Trailing-Edge
-
PDP-10 Archives
-
APLSF-20_V2_bin_10-12-79
-
aplsf/doc/aplsf.doc
There are 3 other files named aplsf.doc in the archive. Click here to see a list.
APLSF.DOC -- Changes from V1C(315) to V2(435)
September 1979
COPYRIGHT (C) 1976,1979 BY
DIGITAL EQUIPMENT CORPORATION, MAYNARD, MASS.
THIS SOFTWARE IS FURNISHED UNDER A LICENSE AND MAY BE USED AND COPIED
ONLY IN ACCORDANCE WITH THE TERMS OF SUCH LICENSE AND WITH THE
INCLUSION OF THE ABOVE COPYRIGHT NOTICE. THIS SOFTWARE OR ANY OTHER
COPIES THEREOF MAY NOT BE PROVIDED OR OTHERWISE MADE AVAILABLE TO ANY
OTHER PERSON. NO TITLE TO AND OWNERSHIP OF THE SOFTWARE IS HEREBY
TRANSFERRED.
THE INFORMATION IN THIS SOFTWARE IS SUBJECT TO CHANGE WITHOUT NOTICE
AND SHOULD NOT BE CONSTRUED AS A COMMITMENT BY DIGITAL EQUIPMENT
CORPORATION.
DIGITAL ASSUMES NO RESPONSIBILITY FOR THE USE OR RELIABILITY OF ITS
SOFTWARE ON EQUIPMENT WHICH IS NOT SUPPLIED BY DIGITAL.
APLSF.DOC Page 2
APLSF.DOC -- Changes from V1C(315) to V2(435)
September 1979
1.0 SUMMARY
APLSF is the implementation on the DECSYSTEM-20 of the APL language
with shared system functions and system variables. It runs on a KL-20
processor under release 3A or later of the TOPS-20 monitor. It
requires the compatibility package, PA1050, during execution, and is
documented in the APLSF Programmer's Reference Manual
(DEC-20-LASFA-A-D). Version 2 is documented in the TOPS-10/20 APLSF
Language Manual (AA-H200A-TK).
Version 2 runs under release 3A or later of the TOPS-20 monitor. It
has been tested in-house under 3A and 4.
The reason for this release is to correct outstanding bugs in APLSF
Version 1C and to incorporate several functional enhancements as
described below.
APLSF Version 2 fully supersedes Version 1C and includes edits #316 to
#435.
APLSF.DOC Page 3
2.0 EXTERNAL CHANGES
1. The current setting of )MAXCOR is now saved across a )CALL,
so it is possible to return from a )CALL even if )MAXCOR had
been expanded in the workspace before the )CALL.
2. There are now TTY mnemonics for diamond (.DM), left bracket
(.LB), left tack (.LK), right bracket (.RB), and right tack
(.RK).
3. Nonprinting elements of the atomic vector (.BXAV) are now
displayed as squish quad (.SQ). Note that this is not a
valid character and cannot be typed to APL.
4. When doing APL output to an ASCII sequential file while in
TTY mode, the default terminal type is now LA36 instead of
2741 (which is not supported by TOPS-20).
5. Error numbers are now catenated to the front of error
messages. Also, the error line printed when an error occurs
occasionally has fewer added spaces and the caret position is
sometimes one position to the right of where it used to be.
6. LIB: now refers to whatever the user has logical name LIB:
defined to be, instead of the APL system library. LIBn:
still refers to the system library.
7. It is now possible to open a PTY as an ASCII sequential file.
As noted in the next item, lower case input will be converted
to upper case just as with any ASCII sequential file. The
new system variable .BXTIMEOUT will be set if the input was
not terminated with a line feed.
8. When doing TTY mode input from ASCII sequential files, lower
case characters are converted to uppercase (instead of random
operators).
9. The omega function (.OM) has been removed from the version 2
manual. Omega still works in version 2 but will not be
supported after version 2. The .OM symbol has other proposed
uses in the APL language and the optimizing effect of .OM
will be implememted for reduce-iota-rho.
10. The definitions of I-beams (.IB) have been moved to an
appendix of the version 2 manual. They are still supported
in version 2 and most will still be supported in later
versions for compatibility with other vendors. However, we
strongly recommend that system variables and functions be
used instead of I-beams.
11. Quote-quad (.QQ) or quad-del (.QD) input following .QQ or .QD
output precedes the input string with a string of blanks
equal in length to the preceding output. If such preceding
output was longer than 390 characters (the maximum input
APLSF.DOC Page 4
length), a LINE TOO LONG error occurs and correct input is
awaited.
12. For niladic function F which returns no value, the
expressions "1;F", "1+F" and "F+1" give value errors since a
value is needed by the primitive function but F did not
return one. The expression "F;1" returns 1 to be consistent
with the expression "F", which executes F but prints nothing.
Note that the ";" function is a monadic output catenator,
operating on the expression to its right. ";" is not
intended to be a statement separator.
13. On the -20, GENSF can now set the device for the library area
and the )HI file as well as the directory. The default is
now PS: instead of DSK: so APLSF will be able to find these
areas even if you are not connected to PS:. You must be
connected to the device specified for these areas when
running GENSF in order for the generation procedure to be
successful. Notice that this device is also used for the
billing file.
2.1 Functional Changes
1. .BXMTP has been modified to simulate some of the TAPOPs not
in PA1050. It still has problems if the file has not been
opened. Specifically, density and mode cannot be set from
inside APL before the file is opened. Use the SET TAPE Exec
command to set these tape parameters.
2.2 Functional Additions
2.2.1 Error Trapping - The following system variables and functions
have been added to allow program handling of error conditions and
other events.
1. .BXBREAK - This is a monadic function which causes execution
to terminate, suspending the current function and entering
immediate mode. It prints its argument before breaking to
the terminal.
2. .BXERROR - This is a variable which is set by APL to the
error message for the last error. (It contains the text
which is normally printed on the terminal.)
3. .BXSIGNAL - This is a dyadic function which signals an error
from the current function. It causes the current function to
terminate before generating the error. Its right argument is
either a valid APL error or a user defined error in the range
APLSF.DOC Page 5
500-999. Its optional left argument is the message to be
associated with the error.
4. .BXTRAP - This variable is similar to .BXLX. It can be set
to an expression which will be executed when an error occurs.
2.2.2 Limiting Input Wait - The following two variables allow a
program to limit the amount of time a user is given to respond to
quote-quad or quad-del input.
1. .BXTIMELIMIT - The number of milliseconds to wait for user
input. If 0, then no limit. If -1, no waiting is done (only
characters which were typed ahead are read).
2. .BXTIMEOUT - Flag indicating whether or not user timed out on
last input. (This flag is also used with PTYs to indicate
whether or not last input was terminated with a LF).
2.2.3 New Character Arrays - The following system variables are
useful subsets of .BXAV.
1. .BXALPHA - The 27 letters, .LD-Z.
2. .BXALPHAU - The 27 underscored letters, [email protected].
3. .BXASCII - The 128 ASCII codes. These codes are sent to
files as is, regardless of the terminal type. They are
useful when doing graphics.
4. .BXCTRL - The 32 control characters, octal 0-37.
5. .BXNUM - The 10 digits, 0-9.
2.2.4 New File Functions -
1. .BXRENAME - This is a dyadic system function which allows a
user to rename a file from within APL. Its right argument is
the channel associated with the file to be renamed. Its left
argument is the new filespec.
2. .BXAPPEND - This is a dyadic system function which allows a
user to append data to /BU or /DA files. Its left argument
is the data item to be appended to the file. Its right
argument has the same format as the right argument to an
output quad (.OQ) and specifies how the data is to be
formatted in the file.
APLSF.DOC Page 6
3. Two new file types, /AS* and /BS*, have been added to allow
users to append to ASCII and binary sequential files. These
file types are analogous to /IS*.
APLSF.DOC Page 7
3.0 KNOWN BUGS AND DEFICIENCIES
3.1 Deficiencies And Restrictions
The known deficiencies and restrictions in APLSF are:
1. The caret symbol indicating where an error in an APL
statement was first detected is sometimes displayed in the
wrong position. With the introduction of error trapping and
.BXERROR, it is desirable to fix this. We will attempt to do
so in the future. With this in mind, we recommend that users
do not write programs which depend on the current positioning
of the caret.
2. It is not possible to display the trace or stop vector of a
function. This is a restriction which will not be removed in
the foreseeable future.
3. It is impossible to enter an invalid TTY mnemonic in a
character string when using a TTY terminal, just as it is
impossible to enter an illegal overstrike from an APL
terminal. This is a permanent restriction.
4. Monadic domino uses an inordinate amount of core. This is a
deficiency due to the algorithm used, which does monadic
domino by simulating a dyadic domino with a left argument of
the identity matrix. Thus, it must construct a double
precision identity matrix.
3.2 Outstanding Bugs
The known outstanding bugs in APLSF are:
1. Erasing a pendent or a suspended function can cause a SYSTEM
ERROR, depending on the state of the function when erased.
2. The compression of a zero-by-n matrix returns a length error
when the left argument is an n-element vector, but works if
it is a scalar.
3. A niladic branch inside a multistatement line will fail with
a SYNTAX ERROR.
4. Linefeeds typed during linefeed editing may appear as
carriage return linefeed pairs. This is a temporary
restriction which will be corrected in version 4 of the
compatibility package. A patch to PA1050 which solves this
problem is included in the beware file on this tape
(APLSF.BWR).
APLSF.DOC Page 8
5. The )LIB command updates the access dates of the listed
files. In particular, )LIB/A updates the access dates to the
current date before printing them. This is a problem with
the compatibility package.
APLSF.DOC Page 9
4.0 INSTALLATION INSTRUCTIONS
4.1 Files Included on This Tape
There are four savesets on this tape: DOCUMENTATION, BINARIES, WSINT
and REL LIBRARIES.
The contents of DOCUMENTATION are:
APLSF.DOC This documentation file.
APLSF.BWR The beware file, containing DDT patches.
APLSF.HLP The HELP message to respond to the monitor HELP
command.
APLSF.RNH The RUNOFF input for the HELP file.
APLSF.RND The RUNOFF input for the .DOC file.
The contents of BINARIES are:
APLSF.EXE The APLSF executable file.
GENSF.EXE Generation program for APLSF. See below for its
use.
BILING.1 The APLSF workspace which processes the APL billing
file.
The contents of WSINT are:
APLCNV.DOC DOC file for the interchange package.
APLTAP.EXE Program to read/write interchange tapes.
WSIN.APL Workspace for conversion to APLSF.
WSOUT.APL Workspace for conversion from APLSF.
The contents of REL LIBRARIES are:
APLS22.REL REL Library for APLSF-20 v2.
APLS22.LNK LINK indirect file for APLSF-20 v2.
APLG22.REL REL Library for GENSF-20 v2.
APLG22.LNK LINK indirect file for GENSF-20 v2.
APLW22.REL REL Library for APLSF-20 v2 WSINT.
APLW22.LNK LINK indirect file for APLSF-20 v2 WSINT.
The following two files are not distributed. They are generated by
APLSF.DOC Page 10
APLSF.
APLHSF.APL The APLSF HI file, if a HI message is desired.
ACCTSF.APL The APLSF accounting file.
The default is to put all files on the subsystem area PS:<SUBSYS>, and
to make the APL library area be PS:<SUBSYS>. There is no need to put
the files from the REL LIBRARIES saveset on disk.
4.2 Changing the Default Installation Parameters
The default installation parameters are as follows:
The installation name is DECSYSTEM-20 APLSF.
The automatic save facility, .BXAUS, is off.
The structure used for the HI file and the library is PS:.
The name of the accounting file is ACCTSF.APL.
The directory used for the accounting file is <SUBSYS>.
The directory used for the HI file and the library is <SUBSYS>.
The default value of )MAXCOR (and thus the default maximum
workspace size) is 40P.
If a change in the parameters is desired, put the .EXE file on device
DSK:, run GENSF.EXE and answer the questions. The program updates the
.EXE file. Then rename and distribute the different files to the
appropriate disk structures and directories.
4.3 Creating a HI Message
If you wish a HI message, login as the operator, sign on APLSF, and
execute the command
)HI 'THIS IS MY HI MESSAGE'
Then sign off APLSF. This generates APLHSF.APL on the structure and
the directory specified for the )HI file by GENSF.
The )HI message is truncated to 239 characters. The error message "WS
NOT FOUND" means that the )HI file directory specified by GENSF does
not exist.
APLSF.DOC Page 11
5.0 INTERNAL CHANGES
5.1 Edits
EDIT # SPR REMARKS
316 27504 Fix CRLF idle-character handling: use correct
idle character for TTYs and reset the idle
character control count properly.
317 27492 Fix input routine for integers of magnitude
greater than 2^27-1.
320 12872 Prevent function line editor from accepting
negative line numbers (with random results).
321 Dev Implement .BXRENAME.
322 Dev Implement .BXAPPEND and /AS* file type.
323 Dev Implement .SQ for all nonprinting characters in
.BXAV.
324 Dev Implement .BXASCII.
325 ----- Modify .BXMTP to simulate some of the TAPOPs not
in PA1050. (-20 only)
326 ----- Correctly handle nonalphanumeric SIXBIT characters
in user-specified installation name.
327 12947 Check allocated space on codestring creation to
avoid overwriting data.
330 ----- Change default PPN to be user's PPN if PPN of
current workspace is [0,0].
331 ----- Change default APL output type for TTYs from 2741
to LA36.
332 Dev Implement .BXTIMELIMIT and .BXTIMEOUT.
333 ----- Have TTY mode input from ASCII sequential files
convert lower to upper case.
334 ----- Internal edit.
335 ----- Internal edit to circumvent TOPS-20 bug. (-20
only)
336 ----- Withdrawn.
337 ----- Correctly print nonalphanumeric SIXBIT characters
in user name.
APLSF.DOC Page 12
340 ----- Internal edit.
341 Dev Implement .BXERROR.
342 ----- Fix LIB: definitions on -20. (LIB: belongs to
user, LIBn: belongs to APL) (-20 only)
343 Dev Implement .BXBREAK.
344 ----- Add TTY mnemonics for diamond (.DM), left bracket
(.LB), left tack (.LK), right bracket (.RB), right
tack (.RK).
345 ----- Internal edit.
346 Dev Don't set .BXERROR on EOF.
347 Dev Append .XQ execute error message to previous
.BXERROR.
350 Dev Implement .BXTRAP.
351 ----- For LA36, shift out after MONITOR: message
instead of before.
352 Dev Implement .BXSIGNAL.
353 ----- When doing )SI, have .XQs show up as .XQ instead
of .EP.
354 Dev Make .BXBREAK work inside .XQ.
355 ----- Internal edit.
356 ----- Fix shadowing of .BXERROR.
357 ----- Internal edit.
360 Dev Allow .BXBREAK in APL-BASIC.
361 ----- Fix SIXBIT to NINEBIT of nonalphanumerics.
362 ----- .BXFLS, .BXENQ and .BXDEQ were returning invalid
error numbers on TOPS20.
363 ----- Make .BXBREAK clear )SI when done from protected
function.
364 Q00429 Prevent loop or system error when generating error
during subscripted catenate.
365 Q00377 Put setting of )MAXCOR into TMPCOR so it is
possible to return from a )CALL even if )MAXCOR
has been expanded.
APLSF.DOC Page 13
366 ----- Correctly save terminal type across )CALL.
367 ----- Internal edit.
370 ----- Correctly restore localized system variables when
)LOADing or )COPYing a workspace saved with a
different version of APL.
371 Q00448 If .BXPW is set to a width less than )TABS, reset
)TABS to new width (as is done with )WIDTH).
372 ----- Prevent workspace damage from occurring when
setting .BXTRAP to a system command.
373 ----- Allow .BXSIGNAL and .BXBREAK in immediate mode.
374 ----- Introduce the idea of system variables which can
be set by the system but not the user (needed for
.BXTIMEOUT and .BXERROR).
375 Q03193 Fix .BXRENAME on channels 2 and 5.
376 Q00431 Fix real to integer conversion of negative
numbers.
377 Dev Implement /BS* (sequential append for binary
files).
400 Dev Allow recursive error trapping. Prevent it only
with errors which occur in the .BXTRAP line
itself, not in any functions which are called.
401 ----- Internal edit.
402 ----- Remove 1030, ONTEL and TYCOM terminal types from
HELP message. They are still supported, but are
not documented.
403 ----- Fix printing of error line when error occurred at
end of execute string.
404 ----- Take CRLF out from before ATTENTION SIGNALED error
message so error number appears on the same line
as message.
405 ----- Internal edit.
406 ----- Fix problem with printing real matrices (e.g. 2
2.RO.FL/.IO0).
407 ----- Make .BXGAG work on TOPS20.
410 ----- Have .NT print @T in escape, and turn on
.LK,.DM,.LB,.RB,.RK output mnemonics.
APLSF.DOC Page 14
411 ----- Add local version numbers.
412 ----- Make product version number available to GENSF and
keep MAKILB from loosing it.
413 ----- Allow inner products with right function = and .NE
to compare alpha's with numerics (i.e. 'ABC'.=1 0
1)
414 QAR561 Fix edit 370 to set ASVSYMTAB on )CLEAR and to
RELWS WS with version = 0.
415 QA3420 Fix edit 367 to contract core on )LOAD: problem
only occurs on the -10 when loading smaller WS
after larger.
416 QA3266 Give LINE TOO LONG error if INBUF would overflow
when putting in leading blanks on .QQ or .QD input
following .QQ or .QD output.
417 QAR560 Prevent INCHAR from discarding any character
except a linefeed following a carriage-return.
420 ----- When the quad-error data entry gets allocated
after a partially-built data entry that gtcore()
was trying to get core for when a wsfull occured,
move the quad-error data entry above the partially
built data entry and update globals as necessary
prevent the workspace from being trashed.
421 Q00546 Correct function-line wraparound during output of
a numeric vector (broken during implementation of
.BXERROR).
422 ----- Prevent system errors during evaluation of the
expression F+1) where F is a niladic function that
does not return a value.
423 Q00511 Correct two more problems with function-line
wrap-around.
424 ----- Make DGETSCI accept double precision -1
(=533400,,0 0,,0) as integer -1.
425 Q00518 Make .BXTIMELIMIT set < -1 a DOMAIN ERROR.
426 QAR547 Fix dyadic .BXSIGNAL to correctly ignore left arg
if right arg is system error number.
427 ----- Be able to load release 4 UDDT, which has pages
below 770.
430 QAR612 Change the wording in GENSF (XLIMCOR) to ask for
the default setting of )MAXCOR, not an upper bound
on memory.
APLSF.DOC Page 15
431 QAR613 Allow -20 to specify LIBn: structure (default is
PS:).
432 ----- Fix the dqren() interface to cmdsyn() to prevent
wsid-filename default, accidentally deleted files,
lost default file extension, and infinite loops.
433 QAR564 INLINE loops when .bxTIMEOUT = 1 on PTY because
FILINP didn't reset TIMOUT (DEVTYP UUO is CALLI 53
octal not decimal).
434 ----- Create PATCH.REL to define PATCH: area to be used
instead of PAT.. (which could get wiped out by an
exapanding WS).
435 QA3587 Fix typos in FS on -10 in /BS*. Make STRTIO leave
BLKNUM and INDEX = 0 so WRTSETUP will do read of
last block which write goes into on append.
APLSF.DOC Page 16
6.0 SUGGESTIONS
6.1 Continuing a detached APL job
In the event of a line drop or system crash, a user's job may become
detached. After reestablishing a connection with the system, the user
should check to see if his/her job is still there. If so, the user
should reattach to the job using the ATTACH command. Then, to restore
the job to its previous state, the user should type CONTINUE. If the
user instead types APLSF, a new version of APL will be loaded and the
job's previous core image will be destroyed, along with the user's
workspace.
[End of APLSF.DOC]