* Remake files to convert the 247Q package-delimiter in DEFINE-FILE-INFO expressions to 30Q
* Remake TRANSOR after removing HIST command
* Remake TRANSOR-LOADTRAN after changing the filecoms variable
* MCCS translations for strings passed to/from UNIX-GETENV and other system interfaces.
* INTERPRET.REM.CM assumes system external format is UTF-8. ISO8859/1 external format is defined in MCCS as a dummy for UTF-8 until UNICODE is loaded
* Add string translation interface to EXTERNALFORMAT datatype,
* Set the external format of the default reader environment to *DEFAULT-EXTERNALFORMAT* = :MCCS
* Add external format :THROUGH16 for 16 bit codes, used by linebuffer
* TEDIT adds TEDIT.PARAGRAPH.BOUNDARIES, fixed TEDIT.MAP.OBJECTS, a few other bugs
* Addresses the page layout menu #2447 and #2457
* TEDIT-STREAM add open method
* Ctrl-p inserts page number object
* Restore page number pretext and posttext in menu
evaluating a variable bound to NOBIND causes errors when run interpreted (doesn't when run compiled). It's not clear why this design decision was made, but in this case compiling MASTERSCOPE was hung up trying to run a subfuncion involved in macroexpansion after LOAD(MASTERSCOPE).
Revise and update the repo's README to address issue #2449 by adding an
introduction to the project; referring to the glossary for
system-specific terms; replacing section Running Medley Interlisp with a
reference to the Medley primer; and removing the obsolete link "docs --
Documentation files".
* Remove code for archaic \FONTFILENAME conventions, MEDLEYFONTFORMAT now calls the generic function.
* Add OCTALSTRING to APRINT so FONT can use it. Eventually remove it from PUP
The file lispusers/BLOCKS-HKB is a duplicate of lispusers/h/H-BLOCKS
except with differing end-of-line convention. This PR removes
BLOCKS-HKB and BLOCKS-HKB.LCOM
Closes#2079
* Fixed dumping of the running page header that was from the HEADER option of TEXTTOIMAGESTREAM.
* Restored (CONSTANT ...) around a TIMES call (that *really* ought to optimize at compilation).
* Separate interface for imagefile creation from the send-to-printer interface
* SEND.FILE.TO.PRINTER recognizes :DEFAULTPRINTER
* Various changes to address #2414
* Move BITMAP properties from PRINTERTYPES to PRINTFILETYPES.
* Fix ShellOpen, add UNIX-TMP-FILE-NAME
* Include COERCEFONTSPEC changes in anticipation of HTML streams
* PDFSTREAM compatible with new imagefile architecture plus able to convert non-local Postscript streams
* SKETCH compatible with new imagefile/printing architecture
* TEDIT compatible with new printing architecture
* Tedit files are of type TEDIT, not TEXT, new interface function TEDIT.TO.IMAGEFILE
* ATTACHEDWINDOW: DOATTACHEDWINDOWCOM allows menu to have a form to EVAL, like the background menu
* WINDOW: fix menus for new hardcopy architecture
* Remove FLUSHFONTSINCORE--FLUSHFONTCACHE is more general
* LOAD character names as suggested in PR #2398
* HARDCOPYW respects file extension
* VIEWER as default printinghost
* SLASHIT interprets '.
* MCCS to UTF8 conversion on printer name
* Deal with {LPT}.LOCAL and upper casing
* @ LPT printers work with exact upper/lower matching
* Using NSPRINT functions to recognize fax
* Let "UNIX" be the name of the default printer of type UNIX
* Improved matching for TEDIT-ABBREVS
* Make CHARACTERNAMES available at runtime (SOFT-HYPHEN etc)
* TEDIT-ABBREV: Better parsing strategy
* Added backslash markup
* Document new abbreviation conventions
* Glitch
* Responded to @MattHeffron
* Include backslash in promptwindow message
* Updated edit date
This should complete the changes for the MEDLEYDISPLAYFONTFORMAT.
The building a composite font from BDF files implemented.
Resolves#2313Resolves#2365
This should make #2364 possible.
(I created a size 20 Noto Sans _composite_ font from the same Google
fonts as used by @hjellinek's HTMLSTREAM.)
**NOTE:** This uses the function `IL:CLSMARTEN` which is from the file `CLSMARTARGS`.
The file `CLSMARTARGS` isn't loaded until almost immediately **after** `XCL-EXTRAS`.
There are no uses of `DEFINE-RECORD` in making the `lisp.sysout` loadup, so this _ought_ to be safe, but this must be verified!
Checking for charset *known* to be EQ to the SLUG charset, didn't verify that the font *had* a SLUG charset, so was comparing to NIL, which excluded *all* unloaded charsets.
* Upgrade Sketch to new font interfaces
* Tedit changes so that TEDIT.GETSYNTAX conforms to documentation
* Remove RUBOUT interrupt
* TEDIT-FNKEYS: uses keywords for action names
* LLKEY: Change keybindings for DELETE and LOOKS
* SEDIT-COMMANDS: Shift-DELELETE is wordelete.backwards
* Sketch implements arrow keys
I'm not sure what I did wrong with PR #2396 but it didn't revert
lispusers/BACKGROUND-YIELD(.LCOM).
I did what I thought was the same process, but it seems to have picked
up the right files this time.
Removed WRITE-BDF-TO-DISPLAYFONT-FILE (i.e., no STRIKE format files).
No multiple-values returned. Instead, use LIST when appropriate.
BDF-TO-CHARSETINFO is now IDEMPOTENT w.r.t. the GLYPHS.
Move resolution and defaulting of FAMILY, FACE, SIZE, ROTATION, DEVICE from WRITE-BDF-TO-MEDLEYDISPLAYFONT-FILE to BDF-TO-FONTDESCRIPTOR.
Keep the MCCS chars present BITMAP in the BDF-FONT structure, instead of needing to schlepping it around separately.
Abstracted testing/setting the MCCS chars present bits to CHAR-PRESENT-BIT (mimicking BITMAPBIT).
Added COUNT-MCHARS to know how many MCCS chars are marked in the BITMAP as present.
READ-BDF now handles when UTOMCODE? returns multiple mappings, and creates the appropriate duplicate GLYPHS with different MCCS char codes.
READ-GLYPH doesn't create an empty BITMAP for spacing glyphs.
Use font code changes:
Set (CHARSETINFO CHARSETNO).
Set (FONTDESCRIPTOR FONTSLUGWIDTH).
* loadup scripts: set MEDLEY-INIT-VARS to NOBIND before loading MEDLEYDIR
so that INITVARS in MEDLEYDIR replaces any previous settings in the loadup-sysout
* FONT: add cache for FONTSAVAILABLE, add FLUSHFONTCACHE. See FONTCODECHANGES.tedit
* Make FONTEXISTS and FONTSAVAILABLE consistent
* Add BACKGROUND-YIELD to Lisp.sysout (and thus to FULL.SYSOUT) so you don't need to load it in INITs.
* remove CAUSE-INTERRUPT subr call; doesn't add value
* Verbose mode (READ-BDF) was implemented incorrectly - fixed
* Cleanup DEFPACKAGE in source file using :IMPORT-FROM, and fewer imports.
* Various renaming for consistency with XCCS -> MCCS changes.
* Use IL:FONTSPEC record instead of using FIRST, SECOND, etc.
* Fix the parsing of IL:FONTSPEC to use COMPRESSED instead of incorrect CONDENSED.
* Zero-width "image" with zero-width "escapement" GLYPHS now get put into NOMAPPINGCHARSET.
* Add (FILES (SYSLOAD) SYSEDIT) under existing (DECLARE: EVAL@COMPILE DONTCOPY ...)
Various renaming for consistency with XCCS -> MCCS changes.
Use IL:FONTSPEC record instead of using FIRST, SECOND, etc.
Fix the parsing of IL:FONTSPEC to use COMPRESSED instead of incorrect CONDENSED.
Zero-width "image" with zero-width "escapement" GLYPHS get put into NOMAPPINGCHARSET.
Add (FILES (SYSLOAD) SYSEDIT) under existing (DECLARE: EVAL@COMPILE DONTCOPY ...)
* Add a CLOSEFN to KINETIC window that forces the kinetic to stop and exit when closew is called.
* Reset the window close flag after window is closed so that can repeatedly close and open the KINETICWINDOW.
* Add :IMPORT-FROM option to DEFPACKAGE
* Fix a couple of variable reference typos.
Change the FILETYPE to be :FAKE-COMPILE-FILE (per #2336)
* Fix error of missing arg in call to IMPORT.
prc status is C if changes requested, prc comparison runs in its own process, initial changes for gwc to treat the clone as UNIX, not DSK, so branch-switching doesn't cause files from other branches to pile up.
* COMPAREDIRECTORIES: don't fail on empty directory
* Bug in slashit
* gwc copies to UNIX--doesn't track Medley version numbers when it copies to the clone
* Coerce MCCS filename strings to UTF8 filename strings in file-name system calls, coerce system filenames back to MCCS codes
* Add UNICODE-TABLES so MTOUTF8STRING gets defined in right place in the loadup sequence
* ADIR: Bug fix: UNPACKFILENAME sets FATSTRINGP
* fix virtualkeyboard bug in code assignment
* Unicode canonicalizes non-SMALLP unicodes
* Estimates suggested width for unformatted documents from the width of the first 20 lines
* Recycled regions satisfy minimum and maximum size specifications
* Fine tuning: suggests recycled no bigger than 90% of screen, no prompt message if old region reused
* A revision to the font, Unicode, Tedit, and other modules to implement the MCCS character coding as the standard for internal text strings. MCCS is a variant of XCCS with arrows switched with circumflex/underscore and $ switched with currency, and allows for additional code assignments over time. :MCCS replaces :XCCS as the default external format, especially for source files. The file XCCS is removed in favor of the file MCCS, which includes the XCCS external format for backward compatibility.
* This includes a single Medley-font formatted font file for each of the family/size/face display fonts. The glyph assignments correspond to the MCCS character encoding (except for fonts with idiosyncratic encodings--Hippo, Symbol). All charsets from legacy font files are included in each file, and the character sets and glyphs in each file have also been extended by offline coercion from related families (e.g. Glyphs not in legacy Terminal are taken from legacy Modern). There should be fewer black boxes, and character-display shouldn't change when you switch fonts.
* The Unicode mapping tables have been redefined to set up correspondences between Unicode and MCCS, not XCCS. Separate XCCS to/from MCCS mapping functions are provided in the file MCCS; they are no longer included in INTERPRESS.
* TEDIT converts characters in legacy fonts to their new MCCS codes as it reads formatted files, marks the file as MCCS compatible and preserves the new codes on writing.
* Default keyboard assignments produce the MCCS uparrow and leftarrow for shift-6 and shift-hyphen, use Function-6 for circumflex and Function-10 for underscore.
See documentation in FONTCODECHANGES.TEDIT MCCS.TEDIT MEDLEYFONTFORMAT.TEDIT in docs/internal, and library/UNICODE.TEDIT.
* EDITINTERFACE--ED searches for symbols, no error when declining a loadfns
* Symbol by type menu pops up when the symbol with different package qualifiers have different types
* CLIPBOARD passes file to xclip (windows)
* More careful to specify {UNIX} as the host on the Medley side
* Make sure the process stream is closed in the xclip case
* This time using ShellCommand
* Use DSK and UNIX-FILE-NAME
* Added .txt to the tmp file
* Move PRESS and PRESSFROMNS to obsolete
* Remove PRESS from the loadup
Also moves MULTI-ALIST forward to the beginning of LOADUP-LISP
* BACKGROUNDIMAGES depended on PRESSFILEP, now moved here from PRESS
Update project contributor guidelines
Replace the documentation link to point to the Documentation section of the project site instead of the discontinued Wiki.
Under Working with Medley, advise opening an issue to report errors and unexpected behaviors.
Under Did you write a patch that fixes a bug?, remove the link to Maiko PRs as the project has more repos a contributor may want to open PRs for.
Signed-off-by: Paolo Amoroso <paolo.amoroso@gmail.com>
* Column major order to match XCCS and Unicode tables
* Compile with exports
* HLDISPLAY added a title argument to EDITBM
* EDITFONT puts the character and font in the title of the EDITBM window
* Put character set in edit window title
* Modernize the EDITBM window
so that it can be moved and resized like other modern windows
* Move lispusers/FASTEDITBM to obsolete/lispusers
It isn't used, doesn't even have an LCOM. But maybe it would be worth resurrecting at some time in the future.
* WIP in changing loadups to tagged nomeclature rather that branch and ability to delete tagged branches as well as thin them.
* In loadup and medley scripts, changed the --branch (-br) command line argument to --tag (-tg). Correspondingly changed the MEDLEYDIR/loadups/branches directory to be MEDLEYDIR/loadups/tagged. Separately, added a + parameter to the --thinw and --thinl command line arguments which causes ALL files to be deleted, not just the versioned files. Added a --thin command line argument that is equivalent to --thinw followed by --thinl. Updated the man page to reflect all these changes.
* Updated medley script man page to reflect the new tag nomenclature instead of branch nomenclature.
* Fix issue#2063 - typo in medley man page
* 1) In loadup script made --help (-h) be a synomym of --man (-z). 2) In both loadup and medley scripts added an optional parameter (+) to the --man (z) command line argument that when specified will display the man page without paging - i.e., sets the pager to cat.
* Fix Issue#2251 LOADUPSDIRECTORIES settings by the MEDLEYDIR fn now takes into account tagged loadups
CHARSET.DECODE, CHARSET.ENCODE added. CHARACTER.ENCODE moved here from TEDIT-FNKEYS. Some character names moved here from a few other files.
This anticipates the ability to use names for the charset argument to FONTCREATE. IOCHAR needed to move a little earlier so that STRING.EQUAL/UPPERCASEARRAY is available before FONTCREATE created the guaranteed displayfont.
This PR contains a large number of changes in support of the implementation of the Medley Dsplay Fon file format.
The changes are documented in the docs/internal/FONTCHANGES.TEDIT file.
For loadup --aux reporting of errors during building whereis.hash:
Just ignore pathnames that are just a directory.
For other errors, report the condition that caused the error, instead of just generic Warning message
* Update lispusers/INSPECTCODE-TEDIT for changes to Tedit.
This could be considered to be incorporated directly into sources/INSPECT.
(This was changed from TCOMPL (LCOM) to COMPILE-FILE (DFASL) because it just wouldn't compile correctly otherwise, for me.)
---------
Co-authored-by: rmkaplan <ron.kaplan@post.harvard.edu>
* Inspectcode scrolls, has the correct window title
* Add promptwindo so M-f search strings are locally visible
* Fix FILETYPE for LLDISPLAY
* MEDLEYFONTFORMAT More efficient store and read of numeric Interlisp arrays
* Fixed#2185 and #2173.
* TEDIT-MENU TEDIT-PAGE Fix a few more bugs
* TEDIT-STRESS--cleanup and harden a bit for future use
* TEDIT-FILE: Allow putting of an APPEND file. There is still a guard against putting a readonly file--use COPYFILE instead. Provide message if won't put
* Get proper insert caret looks before EOL
All display fontfiles are now in the single fonts/displayfonts/ directory. The subdirectories adobe, big, other, altofonts have been deleted, and are no longer on DISPLAYFONTDIRECTORIES. The few additional displayfonts from the envos repo are included.
This is the result of:
Removing all byte-identical files, leaving a representative of each such equivalence class in displayfonts/
Removing all files in displayfonts that are renamed byte-identical copies of other files in displayfonts/ (e.g. helvetica1-MRR-c0.displayfont is a fake-renamed copy of helvetica4-MRR-c0.displayfont). These are to be constructed on the fly by the font substitution mechanism.
Roughly comparing the bitmaps of non-identical files with the same font family/size/face/charset in the different subdirectories, and making displayfonts/ have the one that looks best (e.g. most glyphs, unless garbage), giving priority to ones that are already in displayfonts/.
Moving the remaining odd-ball files from the subdirectories to displayfonts/ (e.g. gates, music).
Copying from Envos the 30 or so files that are not byte-identical copies of files already in displayfonts/.
* Add to loadup and medley scripts a --branch argument whereby which loadup outputs are stored in MEDLEYDIR/loadups/branches/BRANCH, where BRANCH is the name of the currently active git branch of MEDLEYDIR (for loadup) or the name of a BRANCH whereby which a loadup had been run with the --branch option (for medley). Man pages updated accordingly.
* Removed --git-branch synonym for --branch from both medley and loadup scripts. Added optional BRANCH argument to --branch in loadup script. If BRANCH is specified iot is used as the (pseudo)branch name instead of the current active git branch of MEDLEYDIR. If BRANCH is "-", the current git branch name is used.
* When doing loadups, a file called gitinfo is created in the LOADUP_OUTDIR that contains the git commit, git branch and git status of MEDLEYDIR at the time of the loadup. Branch names in loadup and medley scripts now contain only alphanumerics, dahes, underscores, periods. Any other character is coerced to underscore. The branch name is now included in the window title.
* Incorporate changes from PR#2208: Use the loadup step script names as the --id string.
* Add loadups/gitinfo and loadups/branches to .gitignore
Fix editing of button...correctly wait for edit to complete before updating the button itself.
Add (FIND-BUTTON label) function.
Minor updates to documentation.
* Tedit window splitting is more robust, interface through menu items instead of split-region on the right of the window. See TEDIT-RELEASENOTES.TEDIT
* Tedit recognizes color as specfied by DSPCOLOR, passes it to hardcopy
* N-way buttons default to unsorted--new items go at the end. Otherwise keyboard shortcut meta-3 for the 3rd font might pick a different one depending on what went before.
* USER.CM can be specified as an opening property for Bravo conversion.
* Adresses/fixes Tedit issues #2173#2172#2171#2142#2105#2062#2059#1972 (maybe some others).
* Changes to rationalize internal interfaces and simplify code, and particularly to eliminate internal dependencies on the STREAMHINT Xpointer backlink. STREAMHINT is only accessed if a client has grabbed the TEXTOBJ and passes it back in. The stream and window are the safe/reliable way of referencing the Tedit state (and the window and stream know about each other, and know about the TEXTOBJ only through the stream).
* Many changes to TEDIT-STRESS, including new defaults CHECKARRAYS NIL, NSYSOUTS 0, ARRAYBLOCKCHECKING T
* lispusers/EQUATIONS: image object no longer saves state on the stream, not the window (which may not be there).
* Rename CHARNAME to be CHARCODE.ENCODE, parallel to CHARCODE.DECODE
Updated Build/Push Medley Release workflow to add wsl1-specific assets
to the releases - based on the new wsl1-specific maiko builds.
Renamed the current wsl release assets to be wsl2-speciic assets.
There are now 20 assets in a release build.
Add a --vnc flag to loadup script, used to tell loadup scripts to run
Medley with the --vnc option.
Used primarily on WSL1 and WSL2 systems where loadup defaults to X
Windows. But can be used on any Linux platform that uses Xvnc.
Not relevant on MacOS and Cygwin.
Man page updated accordingly.
* Fix POSTSCRIPTSTREAM to accept SMALLP values 0 and 1 as COLOR.
(Don't require them to be FLOATP.)
* Changed handling when color values are not directly interpretable without other context (e.g., as an index into a color map).
POSTSCRIPTSTREAM should behave similar to INTERPRESS and just ignore a BRUSHCOLOR if it is a SMALLP that's neither 0 nor 1, and use the current DSPCOLOR of the stream.
If someone wants a specific color, or grayscale, they can use the color descriptions as documented in POSTSCRIPTSTREAM.TEDIT.
* Update documentation to better describe allowed COLOR specification.
* Fix Issue #2155 (loadup -f -b -x fails can't find full.sysout). Reordered loadup so that the sysouts are copied to loadups before loadup-aux and loadup-db are run, and that the product of loadup-aux are copied to loadups before loadup-db is run -- all to make sure that the right prerequisites are available in loadups when needed. Also forced a run of loadup-aux if full.sysout is newer than exports.all when -db is specified.
* Fix Issue 2148 - loadups fail on WSL1 with Xvnc server error. Root cause of 2148 is that certain Medley sessions in loadups run very quickly - less that a second or 2. This causes the medley script to improperly detect an Xvnc server error. Fixed by adding a cli flag --automation that forces the medley script to skip the check for Xvnc server errors. Changed loadup script to call medley with this flag set. Also add a cl flag to loadup --forcevnc that forces loadup to use vnc even on WSL2. This is intended for testing purposes only.
* Collect TALK files into lispusers/talk/
renamed with hyphen convention. Also pull over TALKGAP files that were previous obsolete, and remove them from obsolete/lispusers
* Replace TALK.TEDIT--file got smashed
* TALK.TEDIT now comments the current situation
Fix Issue #2155 (loadup -f -b -x fails can't find full.sysout).
Reordered loadup so that the sysouts are copied to loadups before loadup-aux and loadup-db are run, and that the product of loadup-aux are copied to loadups before loadup-db is run -- all to make sure that the right prerequisites are available in loadups when needed.
Also forced a run of loadup-aux if full.sysout is newer than exports.all when -db is specified.
Here is what this PR does:
Fix Issue 2139: (MAKE-INDEX-HTMLS) was not handling pseudohosts correctly and an errant LI pseudohost was causing MAKE-INDEX-HTMLS to terminate early. Adjusted MAKE-INDEX-HTML so it uses psuedohosts only for the top level directory and everything further down in the tree uses the truenames relative to the top-level pseudohost, Results in a MAKE-INDEX-HTMLS run that works in the presence of random pseudohosts and in a collection of index.html files without difficult to understand and out of context references to pseudohosts.
Remove loadups/build directory from all HCFILES runs (on desktop and via github actions)
Added maiko source code and removed maiko lde executables from HCFILES outputs for github actions - thus adding maiko code and removing maiki executables @ files.interlisp.org.
Fixed scripts/clean_hcfiles.sh so that it actually cleans off all of the index.html files - was missing some.
.github/workflow directory was being left out of HCFILES. Put it back it.
Expanded the (PSEUDOHOSTS) function to have an argument, which is a list of replacement pseudohost pair to replace the current list of pseudohosts. Still returns the current list of pseudohosts as before. Designed to be used cleanly with RESETSAVE.
Three changes to loadup scripts:
1) dribble files are now copied from the workdir into loadups if loadup completes successfully,
2) if the lock preventing simultaneous runs is already set when loadup starts, it now asks the user if they want to override the lock or to exit (previously it just exited),
3) there is now a --override flag that will automatically override the lock without asking the user.
Man page updated accordingly.
Fix Issue#2151
MEDLEY-INIT-VARS now resets the LI pseudohost whenever it sets/resets the value of LOGINHOST/DIR - providing PSEUDOHOSTS is loaded and LI pseudohost already exists.
* I put in support for standard color specifications. I think this is pretty close.
BLTSHADE, FILLCIRCLE, and FILLPOLYGON with the TEXTURE as TEXTUREP or BITMAPP convert to a gray scale as a function of the number of bits set.
It'll take a bunch more refreshing of PostScript knowledge to figure out how to do these with real colors, and actual pixel-by-pixel textures.
* EXAMINEDEFS uses the Tedit atom-bound table for word selection
* REGIONMANAGER gives client better control over region recovered
* TEDIT-PF-SEE: Meta-T defaults to showing the definition in the curren…
* REGIONMANAGER and GITFNS documentation update
* Better handling of a glyph with ENCODING of -1.
I treat it as the _slug_ glyph, instead of the _default_ of a solid block.
* Handle scrambled bitmaps issue #2109.
Glyphs with zero width bitmap *and* zero advance (_escapement_) caused miscalculated glyph offsets into the CHARSETINFO bitmap.
* Allow and ignore COMMENT lines preceding the STARTFONT line.
Add error checking for extracting font FAMILY, SIZE, FACE, etc. from the BDF-FONT object.
Add recommendation to documentation to write the DISPLAYFONT files to a directory separate from the system's IL:DISPLAYFONTDIRECTORIES locations.
* Account for glyphs with a negative initial offset.
* Add VERBOSE optional parameter to READ-BDF to report font internal FAMILY, FACE, etc.
Change &OPTIONAL parameters of WRITE-BDF-TO-DISPLAYFONT-FILES to &KEY to simplify calling. (No need to remember the order.)
Add CHAR-SETS and WRITE-UNMAPPED parameters to WRITE-BDF-TO-DISPLAYFONT-FILES to allow some level of control of which DISPLAYFONT files are written.
Updated documentation, and added warning note about font's FAMILY containing any digits.
This PR is an extensive revamp of the loadup scripts including the following changes:
New omnibus script scripts/loadup which is meant to be the single interface to the loadup system. The man page for this script can be found here: https://online.interlisp.org/downloads/man_loadup.html
The new loadup script allows you to restart the loadup process from any particular stage - init, mid, lisp, full, etc. For example, you can start the loadup from an existing init-mid.sysout and have it run thru creating the full.sysout. The call for this would be: ./scripts/loadup --target full --start mid. (See man page for all options to the loadup script as well as examples of their use.)
In order to facilitate this target/start feature, the loadup workdir has been moved to a single location per MEDLEYDIR, specifically to MEDLEDIR/loadups/build. (Previously, it was different for every invocation in /tmp/loadups-$$)
When restarting the loadup from, say, lisp.sysout, the script will look for the lisp.sysout to start from first in MEDLEYDIR/loadups and then in MEDLEYDIR/loadups/build. If the starting sysout is found in loadups, it will be copied (non-versioned) into the workdir before doing the loadup, overwriting whatever is already there.
There is now a lock (MEDLEYDIR/loadups/build/lock) that prevents concurrent loadups from running (and very occassionally needs to be manually removed).
At successful completion of a loadup run, the created .sysouts and other files are moved into the loadups directory as before (unless the --nocopy option is specified). BUT the build files - including the dribble files - are left in the working directory (i.e., in loadups/build)
The loadup script now supersedes the former loadup-all.sh, loadup-full.sh and loadup-db.sh scripts. But I have left in their place scripts that call the new loadup script with the right options. loadup-all.sh is now just a link to the single loadup script, who's options are (hopefully) a superset of loadup-all.sh.
All of the worker scripts (e.g., loadup-full-from-lisp.sh) have been updated to use better mechanisms to catch and report errors, including the new (LOGOUT T EXITCODE) feature. The main script has been updated to better catch errors in these worker scripts when they do happen.
You can now specify a MAIKODIR for the loadup, either using the MAIKODIR env variable or thru the --maikodir command line option.
All of the loadup scripts have been moved down one level into scripts/loadups. But there are links from the scripts/ directory into the scripts/loadups directory for all of the top-level legacy scripts - loadup-all.sh, loadup-db.sh, loadup-full.sh so that you do not have to change your own scripts unless you need to use some of the new features. More importantly, none of the github workflows need to change right now.
The SYSOUTCOMMITS mechanism had to be changed since there is no guarantee that e.g., lisp.sysout and full.sysout are built on the same commit. So (ASSOC 'MEDLEY SYSOUTCOMMITS) now returns an assoc list of sysouts, e.g., ((INIT "aaa")(LISP "bbb")(FULL "ccc")(APPS "ddd")), showing the commits for the various layers of the loadup.
The new loadup scheme allows different sysouts to be created from different commits, so there is an issue with RDSYS (and RDSYS.LCOM) being out of sync with one or more of the sysout in loadups. RDSYS(.LCOM) are copied into library when ever a loadup of the Init stage completes successfully (unless the -nocopy option is specified). The only way to solve this issue when it arises is to do a complete loadup from starter.sysout to full.sysout (or apps.sysout) to ensure evrything is built on the same commit.
* Now can create the FONTDESCRIPTOR with all non-empty charsets.
Can write DISPLAYFONTFILE format ("STRIKE") files for the charsets.
Add ability to use mapping of Unicode charcode to unknown XCCS charcode in the private space.
* Create 2nd FONTDESCRIPTOR for unmapped Unicode to XCCS charcodes, organized by charset-like (8-bit splitting of charcode) of Unicode encoding value.
* Added option to create and write files for RAW FONTDESCRIPTOR which does NO mapping from Unicode to XCCS.
All glyphs are at the Unicode encoding positions.
Any glyphs with Unicode encoding > xFFFF are not included in the FONTDESCRIPTOR or DISPLAYFONT files.
* Fix a bug where I assumed glyph names couldn't be parsed as a number; and a little cleanup.
The linux otf2bdf utility uses the hex of encoding value as the name, which can appear to be a FLOAT and overflow (i.e., 3D39).
Similar parsing problem fixed and corrected an error message.
* Initial documentation file written.
1961 - RDSYS incomplete path for copying
2061 - delay starting making loadups due to "expensive" operation just to check if connected dir is inside of a git repo
* Distinguish public TEDIT.NTHCHARCODE... from private \TEDIT.NTHCHARCODE...
* Fix screen-update bug: deleting the character before the first character in a window
* glitch in region code
* Implement new key binding architecture for Tedit actions, to make mappings more user-accessible. Includes new Buttons item on Tedit menu. See TEDIT-RELEASENOTES
* Prompts for Tedit initial-window region based on document properties
* More flexible Tedit abbreviations
* Size + or - in Tedit character looks go to next available
* Other bug fixes and code adjustments.
* Fix Issue 2072: update do_hcfiles.sh to use new rem.cm argument to medley script. Also update medley script to print out the REM.CM file used (if any).
* Update buildReleaseIncDocker.yml to account for the change on interlisp/online repo from 'master' to 'main' as primary branch
* Do compile.sh after update to medley-run.sh to add REM.CM to printout when using the medley command.
DLIONFNKEYS created button-images on the screen corresponding to some of
the Dandelion edit buttons for Tedit. EDITKEYS is a newer and better
version of the same functionality, and even that will be replaced by a new internal Tedit keyaction architecture.
Currently, (CHARCODE 1) produces 49, just like (CHARCODE a) produces 97.
That's because CHARCODE.DECODE interprets a single character (that is
not also defined as a cHARACTERNAME) input as exactly that character,
and gets its character code.
But a character name in the charset,charnum format is treated in a
different way. If either part of the name is a number, it is interpreted
as an octal number, to make it easier to match against the XCCS
specifications. So (CHARCODE Meta,1) does not map to the position of
digit 1 in the Meta (currently 1) character set (= 256+49=305), it maps
to 256+1=257. In order to get 305, you have to specify Meta,61 (where 61
is the octal equivalent of 49).
That's rather opaque, and I found that TEDITKEY had a workaround based
on the obscure fact that CHARCODE.DECODE adds 128 for each # it sees in
the beginning of a character name. So ##1 adds 2*128 to 49 to get 305,
corresponding to digit 1 in the Meta character set. But that depends on
the fact that Meta is currently defined as charset 1, it won't work when
we move Meta up to the Unicode/XCCS undefined region of the code space.
This PR adds (Zero 48) (One 49)...(Nine 57) to the initial value of
CHARACTERNAMES in LLREAD, so that 305 for example can be specified as
Meta,One. This is more obvious than Meta,61 and will remain valid when
Meta is moved, unlike the ## hack..
Lafite is not in the fuller ms database, so I missed these references in
TEDIT-INDENT.
But note also: there are 2 definitions of the function
\GV.PARSERECIPIENTS1, one on lafite/TEDIT-PRIVATEDL and one on
internal/MAILCLIENT.
The one on TEDIT-PRIVATEDL references the variable ADDRESSPARSERRDTBL,
which is not initialized.
Selection highlighting should be correct if line contains a preceding
diacritic.
Find from expanded menu should work.
Screen and selection should not be garbled after substitute.
I have not yet replicated the break-on-put also reported in #2055
* Paren error when selection looks set to INVERTED
* Add SET, SHADE, and SHADEHEIGHT properties to TEDIT.SELPROP
* Add QUIET as optional last argument to TEDIT.FORMAT.HARDCOPY, suppress prompt-window message
* COLLECT? argument to TEDIT.MAP.OBJECTS can specify what kind of information to collect.
- Fixed INTERPRET.REM.CM so that it no longer tries to load the file pointed to by LDEINIT and instead loads the file pointed to by LDEREMCM.
LDEINIT remains the file used by greet.
- Adjusted the medley script to have a new argument -cm (or --rem.cm) which sets LDEREMCM as appropriate before launching lde.
- Updated the loadup scripts as required to use this new -cm argument when calling medley.
Finally, added a new feature to the medley script -cc (or --repeat) whereby which when medley finishes it checks for a nonzero file given as the argument to -cc. If that file exists, medley is run again (i.e., repeated) with LDEREMCM set to that file. This repeats until this file no longer exists or is zero-length. The file can be found as the vale of LDEREPEATCM so that each invocation of medley can modify (or delete) this file so as to change the subsequent run of medley.
* Fix up Move command
* change paralooks record, recompile, other updates
* Fix right arrow and EOF
* Caret position after NEXT, add FILESTREAM textprop, CH#/LEN args to TEDIT.SEL.AS.STRING
* Added LLK
* Arrows-movement in menus is confined to fields
* Slightly better menu test
* TEXTPROP glitch, create window after getting file
* remove a debug call
* forward/backward forgets original x position
* Opentext stream preps for window-region prompt width
* typing resets the leftarrow X position
* Removing cached X position for up and down arrows
I tied this to the DIRTY flag and recompiled its users. Also separately fixed the setsel and find functions to clobber the cache
* Fencepost glitrch
* add OPENWIDTH property
* Selection past the middle goes to the right
* bug fix
* Set up for paralooks name changes to avoid future confusions
* Update tedit-exports.all
* One more systematic name change for char/para looks parallelism
* TEDITSTRING applies initial props
* TEDIT.CARETLOOKS returns oldlooks, TEDIT.PUT has a QUIET flag
QUITEFLAG suppresses prompt printing
In github actions, remove build of Medley Docker for arm64 since the Dockerfile_medley build fails when loading tigervnc-standalone-server for arm64. Arm64 docker image has never been used anyway. It was meant for running online on arm64 servers, but we have never done that and have no plans to do so. So just getting rid of arm64 docker image rather than trying to fix it.
Enable **FONTSAMPLER** to display glyphs from **.DISPLAYFONT** (bitmap
font) on non-`DISPLAY` stream (e.g., **PDF**)
Change to match 2025-02-03 discussion on Ascent/Descent
per-`CHARSETINFO` _vs._ whole `FONTDESCRIPTOR`.
(Use whole `FONTDESCRIPTOR`.)
* Add back character sets that had characters outside 16 bit plane
* Update XCCS-353=SYMBOLS3.TXT
Update title line
* Update UNICODE.TEDIT
* Fix charset names
* Reorganized the tables, added requested interfaces
* Use a single hash
* Top-level array branch beats a single hash
* cleanup UNICODE.TRANSLATE macro
* Fix slug in outcharfn
* Remove a stray line
* Another try, would work for raw
* Remove duplicates, redo hashing
* Getting complete maps in both directions
* Initializing
* Only the latest file versions
* Add back gothic mappings
* Enable FONTSAMPLER to display glyphs from DISPLAYFONT (bitmap font) on non-DISPLAY stream (e.g., PDF)
Added .LCOM to repository.
Corrected PR.
* Relocate FONTSAMPLE files to obsolete.
* Files renamed. Internal names and documentation were NOT updated.
---------
Co-authored-by: rmkaplan <ron.kaplan@post.harvard.edu>
These are macros that I have been using for years to simplify the
storage and retrieval of items in alist structures with arbitrary
numbers of keys. They may prove useful to others.
* Add back character sets that had characters outside 16 bit plane
* Update XCCS-353=SYMBOLS3.TXT
Update title line
* Update UNICODE.TEDIT
* Fix charset names
* Reorganized the tables, added requested interfaces
* Use a single hash
* Top-level array branch beats a single hash
* cleanup UNICODE.TRANSLATE macro
* Fix slug in outcharfn
* Remove a stray line
* Another try, would work for raw
* Remove duplicates, redo hashing
* Getting complete maps in both directions
* Initializing
* Only the latest file versions
* Add back gothic mappings
* Add a call to check for orphaned versions after any checkout
* use == instead of -eq for optional
* Add a scripts/install-repo-checks for things to run after checkout; only this versioning error checked for now
The files were **copied** from `lispusers` to `library`, and then
modified there.
They should have been **moved** so the git history was preserved.
I replaced these files _with themselves_ by a 3 step process that
_appears_ to have gotten the git history to be correct.
(And the `PSEUDOHOSTS.TEDIT` from `lispusers` was also moved. It had
been left behind previously.)
* Addresses #1943
* Better prompt messages for meta-O and meta-D
* Cleanup TEDIT.SETSEL
* Meta-O shows menu of types to edit
* Extra variable
* More cleaning of TEDIT.SETSEL
Start it running with `(START-TAB-WINDOWS)`.
Bonus debug tool: `(KEY-WINDOW)` starts a process that monitors keyboard
and mouse button events and displays them in a little window.
* Shakedown of field menu items
* Suppress shift-select in menus
* Typo, plus inserting EOL-containing strings
* Fix selection display glitch revealed by DOCUMENT
* Fix arg order, eliminate U-CASE
* U-CASE only if coercing IDENTIFIER from LABEL
* Allow SMALLP for identifiers, INITSTATE for SELECTION
Also a fence-post glitch in paragraph selection
* Fix MB.GET
* TEDIT-MENU: Right button doesn't invert Marginbar, copying suppressed
* TEDIT-STREAM: Error if copying an image object that doesn't allow copying
* Missed another LITATOM/SMALLP test on IDENTIFIERS
* Clicking in a menu removes stale promptwindow text
---------
Co-authored-by: Frank Halasz <frank@halasz.org>
* FILEIO: EOL for input defaults to ANY, EXT-FORMAT can specify EOL
As per technical meeting on 7/15/2024
* Revert "FILEIO: EOL for input defaults to ANY, EXT-FORMAT can specify EOL"
This reverts commit 6a7e8c3665.
* FILEIO: Fix comment
* Added DETECTEDEOLCONVENTION to STREAM declaration
and recompiled calls to macro \CHECKEOLC.
* COMAPARETEXT: was trying to set EOL to ANY on a Tedit stream
* LCOMS needing to be recompiled for \CHECKEOLC macro and Create STREAM
(plus a new (unchanged) version of IOCHAR needed to get the cleanup to work for the recompile)
* EXTERNALFORMAT macro and function implement EOL detection
* FILEIO: stream records detected EOL, also RENAMEFILE uses COPYBYTES
UFS doesn't check file devices identity, doesn't give type-change message. Recompiled for create stream
* ADIR has TRUEDEVICE
* Revert "FILEIO: stream records detected EOL, also RENAMEFILE uses COPYBYTES"
This reverts commit fa97aa6157.
* Revert "EXTERNALFORMAT macro and function implement EOL detection"
This reverts commit eb098615ed.
* Revert "LCOMS needing to be recompiled for \CHECKEOLC macro and Create STREAM"
This reverts commit 5967452c63.
* Revert "Added DETECTEDEOLCONVENTION to STREAM declaration"
This reverts commit 196f105cf5.
* Trying to complete the ANY/EOLC and binary RENAMEFILE issues
* loadup glitch
* TMAX updates for compatibility with Tedit changes
* DOC-OBJECTS changes for compatibility with Tedit changes
* MODERNIZE update for Tedit split windows
* Core Tedit files
* IMAGEOBJ: Remove dependency on Tedit internals
* WINDOW: Remove dependency on Tedit internal declaration
Still strange that WFROMDS should have to branch on Tedit
* WINDOWOBJ gets window of TTY process before the window of the stream of the TTY process
So insert into Tedit works
* TEDIT-CHAT: try to use TEXTSTREAM vs TEXTOBJ
* Fix tab-initialization problem in SLIDES.TEDIT
as reported by @nbriggs
* TEDIT-CHAT: use TSTREAM rather than TEXTOBJ
* Updates after lots more testing, particularly scrolling
Some other files dragged along to avoid dependence on Tedit internals
* Remove unwanted SAVE.SYSOUT
* Addresses more end-of-file and empty-file display issues
Try it again
* TEDIT-DEBUG tracking other changes
* Odds and ends
* Adjust EOF selection and caret-scrolling on copy
* More cleanup, plus fixing a few more ancient (Venue) glitches
As usual, the problems have to do with the funky behavior of EOL's in the middle and end of the document. More abstraction and refactoring to get better control of this (I hope).
* TEDIT-WINDOW: Scroll down of big objects
Trying to fix what happens at the transition when scrolling down brings a big-object's top down in the window. Approach is to bring down the line above, which may make for a little jump. I hope that solves it.
Scrolling up still needs some adjustment.
* Eliminate junk at top of window after up/down scrolling of big objects
BLTSHADE is OK there for scrolling, but not for redisplay after editing. In the edit case, the top of the pane above the last valid line is preserved.
Scrolling still has the problem that the window can go blank at the first scroll that brings a tall object into the pane--still working on that.
* Scrolling with tall lines should be more continuous
* Another tweak for scrolling
plus interface extension to TEDIT.MOVE and TEDIT.COPY, a little more on field menus
* More robust strategy for field menu buttons
Surround the field with prefix and suffix pieces with image objects that print the pre and post labels and shift the selection forward or backward into the field. Doesn't depend on inherited quirky logic in the selection line-scanner.
* Field selection ignores right and middle clicks
* A little more menu/selection tweaking
You can't extend through fields and buttons
* Added CUSTOMBUTTONEVENTFN to menu field buttons
Also, menu buttons in general can't be deleted
* A few more glitches, plus a little selection refactoring for buttons
* Reduce flicker in pargraph menu margin bar
* screen update glitch
* DOC-OBJECTS, TEDIT-SCREEN: Fixes the HCFILES DOC-OBJECTS failure
* TEDIT-BUTTONS: Field values should always be shown in the specified FIELDFONT
* Abstracting the structure of the history lists
cleanup, but mostly as a precursor to maybe doing a ring buffer of a specified length
* TEDIT-FILE, a little font-reading cleanup
* Include the files from rmk-39 that deal with the text/binary renamefile problem
* TEDIT-PAGE addresses #1905
* Fix BUTTONSTART to STARTPC in Put/Get menu buttons
* Use width of M as width of EOL--easy to select
Also put in function call for potential kerning--needs eventual FONT support
* Rename a few internal functions from TEDIT.-- to \TEDIT.--
* Doesn't make sense for a charlooks to not have a font
* TEDIT-BUTTONS - Fix comment
* Take out Tedit internals from \CARET.FLASH?
Should have included this in fifth round long ago
* TEDIT-FILE: use DEFAULTFONT for .sh files
Easier to follow the layout
* TEDIT-LOOKS: fix loadup order
* tedit-exports.all Remove line-has-protection field
Useless
* Make sure that charlooks change as expected
* External format for .sh files is UTF-8
* Better display of history information for debugging
* Button changes: show document font families, better fields
* TEDIT-WINDOW, remove extra truncated line with down-scroll
* TEDIT-SELECTION: suppress line/paragraph selection for built-in menus
Line/para selection would be reasonable for multi-line fields, but most menu lines have protected text that would behave inconsistently. So just suppress
* Better support for potential kerning
* Fix empty field value
* Simplify ASCII translation code
* Make sure headings have a default tab
* Word boundary at character 1
* Remember that you specified a font class instead of a font
For the charlooks menu, but also so that it is saved on a put
* More items on the Family NWAY-button line
* glitch
* TEDIT-LOOKS: Better algorithm for Ascii translation
* Charmenu remembers previous "Other" fonts, even if not installed
* Fix initial piece index
* Fix fontclass changes (again)
Start it running with `(START-TAB-WINDOWS)`.
Bonus: `(KEY-WINDOW)` starts a process that monitors keyboard and mouse button
events and displays them in a little window.
2024-07-19 10:59:52 -07:00
1641 changed files with 63596 additions and 47594 deletions
First, we want to thank you for helping reach the goal of restoring Medley Interlisp
to the point where it is as useful today as it was 30 years ago.
This guide is meant to help you make useful contributions, whether to the [Maiko](https://github.com/Interlisp/maiko) C-based virtual machine implementation, the [Medley](https://github.com/Interlisp/medley) Lisp code (in Interlisp and Common Lisp), or [documentation](https://github.com/Interlisp/medley/wiki). There are a number of [GitHub](https://github.com/Interlisp/medley/discussions/categories/github-use) problems that could use some attention.
This guide is meant to help you make useful contributions, whether to the [Maiko](https://github.com/Interlisp/maiko) C-based virtual machine implementation, the [Medley](https://github.com/Interlisp/medley) Lisp code (in Interlisp and Common Lisp), or [documentation](https://interlisp.org/software/using-medley). There are a number of [GitHub](https://github.com/Interlisp/medley/discussions/categories/github-use) problems that could use some attention.
## Working with Maiko
@@ -17,6 +17,8 @@ The current arrangement of files and extentions is awkward for working on the im
* The most useful contributions are reproducible errors -- things that don't work as documented.
* Second most useful are reports of unexpected behavior -- things that aren't documented but behave unexpectedly.
To report these errors and behaviors [open an issue](https://github.com/Interlisp/medley/issues).
## Reporting a bug or feature request
* Ensure the bug was not already reported by searching on GitHub under [Issues](https://github.com/Interlisp/medley/issues) or [Discussions](https://github.com/Interlisp/medley/discussions). Note that all issues and Discussions are found in the Medley repository, using labels to distinguish. Discussions are for questions or topics where there is some disagreement or uncertainty about the "right" direction.
* If you're unable to find a discussion or open issue addressing the problem, open a new one. Be sure to include a title
@@ -24,7 +26,7 @@ and clear description, as much relevant information as possible. Use the issue t
## Did you write a patch that fixes a bug?
* Some bug fixes and "improvements" have unintended consequences, well beyond what you might expect for well-written modern code. We don't have testing new builds automated or integrated. Be sure you've tested your patch.
* Open a new [GitHub pull request](https://github.com/Interlisp/maiko/pulls) with the patch.
* Open a new GitHub pull request with the patch.
* Ensure the PR description clearly describes the problem and solution. Include the relevant issue number if applicable.
* Keep Pull Requests small and easily reviewable. https://www.thedroidsonroids.com/blog/splitting-pull-request for
This repository is for the Lisp environment of [Medley](https://interlisp.org).
The [Medley Interlisp Project](https://interlisp.org) aims to preserve, revive, and modernize the [Interlisp](https://interlisp.org) software development environment for rapid prototyping, research and Artificial Intelligence created at Xerox PARC since the 1970s.
[Install and Run](https://interlisp.org/software/install-and-run) covers ways to install and start up Medley on Linux systems, MacOS, and Windows (with or without WSL).
This repository is for the Lisp environment of the [Medley](https://interlisp.org) release of Interlisp. Other repositories hold additional subsystems and applications such as [Maiko](https://github.com/Interlisp/maiko), the implementation (in C) of the Medley virtual machine, the [LOOPS](https://github.com/Interlisp/loops) object-oriented extension of Interlisp, and the [NoteCards](https://github.com/Interlisp/notecards) hypermedia system.
[Install and Run](https://interlisp.org/software/install-and-run) covers ways to install and start up Medley on Linux systems, MacOS, Windows (with or without WSL), and in a web browser.
[Using Medley](https://interlisp.org/software/using-medley/) has an overview and pointers to documentation.
[Interlisp/maiko](https://github.com/Interlisp/maiko), is the repo for the implementation (in C) of the Medley virtual machine.
The [Glossary](https://interlisp.org/history/glossary) defines system-specific terms such as "loadup" and "sysout".
## Releases
@@ -67,25 +69,9 @@ If you have a high-resolution display, note that much of the graphics was design
Medley presumes you have a 3-button mouse; the scroll-wheel on some mice acts as one, with some difficulty. Go into XQuartz Preferences/Input and check "Emulate three button mouse" option.
### Running Medley Interlisp (obsolete)
### Running Medley Interlisp
The `run-medley` script in this repo sets up some convenient defaults. Running Medley can be done by typing:
```
$ cd medley
$ ./run-medley
```
Or, if you wish to start Medley up with a different SYSOUT:
```
$ cd medley
$ ./run-medley <SYSOUT-file-name>
```
The first time the system is run it loads the system image that comes
with the system. When you exit the system (or "do a `SaveVM`" menu
option) the state of your machine is saved in a file named
`~/lisp.virtualmem`. Subsequent system startups load the
`~/lisp.virtualmem` image by default.
The primer [Medley Interlisp for the Newcomer](https://primer.interlisp.org) eases new users into the Interlisp environment. It assumes no prior knowledge of Lisp and covers the user interface, programming and debugging, windows and graphics, and more. We recommend consulting this document to learn how to run and use the system.
### Exiting The System
@@ -123,7 +109,7 @@ Each directory should have a README.md, but briefly
* BUILDING.md -- instructions on how to make your own loadups
* clos -- early implementation of Common Lisp Object System
* CLTL2 -- files submitted to bring Medley up to the conformance to "Common Lisp, the Language" 2nd edition. Not enough to conform to the ANSI standard lisp.
* docs -- Documentation files (in TEdit format, PDFs, or online help; look [here](https://github.com/Interlisp/medley/Documentation))
* docs -- Documentation files (in TEdit format, PDFs, or online help
* fonts -- raster fonts (or font widths) in various resolutions for display, postscript, interpress, press formats
Some files were not shown because too many files have changed in this diff
Show More
Reference in New Issue
Block a user
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.