardware #insert/delete strings, allocating a couple of definitions for the #library, including code in 'vtterm' to fill in the common block, #and writing the code for the routines. !Tuesday, 08/16/83 14:32:18 *roy $ring #Improved "build" and "install" procedures to accept arguments to #select what to build and install. !Thursday, 08/18/83 08:23:36 *Ed Hunt $=bin=/vcg =ebin=/c1 =src=/spc/(c1.u vcg.u) #Fixed c1_alloc.r c1_out.r & c1_sym.r to do some things they #weren't doing. Installed new c1 in =ebin=/c1. # #Fixed static initialization in vcg_load1.r, installed new vcg in #=bin=. !Thursday, 08/18/83 10:32:16 *jeff $dynts #Added dynt for 'spawn$'. !Tuesday, 08/30/83 10:51:50 *Ed Hunt $=doc=/man/s1/(cc ccl).d =src=/ext.c/(ccl cc).c =ebin=/(cc ccl) #Fixed up 'cc' and 'ccl' to crack arguments appropriately for use #by 'c1' and 'vcg' (and 'ld'); 'ucc' is still insane. # #Revised doc & installed it all. !Wednesday, 08/31/83 08:21:45 *Ed Hunt $=bin=/vcg, =src=/spc/vcg.u/vcg.r #Changed command line argument cracking to handle named binary & #PMA output files. Installed. !Saturday, 09/03/83 18:31:54 *Terry $initswt, lutemp, =template=, lib_def.r.i #Changed '=home=' to be a dynamic template. This called for #adding a new definition (TEMP_HOME) to 'lib_def.r.i', a new entry #in the dynamic template table in 'initswt', and adding code in #'lutemp' to look up the value. # #This change will affect the '=ubin=' template also. It is now #defined as '=home=/bin'. This means that it is possible to have #different ubin's for different projects. But should =ubin= #change based on login project?? !Friday, 09/16/83 14:14:18 *Terry $swtseg #Updated 'swtseg' to the rev19.2 version. !Friday, 09/16/83 17:57:03 *Terry $copyout #Updated its version of Primos SPOOL$ to the Rev19.2 version. !Saturday, 09/17/83 19:33:56 *Terry $swt #Modified the 'swt' program to recognize user types (terminal user #or phantom) better than before. The old way depended on the #return value from DUPLX$; the new Primos UTYPE$ routine gives #better results. DUPLX$ does not seem to work from the console, #so the screen editor is not usable when a video terminal is used #as console. !Saturday, 09/17/83 19:35:18 *Terry $=incl=/user_types.r.i #Added this include file that defines the possible user types as #returned by UTYPE$ and GMETR$. !Friday, 09/23/83 20:37:04 *Terry $getto #Corrected a bug which prevented attaching to a specify MFD for a #non-existant top-level UFD. E.G. "mkdir /part/new_dir" was not #working. 'Getto' now attaches to the MFD first and then checks #to see if it needs to go further. !Tuesday, 09/27/83 14:38:22 *Ed Hunt $=lib=/ciolib(strtok) #Minor surgery on 'strtok' to make it skip leading separators..... !Wednesday, 09/28/83 17:30:55 *Ed Hunt $=bin=/xcc =ebin=/xcc =doc=/man/s1/xcc.d //cmdnc0/cc #Installed Prime's C compiler. The object files are collected #under /sc/cc; the compiler is spread about the file system, with #important hunks in //cmdnc0, //system (shared segments 2175, #2176, 2177), //syscom, //lib, //segrun*, //syscom (?). # #Installed SWT interludes to call the compiler and load the resul- #ting object code. 'Xcc' only knows about some of the more useful #features of the compiler.... it doesn't know about the debug #flag, statistics, and some other options. If you want these, you #better call the compiler yerself from PRIMOS (yecch!). !Monday, 10/17/83 18:54:18 *Terry $cobcl #Corrected a typo in 'cobcl'. It was calling 'cobc' with the #source file having a ".f" suffix. From St. Joseph's. !Friday, 10/21/83 13:27:11 *Ed Hunt $=src=/lcl/spc/lfo.u/lfo.r =doc=/man/s3/lfo.d #Hacked 'lfo' to search for a user name given as a command-line #argument (in addition to printing pid info). Installed. !Monday, 11/07/83 12:50:16 *Terry $ld #Changed 'ld' so that the share filenames can be up to 28 charac- #ters long. This changes affect the "-g" option and also the #normal "FILE ID" name. This functionality is new as of Primos #19.2. !Wednesday, 11/23/83 19:52:51 *Terry $tscan$ #Changed 'tscan$' to only check the type bits in the ECW instead #of checking the whole word. At 19.2 the number of words in the #directory increased to 27 and the ECW returned by DIR$RD (or #RDEN$$) changed from 8r1030 or 8r1430 to 8r1033 or 8r1433 (for #file or acat respectively). Bug was reported by St. Joseph's #College. !Monday, 05/28/84 11:50:25 *arnold $ptar.r ptar_com.r.i #Installed them in =src=/lcl/std.r. Doc put in #=doc=/man/s3/ptar.d Look at the doc to see what it does. !Sunday, 06/17/84 19:00:23 *Arnold Robbins $se #Added multiple features to it: # 1) trailing delimiters are not needed # 2) reads commands in =home=/.serc for startup # 3) can call the shell # 4) can switch between UNIX and SWT patterns and commands # 5) global substitutes can continue if one fails # 6) auto indent # 7) the join command defaults to a blank # 8) 's' by itself means 's//&/p' #The help file (se.d) has been updated. #Totally reworked the files in =doc=/se_h, for a massive improvement !Sunday, 06/17/84 19:06:52 *Arnold Robbins $vedtlb #Added to the 'ed' editor the same features as I put in the 'se' #editor, with the exception of the .serc file and auto indent. #Updated the help file (ed.d). # #Any programs which use the vedtlb must now also load the vshlib, #since the editor can now call the shell. Changed the help on the #edit subroutine to reflect this. !Monday, 06/18/84 12:42:01 *arnold $cat #Fixed it (finally!) to print an error message on ERROUT if it #can't open a given file. Fixed the help page also. It's about #time, too! !Monday, 06/18/84 13:10:41 *arnold $ed guide #Revised the guide to reflect the changes made in 'ed' and 'se'. #Installed the guide. Got revision bars first. Documented the #join command also. !Tuesday, 06/19/84 13:43:53 *jeff $sh #Changed the shell to have "while ... endw" and "repeat ... #until" loops. This involved writing a couple of routines to scan #forward in the current input file and squirrel away the commands #until we hit the loop terminator. Then rewind the input file and #call the shell after the loops have been turned into labels and #goto's. You can call exit to terminate a loop but you cannot use #a goto to jump out of a loop. !Tuesday, 06/19/84 14:21:06 *arnold $cset #Installed it, and the help file. It is now an official part of #Software Tools. !Tuesday, 06/19/84 15:59:44 *arnold $fmt, fmt.d, fmt guide #Installed new version of fmt with traps, condition handling, and #more functions. Updated help page, and the fmt guide. See guide #for details of new features. !Thursday, 06/21/84 15:11:49 *arnold $spell #Fixed it to handle trailing 's on a word. Also now checks for #=new_words=. If it exists, spell treats it as a file to append #unfound words onto. Fixed the help page. Added =new_words= to #the template lst. !Thursday, 06/21/84 15:13:12 *arnold $shar #Wrote and documented a shar program, which will create a 'shell #archive'. It's installed. !Thursday, 06/21/84 15:45:19 *arnold $sprint #Updated to handle all the new characters now in 'fmt'. Fixed the #help page as well. !Friday, 06/22/84 14:02:42 *arnold $pg, page #Fixed the page subroutine to act like the editors and allow mis- #sing final delimiters. Changed the help on pg and page to #indicate this. The code for pg itself didn't need changing, #since page does the real work. !Wednesday, 06/27/84 15:07:16 *arnold $=aux=/spelling/?* #Updated the dictionary from the 'new_words' file and the now #deleted 'new_dictionary' file. I culled words from gt.a and #gt.b, along with the words from gt.c. About 130 words altogether #were added. !Monday, 07/09/84 10:00:40 *Arnold Robbins $//src/std.r/news.r #Fixed the call to 'page' to use the vth option !Tuesday, 07/10/84 17:18:09 *Arnold $xcc and its files #Rewrote xcc in ratfor so that customers could compile it. Made #it into a compile-only interlude. Two shell files, xcc and xccl #do compile and compile+load respectively, while =ebin=/xcc #produces the primos command. Rewrote the xcc.d help page, and #added one for xccl. !Tuesday, 07/10/84 17:31:57 *Arnold (Friend of the Arithmetic IF) Robbins $ld #Add -w option to support programs compiled with Prime's C com- #piler, in particular for xccl to use when loading programs. #Fixed the doc also. !Tuesday, 07/10/84 17:35:13 *arnold $se, =vth=/?*, =ttypes= #Added code for several new terminals, gotten from the University #of Iowa. Terminals added were several ADM terminals, another HP, #Infoton, Terak micro computer, and maybe some others. I did this #a while back, so I don't remember exactly. I also added the #line_delete and line_insert capabilities to the vth files for #terminals which support it. !Wednesday, 07/18/84 13:43:56 *arnold $=src=/std.sh/source.sh #Rewrote it to be smarter about 1) the number of arguments passed #to it, and 2) about extracting source files from archives (in #particular, =src=/lib/swt/src.a). It now uses repeat loops and #all kinds of other fancy shell stuff. !Wednesday, 07/18/84 14:08:24 *arnold $fmt and its paraphenalia #Restored the old fmt from the attic. The new one has bugs in the #trap handling which screws up the footers and headers. The new #one has been installed as 'xfmt' in =lbin=, =src=/lcl/spc/fmt.u #and the other places. The guide is now 'xfmt'. (sigh) !Thursday, 07/19/84 10:21:57 *arnold $//src/std.r/mail.r #Fixed mail to use the VTH option with page. I don't think the #doc needs to be changed, but I may. !Thursday, 07/19/84 10:41:08 *arnold $=doc=/man/s1/source.d #Updated it to reflect the way the source command now works. I.e. #can do both commands and subroutines, and can do more than one. !Thursday, 07/19/84 18:05:34 *arnold $//src/std.r/which.r --> //src/spc/which.u/?* #Moved the which command into the spc directory, since it requires #being loaded with the vshlib. Created the necessary build, #clean, install and loc files also. !Thursday, 07/19/84 18:48:30 *arnold $=doc=/man/s3/ap.d, s5/cck(1 2).d s6/pg$brk.d #Created these files to document the corresponding programs and #subroutine. !Friday, 07/20/84 15:59:26 *arnold $vswtml: gcd, invmod, prime, pwrmod, set_?* (sect 4) #Integrated the routines remaining in the old vswtml with the new #routines Gene wrote in vswtmath. There is now only one math #library; vswtmath. Updated the documentation to reflect the #library that they are in, and the location of the linkage file #for the set_?* routines. I left them in section 4, even though #the other routines are in section 2. I will try to write an #addendum or some such for the math library user's guide that Gene #wrote. !Wednesday, 07/25/84 07:30:04 *jeff $sh, hist.r #Added the module 'hist.r' to the shell source. A single call to #a history routine was added to 'get_cl' in 'ci.r' to accomodate #the insertion of the history into the command line. Recompiled #and installed as the standard shell since we are short on time #for testing. !Wednesday, 07/25/84 07:42:17 *jeff $hist_cmd.r #Added the 'hist' command for the shell. Documentation for all of #this history stuff will be forthcoming. !Wednesday, 07/25/84 09:29:45 *arnold $fmt, xfmt, guides, help pages for them #1) Moved old fmt (source, fmt.d, fmt_guide) to the attic. 2) Did #the same for xfmt. 3) Took xfmt, removed the code for trap hand- #ling, saving and restoring the output buffer, and marking and #returning lines. Left in the code for appending to macros, and #the .if command. Installed this version as plain old fmt. Fixed #the help page for it. 4) Took the fmt guide and updated it for #.if, .am, and the new functions and special characters. Wrote a #brand new section on the =fmac=/ugm macro package, including #=fmac=/(ugh ugnh). Made mention of the report macros, and the #thesis macros in =fmac=/gt_thesis. 5) Installed the thesis #macros there. They could use some work, but I don't have time at #the moment. 6) The new fmt has been built and installed. Xfmt #was deleted from =lbin=. !Wednesday, 07/25/84 09:37:43 *arnold $=doc=/guide/math/?* #Added an addendum to the math guide discussing the routines which #were previously in vswtml. Basically, mentioned the deleted #routines, then did .so's on the macros and help pages for the #remaining routines. !Wednesday, 07/25/84 10:24:56 *arnold $=src=/std.sh/history.sh #Changed history to use repeat loops, and to crank up the screen #editor on the description file. This is especially useful when #the initial entry wasn't quite right; the user can edit his #previous description. !Wednesday, 07/25/84 11:14:17 *arnold $=src=/std.r/help.r #Fixed 1) problem with common (an array wasn't being declared in #one routine which included the common block, causing a "smaller #redefinition of common" error from the loader, and causing page #to occasionally screw up). and 2) to check for EOF return status #from page, so that it will exit immediately. !Tuesday, 08/14/84 16:04:40 *arnold $vcglib #Fixed the alloc.s and free.s files to be closer to the ones in #the C Standard I/O Library. I.e, the alloc now does some error #checking, and they now both have code to set the debug info (i.e. #the name of the routine for the runtime stack dump). I also #added the shifts.s file from the C lib, since Ed Hunt had changed #VCG to use these, instead of having it build a shift instruction #and EXECing it. !Thursday, 08/16/84 11:49:52 *arnold $cc, ucc, ccl, and new module, 'compile' #Backed up everything in //src/ext.c to the attic, along with #//doc/man/s1/(cc ucc ccl).d and //src/std.sh/(cc ucc ccl).sh. #Put in totally new versions of cc.c and ucc.c. Rewrote ucc.c to #use new utility 'compile'. Put compile.c in //src/ext.c along #with ucc.c and cc.c. Changed ccl.sh to just call cc and ld. #Fixed cc.sh and ucc.sh, and created compile.sh. Put the shell #files in //src/std.sh. Rewrote the doc for cc, ucc, and ccl, and #wrote doc for compile. # #Installed everything out in the live bins, and rebuilt the #documentation. Boy were things a mess! But it's all better now. !Friday, 08/17/84 14:19:16 *arnold $change #'Change' had its own definition of the constant MAXPAT, which was #considerably *smaller* than the current definition in #=incl=/swt_def.r.i. This caused it to bomb on large patterns, #e.g. the 'loc' file for the C Standard I/O Library. Deleting #this definition, and using the standard one, allowed things to #work properly. !Friday, 08/17/84 14:20:47 *jeff $vswtlb, nvswtlb #Added the subroutines 'gfdata', 'sfdata', 'mkpacl', 'mksacl', and #'lookac' to the shared and non-shared library. Documentation for #'gfdata' installed. The rest of the doc is forthcoming. !Friday, 08/17/84 14:23:30 *jeff $lacl, sacl, touch, quota #Installed them and placed their documentation in section 1. !Friday, 08/17/84 14:25:40 *arnold $c$main #Backed the old one up into the attic. Installed a new one that #is 1) commented, 2) better coded (less duplication), and 3) knows #about the environment stufff for SWT 9. !Friday, 08/17/84 14:25:46 *jeff $bind #Fixed to accomodate Rev 19.2 BIND. !Friday, 08/17/84 14:27:34 *jeff $swt_def.r.i #Added the includes and support for all the new file system #routines and for the inclusion of the VTH library into the stan- #dard library. !Friday, 08/17/84 14:32:00 *jeff $ld.r, swt_com.r.i #Changed the location of swt$cm (probably much to the chagrin of #our customers). This was to accomodate the inclusion of the vth #library into the standard library and because the shell's common #blocks were expanding to the point they threatened to run into #swt$cm. Everything was recompiled and reloaded, in addition to #'ld' which now thinks swt$cm is at "4040 10000". !Friday, 08/17/84 14:32:35 *arnold $=incl=/(swt_def.c.i stdio.h ctype.h setjmp.h) #Backed these files up into the attic (there is a cstuff.u direc- #tory there, for all the old C stuff). Installed new versions of #these files, and some others (see next history item by me). The #main reason for this is that I have totally reorganized these #files. A full discussion is in the C guide (which I will install #shortly. Stay tuned to your current history channel!). !Friday, 08/17/84 14:35:04 *jeff $vcg_otg.r, vcg_otgmem.r #Split each of these routines into two parts (vcg_otg1.r, #vcg_otg2.r, and vcg_otgmem1.r, vcg_otgmem2.r). 'Rp' would run #out of dynamic memory space because of the huge number of symbols #they use and the fact that I included so many more symbols into #the swt_def.r.i when including the VTH library and the new file #system routines. !Friday, 08/17/84 14:46:42 *arnold $=incl=/(ascii assert keys math memory swt swt_com vcg_defs).h #Put all these files out into the =incl= directory. The compile #time facilities available to the C programmer are now greatly #improved, closely rivalling what the Ratfor programmer has access #to! (Cheers, stirring band music, etc.) The C guide (coming #soon to a history entry near you) explains what all these things #are. !Friday, 08/17/84 14:51:07 *arnold $//src/lib/cio/?*, =lib=/(ciolib nciolib) #Backed up the C run time library to the attic. Installed new #version. This version is 1) written entirely in C and assembler, #2) uses the new math routines, 3) implements almost everything #reasonable in section 3 of the UNIX User's Manual for System V, #Release 1, 4) has stuff to do the environment, a la Unix, and 5) #everything is initialized at compile time, not run time. Has new #build file, and added a loc file. Installed the new library. #See the new guide for details of what the library contains. !Friday, 08/17/84 14:54:55 *arnold $//src/ext.r/ld.r #Modified the -b option ("do C runtime stuff") to automatically #load the vswtmath library (the C math routines call them), and #the vshlib library (the environment stuff needs it), along with #the regular c$main startoff and ciolib run time library. This #way ccl, ucc, and compile don't need to know too much about what #C programs need, just the -b option to ld. !Sunday, 08/19/84 11:42:33 *arnold $//doc/man/s1/ld.d #Added info to the doc on the -b option to indicate that ld will #include "ciolib", "vswtmath", and "vshlib" automatically whenever #a C program is loaded. Changed the date in the .hd macro; I for- #got to (apparently) the last time I changed the doc (when the -w #option was added). !Sunday, 08/19/84 16:49:23 *arnold $=doc=/guide/cc/?*, =doc=/fguide/cc #(This is it! The history entry you've been waiting for!) I bac- #ked the stuff in //doc/guide/cc up to the cstuff.u directory in #the attic. The guide was in absolutely disgusting shape. #Replaced it with an almost totally reworked guide for the com- #piler. Formatted and installed it. # #The new guide documents the existence of the new 'compile' inter- #lude, has a whole new chapter on the compile time facilities #available (the include files), discusses loading C programs for #bare Primos, and contains a totally reworked chapter on the run #time library. I also just cleaned up the guide itself, in terms #of how it made use of the ugm macros, and how some of the stuff #was formatted. It is now somewhat easier to deal with. Where #appropriate, there are margin bars; but for the compile time and #run time chapters, the entire chapters have margin bars. The #guide is now somewhat longer, but it is considerably more com- #prehensive, and better organized. !Tuesday, 08/21/84 14:53:24 *arnold $//src/std.sh/rfc.sh, //doc/man/s1/rfc.d #Created and installed these files in the source and man direc- #tories, and then in the bin and fman directories. This program #is rfl-without-the-l, i.e. just ratfor and compile, but don't #load; useful for multi-module programs where you just want to #recompile one module and then link everything together. It is #something that should have been done quite a while ago (in my #opinion, anyway). !Tuesday, 08/21/84 15:39:15 *arnold $//src/std.r/cat.r, //doc/man/s1/cat.d #Added a '-s', for 'silent' option to 'cat', at Jeff's request. #This is so shell files, like sep and the build procedures, don't #give unnecessary "foo: can't open" messages. Fixed the doc as #well. !Tuesday, 08/21/84 15:40:24 *arnold $//src/std.sh/sep.sh, //bin/sep #Fixed to use the 'cat -s' option. (See previous history entry.) !Wednesday, 08/22/84 15:23:43 *arnold $rd(att avg count diff int max min nat sum) #These are the new RDB programs that Peter had worked up, but #hadn't installed yet. I installed them and their doc in the #appropriate source directories. They will get (re)built and put #in the right directories later. We have set up some cron com- #mands to rebuild the subsystem and the manual and guides every #night. !Friday, 08/24/84 18:23:50 *jeff $hist.r #Fixed bug where if there were two failures in a row in a history #substitution, history went into a loop with breaks disabled. #Also added code to dispose of the substitution string when we're #done with it. This prevents losing some linked string space. !Friday, 08/24/84 19:48:54 *jeff $history.sh #Fixed 'history' to use the appropriate editor for the #individual's terminal, instead of always using the screen editor. !Friday, 08/24/84 19:58:49 *jeff $source.sh #Fixed 'source' to declare the variables 'file', 'curarg', and #'srcfil'. Otherwise they would change the value of a user's #variables that happened to be named the same. !Friday, 08/24/84 20:44:16 *jeff $isph.r #Changed to use subroutine call 'isph$' to determine if it is a #phantom. !Tuesday, 08/28/84 17:57:30 *jeff $stacc #Installed the new 'stacc'. It now handles Pl1g, Plp, and Pascal. !Tuesday, 08/28/84 17:58:29 *jeff $rmfil$ #Fixed to understand how to delete access categories. !Thursday, 08/30/84 08:18:13 *jeff $del #Fixed to understand the delete protect switch and how to force #privilege on acl protected items. !Thursday, 08/30/84 08:29:28 *jeff $del.d, rmfil$.d #Updated the documentation for 'del' and 'rmfil$'. !Thursday, 08/30/84 11:44:56 *arnold $=src=/lcl/spc/wallclock.u/?*, =doc=/man/s3/wallclock.d #This is Jeff's large CRT clock program. I've installed and #documented it, and called it 'wallclock', to avoid conflict with #the previously existing 'clock' program. This stuff went into #the local section and bin, not the standard "supported" stuff. !Friday, 08/31/84 16:04:21 *jeff $del.r #Finally fixed 'del' (again) and it really works this time. #Updated documentation. !Sunday, 09/02/84 17:34:42 *arnold $//doc/guide/v9/guide #Finished writing the Version 8.1 to Version 9 Conversion Guide, #and installed it in the guide/ and fguide directories. Also told #=doc=/fguide/general about it. The conversion guide is about 20 #pages long, and is accurate, as far as I can tell. !Thursday, 09/06/84 10:59:51 *arnold $=src=/std.sh/source.sh #Fixed a bug in the handling of the library source archive. If #the file isn't in the archive, it was just printing the whole #archive. It no longer does this.