github.com/PDP-10.its
1
0
Fork 0
mirror of https://github.com/PDP-10/its.git synced 2026-01-11 23:53:12 +00:00
Code Issues Releases Wiki Activity

Add some documentation files.

Browse Source
This commit is contained in:
Lars Brinkhoff 2018-09-21 09:26:52 +02:00
parent ef92edf3fa
commit 823ed80c6b
9 changed files with 653 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
Makefile
View File

@ -30,7 +30,7 @@ DOC = info _info_ sysdoc sysnet syshst kshack _teco_ emacs emacs1 c kcc \
chprog sail draw wl pc tj6 share _glpr_ _xgpr_ inquir mudman system \
xfont maxout ucode moon acount alan channa fonts games graphs humor \
kldcp libdoc lisp _mail_ midas quux scheme manual wp chess ms macdoc \
aplogo
aplogo _klfe_
BIN = sys2 emacs _teco_ lisp liblsp alan inquir sail comlap c decsys \
moon graphs draw datdrw fonts fonts1 fonts2 games macsym maint imlac \
_www_ hqm

26
doc/_info_/carpet.recent Normal file
View File

@ -0,0 +1,26 @@
SLIGHT IMPROVEMENTS TO CARPET 4/30/76 --RG
TWO NEW COMMANDS HAVE BEEN ADDED.
300:PPRI (FOR EXAMPLE) SETS THE BPT VECTOR PRIORITY TO 300.
(IT IS NORMALLY 340.) THIS COMMAND IS USEFUL IF THE 10-11
INTERFACE TO THE PARTICULAR 11 IS CONNECTED TO OTHER THAN
NPG. IF IT IS, SETTING THE PRIORITY TO 340 WOULD LOCK OUT
THE 10 COMPLETELY, CAUSING LOSSAGE. AT THE PRESENT TIME,
THE LISP MACHINE GT40 IS THE ONLY 11-END NOT CONNECTED TO BG7.
IT IS CONNECTED TO BG7, SO 300:PPRI IS THE RIGHT THING.
:BOOT CAUSES CARPET TO GRONK IN THE BOOTSTRAP. THIS IS USEFUL IF
IT HAS BEEN CLOBBERED, OR IF CARPET DID NOT LOAD THE PROGRAM
ORIGINALLY WITH A ;L . WHEN :BOOT IS TYPED, CARPET WILL PUT THE
BOOT AT WHAT IT THINKS IS (NEAR THE) TOP OF MEMORY, AND TYPE:
BOOT LOADED, HIT START AT XXXXX.
CARPET THEN HANGS WAITING FOR A TYPED CHARACTER. THUS, IT IS
"SAFE" (GIVEN 10-11 INTERFACE LOSSAGE WHEN TALKING TO STOPPED MACHINES)
TO STOP THE 11 AND FOLLOW THE INSTRUCTION.
WHEN A CHARACTER IS TYPED, IF CARPET THINKS IT HAS "MADE CONNECTION"
IT WILL TYPE AN ENTRY MESSAGE, OTHERWISE A QUESTION MARK.
ADDITIONALLY, THE ROUTINES WHICH EXAMINE AND DEPOSIT IN THE 11 HAVE BEEN
HAIRED UP SOME. THEY ATTEMPT TO CHECK ALL TRANSFERS, AND TRY AGAIN IF LOSSAGE
OCCURS, ALSO ADVISING THE USER. ALSO, SINGLE WORD DEPOSITS ARE DONE, SO
LOSSAGE THAT FORMERLY OCCURRED READING IO-REGISTERS WHICH ARE AFFECTED BY
BEING READ MAY BE AVOIDED.

388
doc/_info_/tj6.info Normal file
View File

@ -0,0 +1,388 @@
TJ6 Recent Changes 1_ NNTJ6 2 -- 8 June, 1974
or, T_h_e_ _N_e_w_ _R_e_g_i_m_e_
There are important new changes to the text justifier. These changes are
incorporated in NNTJ6. After a while, NNTJ6 will become the standard TJ6,
so use NNTJ6 and help get the bugs out.
Most of these changes are designed to straighten out the mess which is the
insides of TJ6. Changes visible to the user are:
NNTJ6 is half p_u_r_e_ _c_o_d_e_. Hopefully this will speed it up in certain
circumstances.
It has some n_e_w_ _c_o_m_m_a_n_d_s_:
._P_W_I_N_C_H_ _n_ -- sets paper width (.PAPERW) in tenths of an inch. E.G.,
.PWINCH 8.5 sets the physical paper width to 8.5 inches.
._T_W_I_N_C_H_ _n_ -- sets text width (.LINE)
._P_L_I_N_C_H_ _n_ -- paper length (.PSIZE)
._T_L_I_N_C_H_ _n_ -- text length (.PAPER)
._S_I_N_C_H_ _a_,_b_,_c_,_d_,_e_ -- sets paper width, text width, paper length, text
length, and left margin in inches respectively to a,b,c,d, and e.
E.G.,
.SINCH 8.5,6.,11,8.0,1.5 sets the physical paper size to 8 1/2x11,
the text size to 6x8, and left margin to 1 1/2 inches.
._F_I_N_C_H_ _n_ -- sets figure size in inches.
._F_I_G_U_R_E_ _n_ nows treats its argument as an absolute number of lines
(irregardless of whether you are single or double spacing).
._F_O_N_T_ _n_ _F_i_l_e_ _N_a_m_e_ -- associates font n with the appropriate font file.
This is used to set font height and width, for use by various size
commands, and to set the font name for the XGPLPT command page.
.FONT 1 LGI makes font 1 FONTS;LGI KST.
(This command actually causes the entire font to be read in and
groveled over, so it takes a while to execute. The only legal
device names are AI and DSK.)
._V_S_P_ -- sets interline space. Units are XGP raster dots, 1/192nds of an
inch. (Default is 6.)
._H_V_W_ _a_,_b_,_c_ -- sets the height of font 0, VSP, and width of a space in font
0 respectively to a, b, and c. The defaults are 25, 6, and 16 for
the XGP, and 24, 6, and 19 for non-XGP. These data are used by the
commands which set size in inches. .HVW is a quicker way of
reading in this information than .FONT.
X_G_P_L_P_T_ _c_o_m_m_a_n_d_ _p_a_g_e_ -- If TJ6 thinks it's outputting to the XGP, (i.e.,
output device is DSK: and there was either a /X in the command line
or a .XGP command,) it automatically outputs a page of XGPLPT
commands:
TJ6 Recent Changes 2_ NNTJ6 2 -- 8 June, 1974
;RESET
;TOPMAR 0
;BOTMAR 0
;LFTMAR 0
;NLINES n
;VSP n
;KSET any font names you have specified with .FONT
;SKIP 1
To prevent TJ6 from doing this, use:
._N_O_S_P_E_C_ -- turns off automatic output of XGPLPT commands.
._U_L_F_O_N_T_ _n_ -- causes printing in font n to be underlined if output is not to
the XGP. Default is .ULFONT 1. (This is helpful if the ulfont is
italicized.) (To turn off, make n <0 or >3.)
_ _(_c_o_n_t_r_o_l_ _Y_)_ -- in text command character which begins underscore.
_ _(_c_o_n_t_r_o_l_ _X_)_ -- in text command character which ends underscore.
Note that underscoring is treated differently by TJ6 and the XGP than
backspacing and overstriking with underbars. For underscoring, TJ6
outputs special XGP underscore commands. For overstriking, TJ6
backspaces and overstrikes. (Complicated, huh?) For most fonts,
the _ character is defined to look like underscoring. For a couple
it isn't. So if you mean underscore, do it, and if you mean
overstruck _, do that. (Backspacing and overstriking of all
characters once again win in .NOFILL mode.)
All the in text command characters should now work on centered, spread,
etc., lines, and in the header. Those in-text characters which
make sense to non-XGP devices are:  (control Y), begin underscore;
 (control X), end underscore; and  (control F), font switch,
which makes sense because of .ULFONT). Other in-text commands are
ignored if not to the XGP.
H_i_d_d_e_n_ _h_y_p_h_e_n_s_ work!
^_C_ _(_c_o_n_t_r_o_l_ _C_)_ is now read by TJ6 like any other input character, but TECO
makes these difficult to insert. Also they are not typed out in
block mode .IOTs to teletypes, due to a system "feature".
Characters which have been quoted (with the quoting character), will be
printed on the XGP in "normal" mode. E.G., quoting character
followed by a backspace will print as lambda, rather than
backspace. To turn this off use:
._N_O_R_M_A_L_ _n_ -- if no argument or n=0, turns off "normalization" of
quoted characters to the XGP, otherwise turns it on. This command
does not break.
F_L_U_S_H_E_D_ _C_O_M_M_A_N_D_S_
TJ6 Recent Changes 3_ NNTJ6 2 -- 8 June, 1974
The following commands have been noops for a while, and are now gone:
.BSPDIR (Back spacing is always direct.)
.QUICK
.TABS
.TABSET
.C2471
.K2471
.S2471
.T2471
.CBALL
.KBALL
.SBALL
.TBALL
The following have simply been flushed:
.PAUSE
.STOP
.STPCHR
.TR
.TRANS
.TSTOUT
The following "abbreviations" and command names have been flushed in favor
of others:
.BP -- use .BEGIN
.CM -- use .C or .COMMENT
.CO -- use .FI or .FILL
.CP -- use .BLOCK
.RA -- use .RI or .RIGHT
.APPND -- use .AP or .APPEND
.DUMCHR -- use .DUMMY
.QTECHR -- use .QUOTE
.PL (for .PAPER) -- use .TL (.PAPER will soon go away)
.PS -- use .PL (.PSIZE will soon go away.)
.L -- use .TW (.LINE will go away.)
.PAPERW -- use .PW (.PAPERW will go away.)
The prior four new names are mnemonic for T_ext L_ength, P_aper L_ength, T_ext
W_idth, and P_aper W_idth. The following have been changed:
._P_H_P_1_ -- again takes no argument, and simply causes the header to be
printed on page one.
._S_E_C_T_ -- now does a .BLOCK 2, so that a section title will not be the last
line on an output page.
._X_G_P_ and /_X_ -- now are equivalent. If output is to DSK, they: do a .SINCH
8.5,6,11,8; a .HVW 25,6,16; and try to output an XGPLPT command
page.
Handling of s_p_a_c_e_s_, ?_ _._ _!_, and ;_ _:_ has been changed. In .NOFILL mode, they
continue to get no special treatment. In fill mode, they have two
spaces inserted afterwards only if followed by a carriage return or
TJ6 Recent Changes 4_ NNTJ6 2 -- 8 June, 1974
line feed. There is a n_e_w_ _m_o_d_e_, .HALFAD, in which the text is
filled, but not adjusted, and in which spaces and ?_._!_:_;_ are treated
the same as in .ADJUST mode. The gist of this treatment is that
two spaces will be inserted after ._!_?_:_;_ at the end of a clause or
sentence, and that more than one space or tab in a row (except at
the start of an input line,) is treated like one space. NNTJ6's
algorithm for determining the end of sentence or clause is:
A_B_._C_D_
The . is always treated like the end of a clause or sentence if C
is a carriage return or line feed; otherwise, This is n_o_t_ the end
of a sentence or clause if:
1 B is a number;
2 B is a letter and A is a space, tab, or one of ._!_?_;_:_;
3 C is not a space, tab, carriage return, line feed, or one or more
of the following: )_"_'_]_;
4 C is one or more of )_"_'_]_, and D is not a space, tab, carriage
return, or line feed.
(In cases like A_B_._)_ _, the spaces are inserted after the )_.)
If you don't like the special treatment of some of these
characters, turn one off with:
._P_E_R_I_O_D_ _n_ _c_h_a_r_a_c_t_e_r_ -- which turns off special treatment of c_h_a_r_a_c_t_e_r_ if
n=0, otherwise, turns it on.
._P_E_R_I_O_D_ with no argument, still turns off special treatment of all
of ._!_?_;_:_.
M_E_M_O_ is now the default output name instead of (MEMO).
C_o_h_e_n_'_s_ _L_a_w_:
P=1/C2, i.e.,
the probablity of my looking at any bug complaint, without an example of
the source file which caused it, is inversely proportional to the speed of
light squared.
O_T_H_E_R_ _N_E_W_ _F_E_A_T_U_R_E_S_:_
In the immediate future, I shall concentrate on continuing to straighten
out the inside of TJ6, and I am unwilling to put in new features. Please
don't waste my time by asking for a .UPSIDEDOWN command, or whatever's your
penchant. However, if you feel you absolutely must have a new feature, you
may go about it as follows:
1) Obtain three "TJ6 Suggested Improvement Forms"
from Suzin Jabari. (These will be available in a couple of weeks.)
2) Fill them out, and give one copy to Marvin, one copy to Berthold, and
one copy to me.
Your request will be carefully filed.
--JDC
NEW TJ6 COMMANDS
RWW 2/27/74
CHANGED COMMANDS:
.PHP1 New .PHP1 takes arg. 0 means dont print header on page
one; not 0 means print it. Default is 0. .DPHP1 is gone.
NEW COMMANDS:
.CHAP (argument is chapter number).
Once this command is executed, all pages will be numbered as
xx.yy, where xx is the chapter number and yy is the page
number. For example, ".chap 4" will cause the pages to be
numbered 4.1, 4.2, etc. Note that chap also causes the page
number to be reset to one. This means that the first page
of a chapter is not numbered (since it should probably have
a heading).
.HE1 These two set the left (HE1) and right half of the
.HE2 header line to the fllowing text line. The page # is
put between them.
.HDFONT Causes font <arg> to be used for all headers.
Font 0 is the default.
Another in-text command character:
~Fd Select font d where d is digit immediately
following the ~f. d can be 0-3.
RWW 2/8/74
.SPAGE Set page number to arg.
.PHP1 causes header/page number to be printed on page 1.
( .DPHP1 Dont print header on page one.)
.RLINE Decrease the line length by arg.
.SELECT .SF Generate XGPLPT commands to select font arg.
.PSCORE Sets position of underline. Arg = 0 will put it
at base line of font; arg + will move it down.
.BSCORE Begin under score. (use at own risk.)
.ESCORE End underscore. " (Takes arg as in .PSCORE.)
.BLADJUST Send xgp commands that cause text to move up or down
by arg. (+ is up) Sticks until next
.SELECT.
.BLUP SET BASE UP AMOUNT. (+ IS UP) INITIALLY 12
.BLDOWN SET BASE DOWN AMOUNT. (+ IS DOWN) "
IF ONE OF THE FOLLOWING CHARACTERS IS ENCOUNTERED IN THE
TEXT, IT HAS AN IMMEDIATE EFFECT.
CHAR EFFECT
^A CAUSE BASELINE TO BE MOVED UP BY "BASE UP AMOUNT".
^K CAUSE BASELINE TO BE MOVED DOWN BY "BASE DOWN AMOUNT".
Ntj6 now will produce files that will accomodate underlined text on the XGP.
Using the command ".xgp" will enable this format output,
as will typing "X" somewhere in the command line. In addition, /x
will also force the page specifications so that the resulting output
will fit in the default XGP page size.
Ntj6 will also read commands from the jcl command line in ddt.
I.E. typing ":ntj6 filenam1 filenam2" will produce "filename1 (memo)".
THE NTJ6 CURRENTLY ON SYS: HAS TWO NEW COMMANDS...
.FNAME INSERTS NAME OF THE FILE BEING INPUTTED FROM,IN THE
FORMAT 'DEV:SNAME;FN1 FN2'
.DATE INSERTS CURRENT DATE IN FORMAT 'MM/DD/YY'
BOTH ARE PRELIMINARY AND NEITHER IS CLAIMED TO CERTIFIED
BUG-FREE....
'!<CR>' NOW MAKES TJ6 TOPLEVEL AND FLUSHES HACTRN, AS IT WAS
SUPPOSED TO. 'Q<CR>' LOGS YOU OUT, ^_^Z REPLACES TJ6 WITH HACTRN
AGAIN.
THIS INFO DESCRIBES ADDITIONS TO TJ6 MADE
AFTER THE TJ6 MEMO (164A) WAS PUBLISHED.
THE ALGORITHM FOR DECIDING WHEN A PERIOD
SHOULD BE USED FOR JUSTIFICATION HACKS
(I.E. HAVE TWO SPACES INSERTED AFTER IT
ETC.) IS NOW A BIT MORE REASONABLE.
COMMANDS CAN NOW BE ENTERED AT THE TOP
OF EACH PAGE (WHEN TJ6 HALTS TO ACCEPT
A CHARACTER FROM YOU), BY TYPING LF
INSTEAD OF SPACE. IT IS NOT YET KNOWN IF
THIS FEATURE REALLY WINS.
.SP AND .SPACE NOW HAVE IDENTICAL ACTIONS
(I.E. BOTH TRY TO BE CLEVER ABOUT NOT
GENERATING BLANK LINES AT THE TOP OF A
PAGE ETC.)
.LS TURNS OFF THE HEURISTICS FOR .SP AND
.SPACE.
.STPCHR DEFINES A CHARACTER WHICH WILL
NOT BE PRINTED ON OUTPUT, BUT WILL CAUSE
TJ6 TO WAIT UNTIL YOU TYPE A SPACE (SO
YOU CAN CHANGE GOLF-BALLS FOR INSTANCE).
.HYPCHR DEFINES A CHARACTER WHICH WILL
NOT NORMALLY BE PRINTED ON OUTPUT UNLESS
IT GETS USED TO BREAK A WORD, IN WHICH
CASE IT IS REPLACED BY "-".
(HIDDEN-HYPHEN FEATURE).
.QTECHR DEFINES A CHARACTER WHICH WILL
QUOTE THE FOLLOWING CHARACTER (SO IT
WILL LOSE ITS SPECIAL SIGNIFICANCE).
.DUMCHR DEFINES A CHARACTER WHICH WILL
BE OUTPUT AS SPACE, BUT BE TREATED
INTERNALLY AS A LETTER.
.SPREAD CONSIDERS THE NEXT LINE AS
COMPOSED OF 3 STRINGS, THE FIRST OF
WHICH IS TO BE LEFT-ADJUSTED, THE SECOND
CENTERED AND THE THIRD RIGHT-ADJUSTED
(ALL IN THE SAME LINE).
THE STRINGS ARE SEPARATED BY
A SEPARATOR DEFINED BY THE FIRST
CHARACTER ON THAT LINE.
SOME OR ALL OF THE STRINGS MAY BE NULL.
.INSRT <FILE-NAME> WILL INSERT THE FILE
<FILE-NAME>. INSERTIONS MAY NOT BE
NESTED MORE THAN 8 DEEP.
.APPND <FILE-NAME> WILL APPEND THE FILE
<FILE-NAME> (AND IGNORE THE REST OF
THE CURRENT FILE).
.OF INCREASES THE INDENTING BY <ARG>
AFTER THE NEXT LINE IS OUTPUT.
THAT IS
.INDENT N
.OFFSET M
IS EQUIVALENT TO
.INDENT N+M
.UNDENT M
.PV EVEN PAGE FORCE (NOT YET REALLY DEBUGGED)
.PD ODD PAGE FORCE (NOT YET REALLY DEBUGGED)
.NVRADJ MAKES ALL .ADJUSTS LOOK LIKE
.FILL S ONLY.
.STOP WAITS FOR CHARACTER TO BE TYPED
BEFORE PROCEEDING.


2
doc/_info_/tj6.order Executable file
View File

@ -0,0 +1,2 @@
A description of TJ6 is available as AI Memo #358

6
doc/_klfe_/-read-.-this- Normal file
View File

@ -0,0 +1,6 @@
THIS DIRECTORY CONTAINS THE KLDCP "FRONT END" FILE SYSTEM.
THIS INCLUDES VITAL FILES SUCH AS MICROCODE AND DIAGNOSTICS.
FILES IN THIS DIRECTORY ARE KNOWN TO THE PDP-11 BY ABSOLUTE
DISK ADDRESS AND SHOULD NOT BE MESSED WITH IN ANY WAY.
USE ONLY THE :KLFEDR PROGRAM TO MANIPULATE THESE FILES.

113
doc/_teco_/-read-.-this- Normal file
View File

@ -0,0 +1,113 @@
THIS DIRECTORY STORES COMMON TECO MACROS. DO NOT DELETE.
IF YOU WISH TO PERMANENTLY STORE A TECO-RELATED FILE(S) ON THIS
DIRECTORY, THEN ENTER THE FILE NAME(S) BELOW IN THE
INDICATED FORMAT. FILES STORED ON .TECO. AND NOT DOCUMENTED
ARE NOT GUARANTEED TO SURVIVE.
DO NOT ALTER ANY FILES ON THIS DIRECTORY WITHOUT FIRST
RECEIVING PERMISSION FROM ONE OF THE PEOPLE MAINTAINING THE FILE.
THESE FILES ARE AVAILABLE FOR PUBLIC USE, BUT USE
AT YOUR OWN RISK.
A SHORT DESCRIPTION OF EACH FILE FOLLOWS:
.DDT_ (INIT) ;(JLK) DDT LOGOUT FILE INCLUDING FINIT (Q.V.)
;(JLK) MANY USERS LINKED TO THIS
.TECO. (INIT) ;(JLK) " " " " "
;PRATT'S DOC PACKAGE
DOC > ;(PRATT) SOURCE FILE FOR THE DOC EDITOR
DOC LOAD ;(PRATT) COMPILED DOC. LOADED BY EJ.TECO.;DOC LOAD
DOC TTY ;(PRATT) COMPILED DOC TUNED TO TTY USERS NEEDS
DOCORD > ;(PRATT) DOC MANUAL
ERROR INFO ;(JLK) INSTRUCTIONS ON HOW TO DEBUG A TECO ERROR.
FINIT 7 ;(JLK) SOURCE FOR BACKUP FILE ZEROING PROGRAM
FINIT BIN ;(JLK) BIN FOR "
FNTLDR 31 ;(JLK) SOURCE FOR FONT LOADER JOB DEVICE
FNTLDR BIN ;(JLK) BIN "
FTN^R > ;(JLK) FORTRAN ^R MACROS
FTN^R MACROS ;(JLK) " "
LISPT 65 ;(JLK) LISP CODE FOR RUNNING TECO INFERIOR TO LISP
LISPT FASL ;(JLK) FASL """
LISPT LOADER ;(JLK) TECO MACROS FOR LOADING STUFF TO RUN AS LISPT
MACLST > ;THIS MACRO CONVERTS A FILE INTO SOMETHING PRINTABLE
ON MEMOWRECK OR LPT. GIVEN A NUMERIC ARG <= 1, WILL
PRODUCE MEMOWRECK FORM; OTHERWISE, LPT FORM. STRING
ARG SPECIFIES INPUT FILE FOR WHICH OUTPUT WILL GO TO
FN2 OF "@". IF EMPTY STRING ARG (E.G. ^R CALL), THIS
MACRO WILL USE CURRENT DEFAULT FILE SPEC.
;RBR MACRO PACKAGE
MACROS <TECO VERSION>DMP ;(RBR) LOADABLE BY EJ COMMAND
MACROS DUMPER ;(RBR) CREATES MACROS <TECO VERSION>DMP
MACROS > ;(RBR) SOURCE FILE -- CONTAINS ^R MACROS AS WELL
MACROS (INIT) ;LOADER FOR MACROS. USER SHOULD LINK HIS
.TECO. (INIT) TO THIS FILE.
MACROS LOAD ;(RBR) SEMI-COMPILED MACROS TO BE LOADED INTO QREGS
LOADED BY .TECO.; MACROS (INIT).
MACROS DOC ;(RBR) DOCUMENTATION FOR SAME
MACROS DOC^R ;(IRA, RBR) DOCUMENTATION FOR CONTROL CHARS IN ^R MODE
;INTENDED FOR PASTING ON CONSOLE.
MACROS COMPLR ;(RBR) "COMPILER" INPUT: MACROS >
OUTPUT: MACROS DOC AND MACROS LOAD
MACROS EXTRA ;(RBR) MISCELLANEOUS AND OBSOLETE MACROS
MACROS SUGGS ;SUGGESTIONS FOR IMPROVEMENTS AND CHANGES.
MGRIND MACRO ;MACSYMA BATCH FILE GRINDER
;RVB'S MODE MACRO PACKAGE
MOD477 > ;(RVB) ?
MODE DOC ;(RVB) DOC FOR MODE MACRO PACKAGE.
MODEDF > ;(RVB) ? USED BY LOADER. ?
MODELD > ;(RVB) LOADER FOR MODE MACROS. (LINK FOR AUTO LOAD).
MODEMR > ;(RVB) EXAMPLES OF MACROS
MODE^R > ;(RVB) EXAMPLES OF MACROS
MYMODE > ;(?) EXTRA MACROS LOADABLE INTO MODE ENVIRONMENT.
NEWT 12 ;(JLK) MACRO TO DUMP MACROS FOR A NEW TECO VERSION
TECMAC RECENT ;(JLK) "RECENT" DOCUMENTATION OF MACROS
TECMAC ; FILES USED BY TECMAC MACRO PACKAGE (JLK@MC)
TECMQR ; Q-REGISTER TYPE MACROS FOR TECMAC
TECM^R ; ^R TYPE MACROS FOR TECMAC
TCMnnn ; DUMPED TECMAC MACROS FOR TECO VERSION nnn
TECMLD ; LOADER MACRO FOR TECMAC FILES
TXJ ; XGP TEXT JUSTIFIER FILES.txj is like tj6 except
it has primitives for equations and
is smart about vertical heights. in addition it
interfaces well with teco allowing arbitrary teco
macros to be executed during justification. thus
abbreviations, special hacks, etc. are easy to
implement.
TXJTST ; TXJ TEST FILE. TO SEE FEATURES, XGP TXJTST TXJOUT
ZUSREP MACRO ;(PJ) USER REPORTER - TYPE 1MZ OR MZ<MONTH>. MOVED
TO PJ; DIRECTORY, 1/2/76.
[MACS] > ;(MOON, EAK, CBF, ECC) Source e file for TMACS
[MACS] (PURE) ;(MOON, EAK, CBF, ECC) Compiled TMACS file
;Not very well documented (if at all)
[MMAC] > ;(MOON, EAK, CBF, ECC) Source for TMACS' M macros
[MMAC] (PURE) ;(MOON, EAK, CBF, ECC) Compiled M macros.
;These macros are self documenting.
;Note: "TMACS" is a mailer equivlance entry
;for all users of the above macros.
_TECM TODO ;(JLK) TODO LIST FOR TECO MACROS.
TXJ MACROS ;(JLK) TEXT JUSTIFIER MACROS
TXJ Q-RDOC ;(JLK) DOC FOR SAME
_@ GENDOC ;COMMANDS TO @ FOR DOCUMENTING VARIOUS PROGRAMS.
_@ ITSDOC ;GENERATES @ DOC OF ITS.
_@ LSPDOC ;GENERATES @ DOCUMENTATION OF LISP.
_@ TECDOC ;GENERATES @ LISTING OF VARIOUS TECO RELATED FILES.
_TXJ FONTS ;FONT SPECS USED BY TXJ
_TXJFL MACRO ;(JLK) FONT FILE LOADER USED BY TXJ

BIN
doc/moon/klddt.order Normal file
View File

Binary file not shown.

3
doc/scheme/howto.rabbit Normal file
View File

@ -0,0 +1,3 @@
(LOAD '|AI:SCHEME;RABBIT DOFASL|)
(INIT-RABBIT)
(COMFILE <FILENAME>) ; WILL MAKE A FOO LISP FILE THAT YOU CAN NCOMPLR

114
doc/wp/dk-10.12h7 Normal file
View File

@ -0,0 +1,114 @@
.he DK-10 Programming Page
.l 62
.comment harware memo 12
.paperw 62
.bspdir
.ds
.center
DK-10 Programming
.sp
1. Introduction
.sp
The Systems Concepts DK-10 terminal controller provides
direct from memory output and buffered input of characters
at program set speeds from a DEC PDP-10 system. The following
memo describes the programming of the DK-10 and some
aspects of its operation.
.sp
2.1 Input
.sp
The DK-10 has provisions for setting the speed of each input line
and for enabling and disabling each input line (see section 3.4). When
a character is received on an enabled input line it is placed in a
16 slot first-in-first-out stack along with the line number on
which it was received. A character received when this stack is full
will be lost.
As long as there is at least one character in the input
stack, the input done flag will be on in the DK-10`s control
register and it will request an interrupt if its PIA is non-zero.
Character and line number pairs may be read in from the stack
by doing a DATAI. The entire input stack can be cleared by doing a
CONO to clear the input done flag.
.sp
2.2 Output
.sp
The DK-10 performs character output directly from memory and
handles enough bookkeeping that, for may applications, no output
interrupts need be taken. The DK-10 has a base address register (see
section 3.2) that is used to access a list of word pairs
for the output lines. The first word of each pair, at the
base address plus twice the line number, has a count of
characters remaining to be output on that line. The second word,
at the base address plus one plus twice the line number, is a pointer
to the buffer of characters to be output. This pointer
is similar to a regular byte pointer with its halves swapped and is
effectively ILDBed by the DK-10 to get characters. The size field is
always treated and stored back as if it were 7. The postion
field should be of the form 36-7k.
.figure 5
When the count word is counted down to -1 for an output line,
the output done flag is set for this line. If output interrupts
have been enabled, the DK-10 will then request an interrupt and
it will be possible to read in the number of the line that stopped. The
DK-10 has provisions for starting and stopping each output line and for
setting the speed of each output line be doing CONOs.
To make it easy to ring buffer output, the DK-10 actually examines
each word that it fetches from memory for the purpose of extracting a character
from it. If the bottom bit is one (this bit is not used in packed 7 bit fields)
then the word is treated as a new "swapped byte pointer" which is
effectively ILDBed to get the character to output and then stored
back on top of the second word of the word pair associated with this
output line. Thus one can store a byte pointer back to the start
of a buffer at the end of the buffer.
If the DK-10 gets a non-existant memory indication on a read or
write or a parity error on a read it sets an error
flag in its control register and stops all output processing until
reset.
.sp
3.1 DATAI
.sp
.figure 2
14-17 input line on which character was received
28-35 charcter code
.sp
3.2 DATAO
.sp
.figure 2
15 if a one, enables output interrupts
16-35 base address for output control words
.sp
3.3 CONI
.sp
.figure 2
14-17 number of an output line with done flag set
27 in on, indicates some output line done
28 state of selected input line
29 selected output line busy
30 parity error
31 non-existant memory error
32 input done flag
33-35 priority interrupt channel assignment
.sp
.block 4
3.4 CONO
.sp
.figure 2
.sp
18 master clear
19 makes function apply to selected line through highest
20-23 line number
24-26 speed if used by function
27-29 function as follows:
0 set output done
1 start output
2 disable input
3 enable input
4 stop output & clear output done
5 set output speed & start output
6 set input speed & disable input
7 set input speed & enable input
30 clear parity error
31 clear non-existant memory error
32 clear input done flag
33-35 priority interrupt channel assignment