Victor@JOHN-JAMESON.DOCS.UU.SE (Bjorn Victor) (02/02/90)
I just started using it, and I wonder if anybody else is, and if so,
have they fixed any of the very irritating bugs in it?
Some that come to my mind:
* White/Black is swapped hardcoded - you have to run it with
(TV:WHITE-ON-BLACK), which is the default off-the-shelf setting, but not
the one we use here. We have to press <Term>-C essentially every time
we press <System>-X.
* Windows don't seem to cover each other completely. I have an Emacs
window covering the left side of an Xterm window, and the text covered
is "printed through" to the Emacs window as the xterm scrolls up.
* Pressing Clear-Screen doesn't do an XRefresh (or whatever it's
called), but clears the X screen.
* Sometimes the server seems to hang ("Request or Event"). Switching
(Explorer) windows seems to be a work-around.
I'd also like the X screen to cover the mouse documentation window, to
make the screen a little bit larger (17" isn't that much to start with).
At least it should clear the doc in it (right now it says "R2: Bring up
the System Menu", which is clearly not the case).
But I'm glad it exists at all!
--Bjorn Victor Victor@John-Jameson.DoCS.UU.SE
Dept. of Computer Systems or
Uppsala University, SWEDEN Victor%John-Jameson.DoCS.UU.SE@uunet.UU.NETBjorn.Victor@docs.uu.se (Bjorn Victor) (11/22/90)
Is there anybody out there, still? I have started to use the X11 server daily (I am doing it right now), and have fixed the black/white swapping in the server (so that (TV:BLACK-ON-WHITE) does what it says, even in the X11 frame). If anyone wants the patches I'd be happy to send them to you, it's not much. I still have some problems, though, that I was wondering if anybody has solved already. (How many of you are really using the X11 server?) (How many of you are really using Explorers any longer...?) * Has anyone got the code for the documented(!) function X11:COMPILE-X11-FONT (appendix B of "SLE X Window System Programmer's Reference"), which is said to make a font usable by X11M out of a BDF file. The stuff in jwz's archive just makes a screen font, without the necessary X11 font info. * Has anyone got the patches that make X11M work better with R4 twm? I have asked TI Sweden to get me either release 7 (which they have claimed should have been released october 23) or an up-to-date patch tape, but this could take a while. Two bugs that I really hate: 1) Typing in GNU Emacs is much more slow than in e.g. xterm. It feels like typing to a 100-baud machine with typeahead. 2) A GNU Emacs window seems to not completely cover an xterm window. Having the Emacs window cover the right side of the xterm window, and then scroll some obscured text in the xterm window makes the obscured text "print through" the Emacs window. Both are with X11R4 clients, GNU Emacs 18.55. What's the story about TI pulling out, officially? TI Sweden still don't know anything about it. Hack hack, -- Bjorn Victor Bjorn.Victor@DoCS.UU.SE Dept. of Computer Systems or Bjorn.Victor%DoCS.UU.SE@uunet.UU.NET Uppsala University, Sweden "I'd rather hack a Lisp Machine!"
Rice@SUMEX-AIM.STANFORD.EDU (James Rice) (11/27/90)
>> * Has anyone got the code for the documented(!) >> function X11:COMPILE-X11-FONT (appendix B of "SLE X >> Window System Programmer's Reference"), which is said >> to make a font usable by X11M out of a BDF file. The >> stuff in jwz's archive just makes a screen font, >> without the necessary X11 font info. A quick grep revealed that this function is in x11:x11m.utilities;font-compiler.lisp Rice; ;------------------------------------------------------------------------------- ;;; -*- Mode:Common-Lisp; Package:USER; Base:10; Fonts:(CPTFONT CPTFONTB CPTFONTBI) -*- #| This program allows one to read in characters from X11 font files and convert them to Explorer font objects. The font files are located in the directory tilde:/lagoon.u1/X11-beta.1/fonts/bdf/. This program reads the font information in a manner similar to the fc program in tilde:/lagoon.u1/X11-beta.1/fonts/compiler/fc.c. To run the program, call the function read-font, and pass a pathname for the font file. The read-font function returns the symbol whose value is the font object. These symbols are prefixed with `X11-' to distinguish them from Explorer fonts which may have the same name. |# ;;; Change history: ;;; ;;; Date Author Description ;;; ------------------------------------------------------------------------------------- ;;; 12/06/88 LGO Ensure font name is all caps. ;;; 11/14/88 LGO Get font height and baseline from ascent and descent when FontBoundingBox is bad. ;;; 11/09/88 LGO Get font height and baseline from the FontBoundingBox. ;;; These font attributes cannot be derived from ascent and descent. ;;; 11/08/88 LGO Correctly skip characters with negative encoding ;;; 11/08/88 LGO Correctly left-justify font bitmaps ;;; 11/08/88 LGO Gather font properties correctly. ;;; 11/08/88 LGO Adjust FED:FD-SPACE-WIDTH and FED:FD-BLINKER-WIDTH for fixed width fonts ;;; 10/27/88 TWE Changed the inner data collection loop index start-row to never ;;; start below the index of 0. ;;; 9/14/88 LGO Use Y-RESOLUTION instead of explorer pixel-resolution ;;; when computing character sizes. ;;; 9/13/88 LGO Fix READ-A-LINE to work on unix files ;;; 9/13/88 LGO Add compile-x11-font function which dumps x11-font-info ;;; in the same file as the font. ;;; 9/13/88 LGO Set char-width to x-width instead of bounding-box-width ;;; 9/13/88 LGO Pad the top of char bitmaps to make the baselines line up. ;;; The prefix for font pathnames and font objects is the same. (DEFCONSTANT X-FONT-OBJECT-PREFIX "X11-") (DEFCONSTANT GOLDEN-RATIO (/ (1+ (SQRT 5)) 2)) (DEFCONSTANT MAX-ENCODING (EXPT 2 16) "Largest number of characters in a font that can be supported") (DEFCONSTANT VERSION-STRING "STARTFONT 2.1") (DEFCONSTANT X-FONT-OBJECT-PREFIX "X11-") (DEFPARAMETER LINE-NUMBER 0 "Number of the line read from a font file. Used for error reporting.") ;;; Simpler version of the read-from-string function, which more easily handles ;;; keywords. This version is less general than the read-from-string function, ;;; but sufices here. (DEFUN READ-FROM-STRING-FIX (STRING &KEY (START 0)) (READ-FROM-STRING STRING T NIL :START START)) ;;; The following defparameters are located here instead of being inside of a LET in read-font ;;; because release 2.1 doesn't allow one to put a large number of locals inside of a function. ;;; Data line read from the font file. (DEFPARAMETER LINE NIL) ;;; Determines the name of the font. This may be a filename, which means ;;; that the font name needs to be extracted. (DEFPARAMETER FAMILY NIL) ;;; Resolution of the font in the X direction. (DEFPARAMETER X-RESOLUTION NIL) ;;; Resolution of the font in the Y direction. (DEFPARAMETER Y-RESOLUTION NIL) ;;; Symbol which is the font's name. Has the X11- prefix. (DEFPARAMETER FONT-SYMBOL NIL) ;;; The number of glyphs in this font. (DEFPARAMETER NUMBER-CHARACTERS NIL) ;;; The lowest character in the font. (DEFPARAMETER LOWEST-ENCODING MOST-POSITIVE-FIXNUM) ;;; The highest character in the font. (DEFPARAMETER LARGEST-ENCODING 0) ;;; The next 4 deal with 2-D fonts. (DEFPARAMETER FI.FIRST-ROW MOST-POSITIVE-FIXNUM) (DEFPARAMETER FI.LAST-ROW 0) (DEFPARAMETER FI.CH-FIRST MOST-POSITIVE-FIXNUM) (DEFPARAMETER FI.CH-LAST 0) (DEFPARAMETER REAL-PROPERTY-COUNTER 0 "Number of properties for this font.") (DEFPARAMETER PROPERTIES NIL) ;;; Total font height (defparameter font-height nil) ;;; Font baseline (defparameter font-baseline nil) (defparameter font-y nil) ;;; Maximum length of ascenders for the font. (DEFPARAMETER FONT-ASCENT NIL) ;;; Maximum length of descenders for the font. (DEFPARAMETER FONT-DESCENT NIL) ;;; Maximums and minimums for components of x-char-info. (DEFPARAMETER MAX-LEFT-SIDE-BEARING NIL) (DEFPARAMETER MAX-RIGHT-SIDE-BEARING NIL) (DEFPARAMETER MAX-CHARACTER-WIDTH NIL) (DEFPARAMETER MAX-FONT-ASCENT NIL) (DEFPARAMETER MAX-FONT-DESCENT NIL) (DEFPARAMETER MIN-LEFT-SIDE-BEARING NIL) (DEFPARAMETER MIN-RIGHT-SIDE-BEARING NIL) (DEFPARAMETER MIN-CHARACTER-WIDTH NIL) (DEFPARAMETER MIN-FONT-ASCENT NIL) (DEFPARAMETER MIN-FONT-DESCENT NIL) ;;; Default character property for the font. (DEFPARAMETER DEFAULT-CHARACTER NIL) (DEFPARAMETER ALL-CHARS-EXIST NIL) (DEFPARAMETER CONSTANT-METRICS NIL) ;;; Point size for the font. (DEFPARAMETER POINT-SIZE NIL) ;;; Family name property of the font. Not used for anything yet. (DEFPARAMETER FAMILY-NAME NIL) ;;; Resolution property of the font. I'm not sure how this differs from ;;; the values of x-resolution and y-resolution, which are supposed to ;;; have the same value (i.e. X resolution = Y resolution). (DEFPARAMETER RESOLUTION NIL) ;;; Height of the font. (DEFPARAMETER X-HEIGHT NIL) ;;; Weight property of the font. (DEFPARAMETER WEIGHT NIL) ;;; Quad width property of the font. (DEFPARAMETER QUAD-WIDTH NIL) ;;; The FED font descriptor (defparameter font-desc nil) (DEFUN READ-FONT (PATHNAME &AUX (x11-metrics (make-array 256))) "Read an X11 bdf font file and generate Explorer font object. PATHNAME is the pathname where the .bdf information exists. X11 font descriptor informations is put on the :X11-FONT-INFO roperty of the font name symbol." (WITH-OPEN-FILE (FONT-STREAM PATHNAME :DIRECTION :INPUT) (SETQ ALL-CHARS-EXIST T X-HEIGHT NIL FONT-HEIGHT NIL FONT-ASCENT NIL FONT-DESCENT NIL LARGEST-ENCODING 0;;; Maximum length of ascenders for the font. LINE-NUMBER 0 LOWEST-ENCODING MOST-POSITIVE-FIXNUM MAX-CHARACTER-WIDTH 0 MAX-FONT-ASCENT 0 MAX-FONT-DESCENT 0 MAX-LEFT-SIDE-BEARING 0 MAX-RIGHT-SIDE-BEARING 0 MIN-CHARACTER-WIDTH MOST-POSITIVE-FIXNUM MIN-FONT-ASCENT MOST-POSITIVE-FIXNUM MIN-FONT-DESCENT MOST-POSITIVE-FIXNUM MIN-LEFT-SIDE-BEARING MOST-POSITIVE-FIXNUM MIN-RIGHT-SIDE-BEARING MOST-POSITIVE-FIXNUM PROPERTIES NIL) ;; The validate-string function compares the first few characters of LINE ;; to those characters in STRING. If they are not the same (case doesn't count) ;; then an error message is generated. (FLET ((VALIDATE-STRING (STRING &OPTIONAL MESSAGE) (IF (NOT (STRING-EQUAL (SETQ LINE (READ-A-LINE FONT-STREAM)) STRING :end1 (LENGTH STRING))) (FERROR NIL "Missing font ~A on line ~D, got ~A instead." (OR MESSAGE (STRING-TRIM " " STRING)) LINE-NUMBER LINE) ;;ELSE (nsubstring LINE (LENGTH STRING))))) ;; Parse the header stuff for the font. (VALIDATE-STRING VERSION-STRING "file, or incorrect version") (SETQ FAMILY (VALIDATE-STRING "FONT " "family line")) ;; Form a font name from the family string. The family can be one of ;; several things: a simple string, a simple filename with an ;; extension, or a fully specified pathname with an optional extension. ;; It is assumed that the characters of the filename component of a ;; pathname contain only alphabetic, numeric, - or _ characters. (SETQ FONT-SYMBOL (INTERN (CONCATENATE 'SIMPLE-STRING "X11-" ;; make the font name the same as the file name. This ;; makes it much easier to deal with fonts, since ;; X uses the file system to locate fonts. (string-upcase (PATHNAME-NAME FONT-STREAM))) 'FONTS)) ;; Read in the point size, x resolution and y resolution. (LET ((RESOLUTIONS-STRING (VALIDATE-STRING "SIZE " "size line"))) (MULTIPLE-VALUE-BIND (VALUE X-RESOLUTION-POSITION) (READ-FROM-STRING-FIX RESOLUTIONS-STRING) (SETQ POINT-SIZE VALUE) (MULTIPLE-VALUE-BIND (VALUE Y-RESOLUTION-POSITION) (READ-FROM-STRING-FIX RESOLUTIONS-STRING :START X-RESOLUTION-POSITION) (SETQ X-RESOLUTION VALUE Y-RESOLUTION (READ-FROM-STRING-FIX RESOLUTIONS-STRING :START Y-RESOLUTION-POSITION)))) (WHEN (NOT (= X-RESOLUTION Y-RESOLUTION)) (FERROR NIL "X and Y resolution must be equal, but are ~D and ~D" X-RESOLUTION Y-RESOLUTION))) (let ((BOUNDING-BOX-STRING (VALIDATE-STRING "FONTBOUNDINGBOX "))) (MULTIPLE-VALUE-BIND (VALUE HEIGHT-POSITION) (READ-FROM-STRING BOUNDING-BOX-STRING) value ;; Ignore font bounding-box width (MULTIPLE-VALUE-BIND (VALUE LEFT-POSITION) (READ-FROM-STRING-FIX BOUNDING-BOX-STRING :START HEIGHT-POSITION) (SETQ font-height VALUE) (MULTIPLE-VALUE-BIND (VALUE BOTTOM-POSITION) (READ-FROM-STRING-FIX BOUNDING-BOX-STRING :START LEFT-POSITION) value ;; Ignore font bounding-box X (SETQ value (READ-FROM-STRING-FIX BOUNDING-BOX-STRING :START BOTTOM-POSITION)) (setq font-y (- value)))))) (SETQ REAL-PROPERTY-COUNTER 0) ;; Read in the properties of the font. (LOOP WITH NUMBER-OF-PROPERTIES = (READ-FROM-STRING (VALIDATE-STRING "STARTPROPERTIES " "properties")) FOR PROPERTY-COUNTER FROM 1 BY 1 WHILE (>= NUMBER-OF-PROPERTIES PROPERTY-COUNTER) FOR LINE = (READ-A-LINE FONT-STREAM) DO (MULTIPLE-VALUE-BIND (NAME VALUE-INDEX) (let ((*package* *keyword-package*)) (READ-FROM-STRING LINE)) (LET ((VALUE (READ-FROM-STRING-FIX LINE :START VALUE-INDEX))) (unless (numberp value) (setq value (intern (string value) 'keyword))) (case name ;; These properties are used, and DON'T go on the property list (:font_ascent (setq font-ascent value)) (:font_descent (setq font-descent value)) (:default_char (setq default-character value)) (otherwise (setq properties (list* name value properties)) (incf real-property-counter) (case name ;; These properties are used and DO go on the property list (:point_size (setq point-size value)) #+comment ;; what is this property anyway??? - LGO (:x_height (setq x-height value)))))))) (VALIDATE-STRING "ENDPROPERTIES" "terminator for font properties") (WHEN (OR (NULL FONT-ASCENT) (NULL FONT-DESCENT)) (FERROR NIL "Must have FONT_ASCENT and FONT_DESCENT properties")) ;; We shouldn't have to do this, but some fonts have incorrect fontboundingbox ;; specifications (probably because bdftosnf donesn't use it) (setq font-height (max font-height (+ font-ascent font-descent)) font-baseline (- font-height (max font-y font-descent))) (SETQ NUMBER-CHARACTERS (READ-FROM-STRING (VALIDATE-STRING "CHARS " "character count"))) (unless x-height (setq x-height (if (and font-ascent font-descent) (+ font-ascent font-descent) (ROUND (* POINT-SIZE ;; Convert points to inches (/ 1 72.27) ;; Convert inches to pixels Y-RESOLUTION))))) (unless FONT-ASCENT (setq font-ascent (if font-descent (- font-height font-descent) ;; Assume that the ascender covers about 80% of a character. (* FONT-HEIGHT 0.8)))) (setq font-desc (FED:MAKE-FONT-DESCRIPTOR :MAKE-ARRAY (:LENGTH (MAX NUMBER-CHARACTERS 256.)) FED:FD-FILL-POINTER (MAX NUMBER-CHARACTERS 256.) FED:FD-NAME FONT-SYMBOL FED:FD-SPACE-WIDTH (ROUND (* ;; Convert height to width (/ POINT-SIZE GOLDEN-RATIO) ;; Convert points to inches (/ 1 72.27) ;; Convert inches to pixels Y-RESOLUTION)) FED:FD-LINE-SPACING (+ FONT-ASCENT FONT-DESCENT) FED:FD-BASELINE FONT-BASELINE FED:FD-BLINKER-WIDTH (ROUND (* ;; Convert height to width (/ POINT-SIZE GOLDEN-RATIO) ;; Convert points to inches (/ 1 72.27) ;; Convert inches to pixels Y-RESOLUTION)) FED:FD-BLINKER-HEIGHT FONT-HEIGHT)) ;; Loop through each character in this font. (LOOP WITH BOUNDING-BOX-WIDTH = NIL WITH BOUNDING-BOX-HEIGHT = NIL WITH BOUNDING-BOX-LEFT = NIL WITH BOUNDING-BOX-BOTTOM = NIL FOR CHARACTER-COUNTER FROM 0 BELOW NUMBER-CHARACTERS ;; 1234567890 WHILE (string= (SETQ LINE (READ-A-LINE FONT-STREAM)) "STARTCHAR " :end1 10) FOR CHARACTER-NAME = (SUBSEQ LINE 10) FOR LAST-ENCODING = NIL THEN ENCODING FOR ENCODING = (LET ((ENCODING (VALIDATE-STRING "encoding "))) (MULTIPLE-VALUE-BIND (VALUE END-POSITION) (READ-FROM-STRING ENCODING) (WHEN (AND (NOT (= END-POSITION (LENGTH ENCODING))) (= VALUE -1)) ;; We have two values. Get the second one since the ;; first one looks kind of trashy (I don't know what ;; the real meaning is behind the -1 value). (SETQ VALUE (READ-FROM-STRING-FIX ENCODING :START END-POSITION))) (WHEN (= VALUE -1) #+comment (FORMAT T "~%Character ~A with encoding = -1 ignored at line ~D" CHARACTER-NAME LINE-NUMBER) (LOOP FOR LINE = (READ-A-LINE FONT-STREAM) UNTIL (EQUAL LINE "ENDCHAR"))) (WHEN (> VALUE MAX-ENCODING) (FERROR NIL "Character `~A' has encoding(=~D) ~ which is too large" CHARACTER-NAME VALUE)) (SETQ LARGEST-ENCODING (MAX VALUE LARGEST-ENCODING)) VALUE)) unless (minusp encoding) do ;; The following three things are unused, but are calculated in ;; the C program, so we do this here in case we need them later. (let* ((CHARACTER-ROW (LOGAND (ASH ENCODING -8) #xFF)) (CHARACTER-COLUMN (LET ((COLUMN (LOGAND ENCODING #xFF))) (SETQ FI.FIRST-ROW (MIN FI.FIRST-ROW CHARACTER-ROW) FI.LAST-ROW (MAX FI.LAST-ROW CHARACTER-ROW) FI.CH-FIRST (MIN FI.CH-FIRST COLUMN) FI.CH-LAST (MAX FI.CH-LAST COLUMN)) COLUMN)) (SWIDTH (VALIDATE-STRING "SWIDTH ")) ;; X component of width (X-WIDTH (READ-FROM-STRING (VALIDATE-STRING "DWIDTH "))) (BOUNDING-BOX-STRING (VALIDATE-STRING "BBX" "bounding box")) (ATTRIBUTES 0)) swidth character-column ;; not used ;; Decode the components of the bounding box. (MULTIPLE-VALUE-BIND (VALUE HEIGHT-POSITION) (READ-FROM-STRING BOUNDING-BOX-STRING) (SETQ BOUNDING-BOX-WIDTH VALUE) (MULTIPLE-VALUE-BIND (VALUE LEFT-POSITION) (READ-FROM-STRING-FIX BOUNDING-BOX-STRING :START HEIGHT-POSITION) (SETQ BOUNDING-BOX-HEIGHT VALUE) (MULTIPLE-VALUE-BIND (VALUE BOTTOM-POSITION) (READ-FROM-STRING-FIX BOUNDING-BOX-STRING :START LEFT-POSITION) (SETQ BOUNDING-BOX-LEFT VALUE BOUNDING-BOX-BOTTOM (READ-FROM-STRING-FIX BOUNDING-BOX-STRING :START BOTTOM-POSITION))))) (WHEN (OR (ZEROP BOUNDING-BOX-WIDTH) (ZEROP BOUNDING-BOX-HEIGHT)) ;; The character has zero size. Make sure that the left and ;; bottom dimensions don't affect the minimum/maximum ;; calculations. (SETQ BOUNDING-BOX-WIDTH 0 BOUNDING-BOX-HEIGHT 0 BOUNDING-BOX-LEFT MIN-LEFT-SIDE-BEARING BOUNDING-BOX-BOTTOM (- MIN-FONT-DESCENT))) (SETQ LINE (READ-A-LINE FONT-STREAM)) ;; 12345678901 (WHEN (STRING-EQUAL (SUBSEQ LINE 0 11) "ATTRIBUTES ") ;; The attributes line is optional. (SETQ ATTRIBUTES (READ-FROM-STRING-FIX (CONCATENATE 'SIMPLE-STRING "#\x" LINE) :START 11)) (SETQ LINE (READ-A-LINE FONT-STREAM))) (SETQ LOWEST-ENCODING (MIN LOWEST-ENCODING ENCODING)) ;; Write information out regarding this character. (setf (aref x11-metrics encoding) (X11:MAKE-X-CHAR-INFO :LEFT-SIDE-BEARING BOUNDING-BOX-LEFT :RIGHT-SIDE-BEARING (+ BOUNDING-BOX-LEFT BOUNDING-BOX-WIDTH) :CHARACTER-WIDTH X-WIDTH :ASCENT (+ BOUNDING-BOX-HEIGHT BOUNDING-BOX-BOTTOM) :DESCENT (- BOUNDING-BOX-BOTTOM) :ATTRIBUTES ATTRIBUTES)) (SETQ MAX-LEFT-SIDE-BEARING (MAX MAX-LEFT-SIDE-BEARING BOUNDING-BOX-LEFT) MAX-RIGHT-SIDE-BEARING (MAX MAX-RIGHT-SIDE-BEARING (+ BOUNDING-BOX-LEFT BOUNDING-BOX-WIDTH)) MAX-CHARACTER-WIDTH (MAX MAX-CHARACTER-WIDTH X-WIDTH) MAX-FONT-ASCENT (MAX MAX-FONT-ASCENT (+ BOUNDING-BOX-HEIGHT BOUNDING-BOX-BOTTOM)) MAX-FONT-DESCENT (MAX MAX-FONT-DESCENT (- BOUNDING-BOX-BOTTOM)) MIN-LEFT-SIDE-BEARING (MIN MIN-LEFT-SIDE-BEARING BOUNDING-BOX-LEFT) MIN-RIGHT-SIDE-BEARING (MIN MIN-RIGHT-SIDE-BEARING (+ BOUNDING-BOX-LEFT BOUNDING-BOX-WIDTH)) MIN-CHARACTER-WIDTH (MIN MIN-CHARACTER-WIDTH X-WIDTH) MIN-FONT-ASCENT (MIN MIN-FONT-ASCENT (+ BOUNDING-BOX-HEIGHT BOUNDING-BOX-BOTTOM)) MIN-FONT-DESCENT (MIN MIN-FONT-DESCENT (- BOUNDING-BOX-BOTTOM))) ;; Can't use validate-string here because the attributes line ;; is optional. (WHEN (NOT (STRING-EQUAL LINE "BITMAP")) (FERROR NIL "Missing font bitmap on line ~D, got ~A instead." LINE LINE-NUMBER)) ;; Loop collecting each row of the character. (LOOP with start-row = (- font-baseline (+ BOUNDING-BOX-HEIGHT BOUNDING-BOX-BOTTOM)) #+comment (min (max (- font-ascent (+ BOUNDING-BOX-HEIGHT BOUNDING-BOX-BOTTOM)) 0) (1- font-height)) ; These min/maxes are a kludge... do better someday WITH CHAR-DESC = (FED:MAKE-CHAR-DESCRIPTOR :MAKE-ARRAY (:element-type 'BIT :LENGTH (LIST font-height BOUNDING-BOX-WIDTH)) FED:CD-CHAR-WIDTH X-WIDTH FED:CD-CHAR-LEFT-KERN 0) FOR ROW FROM start-row ;;(+ start-row fudge) FOR RAW-DATA = (READ-A-LINE FONT-STREAM) WHILE (NOT (EQUAL RAW-DATA "ENDCHAR")) ;; Loop through all of the bits of this row. ;; The data comes in left justified, so we need ;; to shift it to make it right justified. DO (LOOP with DATA = (parse-integer raw-data :radix 16.) WITH SHIFTED-DATA = (ASH DATA (- BOUNDING-BOX-WIDTH (* 4 (LENGTH RAW-DATA)))) FOR COLUMN FROM (1- BOUNDING-BOX-WIDTH) DOWNTO 0 DO (PROGN (SETF (AREF CHAR-DESC ROW COLUMN) (LOGAND SHIFTED-DATA 1)) (SETQ SHIFTED-DATA (ASH SHIFTED-DATA -1)))) FINALLY (SETF (AREF FONT-DESC ENCODING) CHAR-DESC)))) (progn ;; WHEN OUTPUT-PATHNAME (SETQ CONSTANT-METRICS (AND (= MIN-LEFT-SIDE-BEARING MAX-LEFT-SIDE-BEARING) (= MIN-RIGHT-SIDE-BEARING MAX-RIGHT-SIDE-BEARING) (= MIN-CHARACTER-WIDTH MAX-CHARACTER-WIDTH) (= MIN-FONT-ASCENT MAX-FONT-ASCENT) (= MIN-FONT-DESCENT MAX-FONT-DESCENT))) ;; Ensure correct space and blinker width for constant-metric fonts (when constant-metrics (setf (FED:FD-SPACE-WIDTH font-desc) max-character-width (FED:FD-BLINKER-WIDTH font-desc) max-character-width)) ;; Adjust the size of the font to only have those characters it defines. (SETF (FILL-POINTER FONT-DESC) (1+ LARGEST-ENCODING)) (ADJUST-ARRAY FONT-DESC (1+ LARGEST-ENCODING)) (FED:FONT-NAME-SET-FONT-AND-DESCRIPTOR FONT-SYMBOL FONT-DESC) ;; Canonicalize metrics vector (unless constant-metrics (loop with metrics = (make-array (1+ (- largest-encoding lowest-encoding))) for i from lowest-encoding to largest-encoding for j upfrom 0 do (setf (aref metrics j) (aref x11-metrics i)) finally (setq x11-metrics metrics))) ;; Calculate all-chars-exists (setq all-chars-exist t) (dotimes (i (1+ (- largest-encoding lowest-encoding))) (unless (aref x11-metrics i) (return (setq all-chars-exist nil)))) ;; Setup the font-info-record (setf (get font-symbol :x11-font-info) (X11:MAKE-FONT-INFO-RECORD :VERSION VERSION-STRING :ALL-EXIST ALL-CHARS-EXIST :DRAW-DIRECTION X11:LEFT-TO-RIGHT-FONT :NO-OVERLAP 0 ; HUH? :CONSTANT-METRICS (if constant-metrics 1 0) :TERMINAL-FONT (IF (AND CONSTANT-METRICS (ZEROP MAX-LEFT-SIDE-BEARING)) 1 0) :LINEAR T :FIRST-COL LOWEST-ENCODING :LAST-COL LARGEST-ENCODING :FIRST-ROW 0 :LAST-ROW 0 :N-PROPS REAL-PROPERTY-COUNTER :PROPERTIES PROPERTIES :LEN-STRINGS 0 ; WHAT IS THIS? ;; Try a bunch of default characters, starting with `?' and ending with ;; the first character in the font. :CH-DEFAULT (LOOP FOR CHAR IN `(#\? #\! #\@ #\# #\$ #\% #\& ,LOWEST-ENCODING) WHEN (PLUSP (AREF (W:FONT-CHARS-EXIST-TABLE (symbol-value FONT-SYMBOL)) (CHAR-INT CHAR))) RETURN (CHAR-INT CHAR)) :FONT-DESCENT FONT-DESCENT :FONT-ASCENT FONT-ASCENT :MIN-BOUNDS (X11:MAKE-X-CHAR-INFO :LEFT-SIDE-BEARING MIN-LEFT-SIDE-BEARING :RIGHT-SIDE-BEARING MIN-RIGHT-SIDE-BEARING :CHARACTER-WIDTH MIN-CHARACTER-WIDTH :ASCENT MIN-FONT-ASCENT :DESCENT MIN-FONT-DESCENT :ATTRIBUTES 0) :MAX-BOUNDS (X11:MAKE-X-CHAR-INFO :LEFT-SIDE-BEARING MAX-LEFT-SIDE-BEARING :RIGHT-SIDE-BEARING MAX-RIGHT-SIDE-BEARING :CHARACTER-WIDTH MAX-CHARACTER-WIDTH :ASCENT MAX-FONT-ASCENT :DESCENT MAX-FONT-DESCENT :ATTRIBUTES 0) :PIX-DEPTH 1 :GLYPH-SETS NUMBER-CHARACTERS :VERSION2 VERSION-STRING ;; Metrics NIL when constant :metrics (and (not constant-metrics) x11-metrics)))))) FONT-SYMBOL) (DEFUN READ-A-LINE (STREAM) "Reads a line from a font file." ;; Loop through each line read, skipping over special case lines. (LOOP FOR STRING = (MAKE-ARRAY 20 :ELEMENT-TYPE :STRING-CHAR :FILL-POINTER 0) with char do ;; Skip newline and formfeed (loop while (or (CHAR= (setq char (read-char stream)) '#.(INT-CHAR (LOGAND (CHAR-INT #\NEWLINE) #O177))) (CHAR= CHAR #\NEWLINE) (char= char #\linefeed))) ;; Read into string until a newline or formfeed (LOOP DO (VECTOR-PUSH-EXTEND CHAR STRING) (setq CHAR (READ-CHAR STREAM)) UNTIL (or (CHAR= CHAR '#.(INT-CHAR (LOGAND (CHAR-INT #\NEWLINE) #O177))) (CHAR= CHAR #\NEWLINE) (char= char #\linefeed))) ;; Skip over comment, empty and dot lines. WHILE (OR (STRING-EQUAL (SUBSEQ STRING 0 7) "COMMENT") ; For apl-s25 font (ZEROP (LENGTH STRING)) ; For cartoon font in crturz.bdf (STRING-EQUAL STRING (STRING #\CENTER-DOT))) ; For cursor font FINALLY (RETURN STRING))) (defun compile-x11-font (bdf-pathname &optional (xld-pathname "")) (let ((font-symbol (read-font bdf-pathname))) (setq bdf-pathname (parse-namestring bdf-pathname)) (compiler:DUMP-FORMS-TO-FILE (merge-pathnames (string-downcase (CONCATENATE 'SIMPLE-STRING X-FONT-OBJECT-PREFIX (PATHNAME-NAME bdf-pathname) "." (string (si::local-binary-file-type)))) xld-pathname) ;; Take font-symbol from the file name, to allow aliasing `((setq ,font-symbol ',(symbol-value font-symbol)) (setf (get ',font-symbol :x11-font-info) ',(get font-symbol :x11-font-info)) (let ((symbol (and si:fdefine-file-pathname (intern (string-upcase (pathname-name si:fdefine-file-pathname)) 'fonts)))) (when symbol (setf (symbol-value symbol) ,font-symbol) (setf (get symbol :x11-font-info) (get ',font-symbol :x11-font-info))))) `(:FONT-SYMBOL ,FONT-SYMBOL :PACKAGE :USER)))) (defun update-all-x-fonts (from to) (let ((files (directory (make-pathname :defaults from :type "BDF" :name :wild)))) (dolist (file files) (format t "~%~a " (pathname-name file)) ;; Skip 16b fonts (not handled yet) (unless (member (pathname-name file) '("K14") :test #'string-equal) (princ (compile-x11-font file to)))))) #| (update-all-x-fonts "stroke:/u1/XV11R3/fonts/bdf/" "tan:/u1/oren/fonts/") ;; Andrew fonts (let ((fudge -1)) ;; The .bdf files in this directory are incorrect, so compensate (update-all-x-fonts "tan:/u1/oren/andyfont/" "tan:/u1/oren/fonts/")) (update-all-x-fonts "tan:/u1/oren/andysrc/" "tan:/u1/oren/fonts/") (update-all-x-fonts "stroke:/u1/XV11R3/fonts/bdf/75dpi/" "tan:/u1/oren/fonts/") (update-all-x-fonts "stroke:/u1/XV11R3/fonts/bdf/misc/" "tan:/u1/oren/fonts/") |# #| (update-all-x-fonts "x11m:fonts.100dpi;" "x11m:fonts.100dpi;") (update-all-x-fonts "x11m:fonts.75dpi;" "x11m:fonts.75dpi;") (update-all-x-fonts "x11m:fonts.misc;" "x11m:fonts.misc;") (update-all-x-fonts "x11m:fonts.rel2;" "x11m:fonts.rel2;") |#
hans@acsu.buffalo.edu (Hans Chalupsky) (11/28/90)
Bjorn.Victor@docs.uu.se (Bjorn Victor) writes: >Is there anybody out there, still? One wonders, doesn't one? > .......... >I still have some problems, though, that I was wondering if anybody >has solved already. (How many of you are really using the X11 server?) Even if you already have the patches that solve the problems with the initial version, or you waited long enough and finally got Rel-6.1, you probably still won't want to use it unless you have never used a SPARC station or thelike before. Maybe our Explorer-II's don't have enough memory (8Meg), but they are awfully slow as X-servers which defeats the purpose of a sophisticated user interface. >(How many of you are really using Explorers any longer...?) This is what I wonder too. I've been a Lisp machine buff for a long time but I think Lisp machines will go extinct soon. Here are some reasons: - other common lisps are getting better and better (if I want to know whether some of the code I write is real portable Common Lisp I try to compile it in Allegro or Lucid, not on an Explorer) - there is no real speed advantage anymore (actually, in our environment Explorers are at the low end of the speed scale now - but the comparison to other systems like Suns might be unfair because of different memory configurations) - development environments (such as various fancy lisp-modes for GNU emacs) are (in my opinion) better then the TI-Explorer environment (there are still a few things missing such as window based debuggers and inspectors, but you can get approximations of those in the Allegro composer or Sun's SPE) - last but not least, I want to have a full integration with all the nice stuff I can have from the UNIX environment, without having to have a re-implementation of each and every thing in Lisp (e.g., news, mail, revision control (rcs), x-server, window managers, nice tex/latex modes, automatic incremental BACKUPS done by system-managers,....). I admit that ZMACS is still nicer than - say - an X-based Emacs such as Epoch which comes pretty close with highlighing, region marking etc., but I'm willing to sacrifice this little extra bit of sophistication for Epoch's portability, wide availability and integration with Unix and X, especially,for the large number of users out there that constantly write new and better packages to make life easier (very important considering the dying audience of this news group). Maybe some of the reasons mentioned above are not quite that compelling if one considers a Mac+Explorer board (or Sun+Explorer board if it exists) configuration, but I don't think this is the right way to go. I'm open to any comments. I changed "religions" and I want to know what other people think about this issue. People in my department often ask me what my ideal Lisp development environment would look like. My answer is: I know what it would look like, it does not yet exist, but it certainly is not a Lisp machine anymore (be it Explorer or Symbolics). Hans hans@cs.buffalo.edu Hans Chalupsky, Dept. of CS, 226 Bell Hall, SUNY@Buffalo, NY 14260.
Bjorn.Victor@docs.uu.se (Bjorn Victor) (11/28/90)
>Even if you already have the patches that solve the problems with the initial >version, or you waited long enough and finally got Rel-6.1, you probably still >won't want to use it unless you have never used a SPARC station or thelike >before. Maybe our Explorer-II's don't have enough memory (8Meg), but they are >awfully slow as X-servers which defeats the purpose of a sophisticated user >interface. Yes, I know. I *have* used SPARCstations running X11R4, quite a lot, and now I'm using an Explorer I (not II), so the performance is... well... not very high. But on the other hand, it's better with an Explorer screen in my office than a SPARCstation in the lab, until we get some fresh funds to buy me a modern workstation. And, to lead into the other subject, it's a real nice hacking machine. I can spend hours fixing a subtle bug, or re-implementing a nice feature from somewhere else, and it's FUN! On the other hand, it's fun hacking Emacs lisp too (e.g. re-implementing Zmacs features), and on the third hand, I really should do something completely different -- my research. >>(How many of you are really using Explorers any longer...?) > >This is what I wonder too. I've been a Lisp machine buff for a long time but I >think Lisp machines will go extinct soon. Here are some reasons: > >- other common lisps are getting better and better (if I want to know whether > some of the code I write is real portable Common Lisp I try to compile it in > Allegro or Lucid, not on an Explorer) >- there is no real speed advantage anymore (actually, in our environment > Explorers are at the low end of the speed scale now - but the comparison > to other systems like Suns might be unfair because of different memory > configurations) >- development environments (such as various fancy lisp-modes for GNU emacs) > are (in my opinion) better then the TI-Explorer environment (there are still > a few things missing such as window based debuggers and inspectors, but you > can get approximations of those in the Allegro composer or Sun's SPE) >- last but not least, I want to have a full integration with all the nice > stuff I can have from the UNIX environment, without having to have a > re-implementation of each and every thing in Lisp (e.g., news, mail, > revision control (rcs), x-server, window managers, nice tex/latex modes, > automatic incremental BACKUPS done by system-managers,....). I agree, mostly. I still haven't seen a lisp-mode for Emacs that beats the Explorer environment, though, but some would say that I haven't seen Harlequin. >I admit that ZMACS is still nicer than - say - an X-based Emacs such as Epoch >which comes pretty close with highlighing, region marking etc., >but I'm willing to sacrifice this little extra bit of sophistication for >Epoch's portability, wide availability and integration with Unix and X, >especially,for the large number of users out there that constantly write new >and better packages to make life easier (very important considering the dying >audience of this news group). This is tricky. Zmacs has some very nice features, and inherits (so to speak) some other very nice features from the LISPM environment. GNU Emacs, on the other hand, has a better design in many aspects (e.g. less Lisp-specific (e.g. syntax), mode-hooks, regexps), but a less good design in some aspects (e.g. no difference between m-X and c-m-X commands, less good undo, no separate redo, less powerful language, worse compiler). I think Emacs will eventually have most of Zmacs' "graphic" features (e.g. multiple fonts), but some of the other bugs are very hard to get out. And it will always live with 8-bit characters, which is a loss from the 12-bit LISPM characters. You can't both have "clean" ISO 8859-1 and a Meta key, for instance. >Maybe some of the reasons mentioned above are not quite that compelling if one >considers a Mac+Explorer board (or Sun+Explorer board if it exists) >configuration, but I don't think this is the right way to go. I agree that this is not the Right Way. >I'm open to any comments. I changed "religions" and I want to know what >other people think about this issue. People in my department often ask me what >my ideal Lisp development environment would look like. My answer is: I know >what it would look like, it does not yet exist, but it certainly is not a Lisp >machine anymore (be it Explorer or Symbolics). Well, it's certainly not an Explorer, the Symbolics environment is *much* better (sorry, RMS and all old CADR-hackers). But there is a difference between "ideal Lisp environment" and "ideal workstation environment". An ideal Lisp environment doesn't (ideally) need anything else than a Lisp environment. On the other hand, I guess that not many of us are plain Lisp hackers anymore, but need and use more and more of the good tools that have/are emerging in the UNIX environment. -- Bjorn Victor Bjorn.Victor@DoCS.UU.SE Dept. of Computer Systems tel: +46 18183169; fax: +46 18550225 Uppsala University, Sweden "I'd rather hack a Lisp Machine!" (yes!)
lgm@cbnewsc.att.com (lawrence.g.mayka) (11/29/90)
In article <9011280928.AA06943@mizar.DoCS.UU.SE>, Bjorn.Victor@docs.uu.se (Bjorn Victor) writes: > I agree, mostly. I still haven't seen a lisp-mode for Emacs that > beats the Explorer environment, though, but some would say that I > haven't seen Harlequin. Harlequin indeed has the best programming environment I've seen on a conventional processor, though I certainly wouldn't trade my Symbolics for it. > >Maybe some of the reasons mentioned above are not quite that compelling if one > >considers a Mac+Explorer board (or Sun+Explorer board if it exists) > >configuration, but I don't think this is the right way to go. > > I agree that this is not the Right Way. I myself think a coprocessor solution is quite reasonable for someone who feels a need for both environments. The people I know who are getting Symbolics UX boards for their beloved Suns apparently agree. > anything else than a Lisp environment. On the other hand, I guess > that not many of us are plain Lisp hackers anymore, but need and use > more and more of the good tools that have/are emerging in the UNIX > environment. And then, others of us are moving in the opposite direction. Lawrence G. Mayka AT&T Bell Laboratories lgm@iexist.att.com Standard disclaimer.
Bjorn.Victor@docs.uu.se (Bjorn Victor) (11/29/90)
>> anything else than a Lisp environment. On the other hand, I guess >> that not many of us are plain Lisp hackers anymore, but need and use >> more and more of the good tools that have/are emerging in the UNIX >> environment. > >And then, others of us are moving in the opposite direction. Really? What do people do in Lisp nowadays? I thought the trend was basically for C++, ML or Prolog, i.e. either more "basic" (but object-oriented) languages or more "clean" languages? -- Bjorn