1 -*-outline-layout:(2 (-1 -1 0 :) 0);outline-stylish-prefixes:nil -*-
4 Features you cannot find in the doco as working, should be
5 mentioned here. This is an assorted collection of stuff that will be
6 done, might be done, or is an idea that I want to think about
8 Most of the items are marked in the code as well
9 Grep for TODO and ugh/ugr/urg.
12 . * junk separate mudela versioning.
14 . * header for PS enteredby = "bla <bla@bar.com>"
15 . * ps/lily.ps see comments.
16 . * devise standard for functioning of Vertical_align_element.
19 . * fix singleStaffBracket
20 . * declare performers in \midi
22 . * \shape 1st dim skipped?
23 . * turn slope-damping on/off
25 . * We need feta-din*.mf files for more sizes than 10.
26 . * fix dynamics decently, ie. use kerning & ligatures.
27 . * dots & rest collisions?
29 . * fix \mudelafile{} to have decent functionality
30 . * P.P.S. It can be cool in mudela-book to distinguish in
31 pre,postMudelaExample, whether MudelaExample is epsed or not: ( if
32 this fragment is floating eps, than 1, otherwise 2). say
33 preMudelaExample[eps]{}, and change it in document body sometimes.
35 . * slurs colllide with lyrics.
36 . * Is there any reasonable way to do this: I would like to have one file
37 that can generate the same tune in two different modes without having
38 to go through and rewrite all the music in the other mode.
40 Long lyrics don't get a space at the end. They just run right up
41 against the next lyric. (I don't think this used to be the case.)
43 Lyrics and scripts collide.
45 \key b; \time 4/4; looks bad. The clef and time signature are too close.
47 restStyle=mensural doesn't seem to work (lots of characters not found
49 . * Phrasing: Indicating the textual underlay of a piece, or
50 otherwise dividing a piece into phrases.
52 There are distinct ways of specifying the first two (although at
53 present, they're not properly distinguished) but no distinct way to
54 specify the third. Should there be?
55 . * rest collision minimum dist.
56 . * 1. is there a "no key signature mode" for lily yet so that i can just
57 enter the notes without worrying about it instead of having to
58 remember to put a "!" after each accidental?
60 a note with a circle after it, meaning:
62 note + circle = note + 1/4 of its length
64 the circle is like a dot that's not filled in. for example, on
65 page three, the c-sharp on the second line should be a quarter with
66 a circle, not a quarter tied to a sixteenth. is this at all
68 tuplets that are written as "number : note", for example
69 "3 : [image of a half note here]". possible?
70 "p+" and "p-" for dynamic markings. possible?
71 . * repeat engraver, gallina.ly
74 . * spacing for prefatory matter
75 . * The line length is 7.5 inch (19.05
76 cm) and the height of normal size is equivalent to the traditional raster 1
77 > size. (The Schott company sells Letraset sheets of musical symbols which are
78 > measured in raster sizes). It is agreed among most engravers I know that the
79 > standard size for adult readers in solo music should 70% of raster 1. Jeff's
80 > example was 75%, so I had to modify my file to match.
82 So that makes for a staff space of 4 2/3 point. I'll add the
83 dimensions to the TODO file.
84 . * accidentals closer to note head
85 . * to space the staffs/staffgroups contained in the score.
87 In the orchestscore.ly file, I tried:
90 \OrchestralScoreContext
91 minVerticalAlign = 3.0*\staffheight;
95 minVerticalAlign = 3.0*\staffheight;
99 minVerticalAlign = 3.0*\staffheight;
108 . * Rhythmic_grouping
110 . * Spring generation
116 . * Morally pure LilyPond.
117 . * Remove mutopia ramblings.
118 . * Remove non-free software links.
119 . * Remove meta article LilyPond.
120 . * break priority setting from SCM.
122 . * remove [] in favour of auto-beamer
123 . * hangOnClef with non-breakable clefs.
124 . * Aha, the problem is the new TeX code generation for changing fonts.
125 > (The explanation below is mostly intended for Han-Wen and Jan)
126 > Before, the font was changed with macros declared separately for
127 > pure TeX and LaTeX, but now all the fonts are declared directly in
128 > the generated code using the low-level TeX commands. This circumvents
129 > the fontencoding mechanism of LaTeX and is a potential source of
130 > problems also for us users of the Latin-1 character set.
131 > I think it still works since the inputencoding latin1 is
132 > declared by ly2dvi, but it still means that e.g. an o with
133 > umlaut is typeset as a combination of the two symbols 'o' and
134 > 'umlaut' instead of as a single character, as would be the case
135 > if fontencoding OT1 had been used. This shouldn't give any
136 > difference in layout, though. One of the main argument for using
137 > fontencoding OT1 is for the hyphenation mechanism to work, an
138 > argument that's clearly irrelevant in our application. Yet we
139 > see the problem for August.
140 . * Abstraction for engravers:
141 . * make "in-between" engraver (Tie, Extender)
142 . * make wide_spanner_engraver (line_group_spanne,r staff_symbol)
143 . * remove Interval dim_ from Dimension_cache and rename the struct.
144 . * do scaled fonts generally
145 . * The Lilypond accidental symbols are available through the TeX macros,
146 \textflat, \textsharp and \textnatural defined in the init file
147 lilyponddefs.tex. All the symbols are defined in the files fetaXX.tex
148 where XX corresponds to the size. I don't know where (or even if) they
149 are installed but they are generated in the mf/out/ directory during the
151 . * make dependencies () a virtual function.
152 . * one big ly2dvi input that includes *all* tests.
153 . * store req -> element, element -> element causes for warnings
154 . * include examples in RPM.
155 . * fix partial measures in meaningful way.
156 . * add scripts to bars eg. |^"bla"
157 . * relative mode for mi2mu
158 . * uniformise recent feta contributions.
159 . * use SCM and Molecules for {accordeon stuff, }
160 . * bigger and fatter 4/4 C
161 . * sort out directory stuff.
163 . * -I option per filetype?
165 . * ly2dvi/lilypond logfile.
166 . * fractional chord durs.
167 . * hang Item on Spanner
168 . * do --safe for PS output?
169 . * convert-mudela --output
170 . * Staff_margin (with a partial measure.)
171 . * fix pletvisibility properties:
172 . * bracket on/off/no-bracket-when-beam
173 . * num on/off/no-num-when-beam
175 . * junk text{sharp,flat,etc}
176 . * fix convert-mudela manpage
177 . * decimal point in \paper {}
178 . * { \voiceone <a'4.*2/3 d''4.*2/3 fis''4.*2/3> [g''8 fis'' e''] d''4 |}
180 { \voicetwo d'4 d'2.*2/3 } spacing
181 . * tied notes for MIDI
182 . * sharp / flat on trills (in MIDI ??)
183 . * scm-ify \property values.
184 . * move class Lookup {} into scm
186 . * \breathmark TeX macro
187 . * add new glyphs to font.ly
188 . * formatting of input stuff.
189 . * \notes{ a \< b \cr } vs \notes{ a \< b \! }
190 . * space after bars?
191 . * 'hinterfleisch' before bar (e.g. wtk1-fugue2)?
192 . * Summary of minor spelling irregularities:
193 . * capitalization/use of underscores in property names
194 . * fix SkipBars -> skipBars
196 lbheel = \script { "bheel" 0 0 -1 0 0 }
197 rbheel = \script { "bheel" 0 0 1 0 0 }
198 lbtoe = \script { "btoe" 0 0 -1 0 0 }
199 rbtoe = \script { "btoe" 0 0 1 0 0 }
200 lfheel = \script { "fheel" 0 0 -1 0 0 }
201 rfheel = \script { "fheel" 0 0 1 0 0 }
202 lftoe = \script { "ftoe" 0 0 -1 0 0 }
203 rftoe = \script { "ftoe" 0 0 1 0 0 }
205 portato= \script { "portato" 0 -1 0 1 0 }
207 . * bottomnote for ly2dvi
210 . * compulsory hshift
211 . * Give # lines in linebreaking.
213 . * rewrite context defs using Scheme
214 . * \selectmusic to cut pieces from music.
215 . * TODO: merge atom & molecule; make tree from molecule.
216 . * Align_element::padding ?
217 . * uniformise property names...
218 . * ydirection <-> yDirection
219 . * rather allow '_' in identifiers first (i.e. junk ^ and _),
221 . * typo checks on property names?
222 . * use streambufs and iostream
223 to provide IO handling for TeX stream, mudela stream, data-file.
224 . * seriously buffer TeX output (do profile of writing .5 mb TeX file.)
225 . * strip EXEs before installing
226 . * zip target for binary windows dist (JBR)
227 . * junking \skip req in lyrics
228 . * mi2mu empty staffs.
229 . * use Fourier transformation for rhythmic quantisation.
230 . * horizontal centering of dynamics
231 . * $DEPENDENCIES_OUTPUT support
232 . * Xdvi zooming ?! Try to understand GS garbage collection.
233 gs: Error: /undefined in draw_beam
235 . * fix vertical alignment and stafflines
236 . * declaring Performers
237 . * *.yo: fix pod manpage layout legacy
238 . * text-items clash with stems/beams
239 . * --include, -I option for ly2dvi (pass on to lily)
240 . * fix placement of beam-less abbrev
242 <{\voiceone cis4. cis8 cis4 | cis4 cis cis | r1 }
243 {\voicetwo cis,4 r r | r2. | r1 }> |
245 . * Musical_pitch (analogous to Duration and Rhythmic_req)
246 think about, analogous to pitch:
247 * { a4 a16 } c <-- duration of c?
248 * < a4 b8 > c <-- duration of c?
250 . * fix audio-items (don't use (input) requests, but copy data)
251 . * make sure all ex's have a mudela-version
252 . * do rest-collisions for Multi_measure_rests as well.
253 . * split error/warning in error/warning/non_fatal_error
254 . * add a Duration_convert member to Duration_iter to set parameters.
255 Junk global duration settings.
256 . * en-,discouraged linebreaking:
257 . * handle DISALLOW < penalty > FORCE
258 . * discourage breaking of slurs
259 . * fix variable define/lookup parser global<->paper
260 . * fix title/instrument placements; see input/test/title.ly
261 . * <\voiceone c4. \voicetwo c4>
262 . * add full cello suites to web site (kom op, Maarten!)
263 . * add mudela-book-doc source to webstie.
264 . * Rethink Time_description
265 \cadenza , \meter, \grouping should all be \properties
268 . * abbrev -> tremolo
269 . * abbreviation-beam -> (measured/unmeasured) tremolo
270 . * partial -> anacrouse
271 . * robustify Beam (in case of no stems, 32nd notes)
272 . * move paper vars into engraver properties
273 . * check for groff / troff/ nroff et
274 . * more intelligent file searching
275 . * disable spaces in TeX stuff
276 . * handle ^C for tmp/file creation.
277 . * make LilyPond RPM fully relocatable
278 . * dots & rest collisions.
279 . * the warning about the negative slur/tie length appears twice
281 . * better hshift (dots, distance, head dependent)
282 . * clefs (AG): The "8" should appear closer to the actual clef,
284 . * put errorlevel in Input class
285 . * junk nesting slurs
286 . * integrate midi-elts from mi2mu and lily?
287 . * distinguish between unset and empty properties.
290 . * the bracket is ugly (wings are too parabolic, should be more circular)
291 . * versioning for Feta
292 . * rewrite dynamic signs. They're kind of spaghetti now.
293 . * The `3' in the meter key is a one or two pixels too thin (at 600dpi)
295 . * design macro for penstroke with rounded endings
296 . * timesig C: fat yet less curved. Check out relations.
297 . * design decent macro for bulbed endings
298 . * printer dependencies.
301 . * work out sizes of character fonts.
303 . * piano pedals: Ped/*
305 . * coda signs: Segno, O+
306 . * glissando, arpeggio
308 . * include important spacing dims in fetalog
309 . * scalability: use blot_diameter everywhere, make font chubbier for
313 . * GNU diff 2.7: diff -rN does not see a new directory with empty file
314 . * check out GCC signatures?
316 f = fopen ("/dev/null", "r")
319 . * tetex: mfplain.mem
320 .* 3RD PARTY PROJECTS:
321 . * make GCC warn about ctor that leaves member vars uninitialised.
323 . * 'double-fix' mode: ignore identical fix, rather than suggest
324 to revert patch direction when (more or less? exactly) the same
328 . * stack scripts in chords:
329 % fingering: footing:
330 < a-4 c-2 g-1 > < c-\ltoe e-\lheel >
332 . * slur parts as a script
333 . * stack parameterised scripts for fingering and footing:
336 . * automatically add a \version if not present.
339 . * convert-mudela -e empty.ly
341 . * emacs mudela-mode
342 . * should handle block comments too.
343 . * handle lexer modes (\header, \melodic, \lyric) etc.
346 . * fontlock: \melodic \melodic
348 . * use properties for:
354 . * optimal pagebreaking.
356 . * Unicode support? -> man 7 unicode
359 . * detect of 16/8 bit files
360 . * examples to go with it.
363 . * write a faster Spring_spacer (without matrices if possible)
364 . * relate energybound to linelen unitspace fontsize etc.
365 . * used fixed point fp?
368 . * output header info to MIDI too.
370 . * a musical dictionary. See Documentation/vocabulary-*, other
380 . * guitar chords (fret diagrams)
381 . * other chord name styles (american(?) style)
386 . * derive lily_stream, Dstream, texstream from ostream?
387 . * indentable stream as baseclass for TeX stream, lily stream, Dstream.
388 . * use dstream feature in mi2mu
389 . * integrate IO checking into stream class
396 . * add to MIDI output:
398 . * ornaments (and trills?)
402 > Would it be hard to add support for proper dynamics in MIDI output? I
403 > would really like to have this feature...
405 To answer this, I have to give a little background. Please say so if
408 Both audio and notation elements are created decentrally, when
409 LilyPond says "Interpreting music ... [8][16] etc." After all
410 elements are created, they are processed again to calculate all
411 parameters. After that is done, they are dumped onto the output (the
412 MIDI stream or the TeX file).
414 The audio and notation elements are created by objects called
415 performer and engraver respectively. They work by taking a chunk of
416 input (a so-called request), and using the request to create an audio
419 After some processing, the engraver sends the created elements to
420 its enclosing notation/performance context.
422 In the case of notation, engravers can also modify existing elements
423 created by sibling engravers. Eg. the beam-engraver will modify stems
424 when it creates a beam over a couple of stems. This is done in
425 Engraver::do_acknowledge_element () and
426 Engraver::do_process_acknowledged ().
428 In the case of audio, there are no mechanisms to modify existing audio
429 elements. After the audio elements are created, they are sent directly
430 to the enclosing context (an instance of Performer_group_performer),
431 through the method Performer::play ()
433 Concretely spoken, for dynamics you have to write a performer
434 (Dynamics_performer), that will take the input requests. If I
435 understand correctly, every MIDI note played must have its dynamic
436 strength set separately. That means the the Dynamics_performer must
437 set the strength of every Audio_note it finds. This means that one
438 has to modify existing Audio items. The best way of doing this is
439 with a broadcast/acknowledge process.
441 So the best way of handling this, is
443 1 supporting dynamic settings in Audio_note
445 2 mimicking the broadcast/acknowledge mechanism of the Engravers in
448 3 using that mechanism to write a Dynamics_performer that will modify
449 any notes it finds to set appropriate strengths.
451 You could also kludge this by deriving from Performer_group_performer
452 a performer that also sets dynamics within notes, but that is not the
453 elegant way to do it.
455 I'd say that the work involved is not hard, but you have to be fluent
456 with C++ and need some insight into the working of Notation Contexts
457 and friends. I guess I could do it in a day or so, but if you are not
458 so fluent with the inner workings of LilyPond, it could take you some
459 more time (A few more days?).
461 Please note, that I am *not* going to code this myself: my days are
462 valuable, and I think I have more interesting things to do than fixing
463 the MIDI output (I don't even have a soundcard). But I *do* want to help
464 you or other people with implementing this, so if you need help, don't
467 . * account for rhythmic position in measure
471 . * make separate class for Grace_*_engravers
472 (which is at Voice_gravs level)
473 . * make encapsulated spacing problems.
475 . * accidentals per Voice_group
477 . * Output class, handles : (smallish)
478 . * help text /(c) notice?
480 . * warning /errors/progress
481 . * abort on error/warning; exit status
482 . * quiet/ignore-version options
486 . * speed up PS code.
487 . * PS lines to .eps files
488 . * write custom prolog
493 . * remove links to non-free SW.
494 . * internal documentation
497 . * left/rightbeamcount
498 . * various other new properties.
503 . * Doc (C) stuff of sheet music
504 . * a better format? SGML? TeX? Info?
505 . * a better tutorial
507 . * more spanners (smallish)
511 . * Rewrite Stem, Beam and Rhythmic_grouping (hairy)
513 . * doc on Rhythmic_grouping
514 . * general shaving on Beam and Stem
515 . * use plet grouping
516 . * separate visual from rhythmic info even more
520 . * Ledger lines, should be separate item: Ledger_lines
521 . * set different line sizes / thicknesses
522 . * beam stem showthrough, inter beam sizes (french vs german style)
525 . * left/right note balls should be handled by Collision:
526 < \multi 2; { \stem 1; <b c> } { <f as b> } >
529 . * merge key{item} & localkey{item}?
530 . * spacing, national styles?
533 . * redo Notehead to allow digits as heads
534 . * add a \tablature {} lexer mode for easy entering
536 . * should adjust stemlength for flag number.
539 . * caching breakpoints
541 . * specify number of lines
544 . * \rhythms 4 16 16 16 16; c c c c c -> c4 c16 etc.
548 . * \bla {} vs \bla ;
549 . * mix engraver hacking with music ?
550 . * \once\property KEY = VAL
551 . * \addtranslator, \removetranslator
552 . * junk ^ and _ for scripts
553 . * junk _ for lyrics.
554 . * abstract grammar.
555 . * Figure out semicolons.
556 . * c4 4 4 4 for c4 c4 c4 c4?
567 . * configure pitch_byte
568 . * rest name configurable
573 . * Faulenzer sign (% repeat). Perhaps also use \repeat syntax?
575 . * fix confusing naming in Engraver vs Engraver_group_engraver (context)
577 . * progress when creating MIDI elts.
579 . * A range for context errors (eg. mark both { and }. )
581 . * lyric in staff (sharpsharp in staff, text below)
583 . * half sharps/flats
585 . * write Dynamic_line (to group dynamics horizontally)
587 . * use Real for all y positions.
589 . * half-sharps, half-flats
591 . * adaptive accidental spacing.
593 . * handle EOF graciously in error messages.
595 . * midi esp.: use I32 iso int where 32 bits are needed (or assumed...)
597 . * stafftypes: voice names/ instrument names.
601 . * % toe to bheel four to five
607 (where "to" is a tiny bow)
609 . * auxilliary file for caching info.
616 . * Output data structures of Mudela in XML/SGML.
617 . * create libmudela, or liblily_frontend
618 . * move MIDI stuff (including Quantization) to a ANSI C libmidi library.
619 . * Spacing_request for manually adjusting spacing
620 . * caching breakpoints
621 . * used fixedpoint arithmetic for min. energy.
622 . * move towards incremental algorithms.
623 . * versioning stuff (cvt mudela, mudela, etc.)
624 . * dynamic loading of engravers?
626 Han-Wen Nienhuys <hanwen@cs.uu.nl> writes:
629 > Foo_smob *smob_ptr_;
633 > Foo_class *class_ptr_
636 Usually you can get away without having this interlinked structure.
637 One piece of information you need when exporting objects to scheme is
638 the smob tag for your class. You might store this value (a long) into
639 a class static variable, or a file static variable. I'll use a class
640 static variable in this example.
642 I typically use code that works like this (untested code ahead):
645 static long *Foo_smob_tag;
646 SCM obj; // representation as a scheme object, description comes later
649 // call this once on startup
650 void init_Foo_class {
651 static scm_smobfuns type_rec;
653 type_rec.mark = foo_mark;
654 type_rec.free = foo_free;
655 type_rec.print = foo_display;
657 Foo_class::Foo_smob_tag = new int(scm_new_smob(type_rec));
660 When you want to export an object to scheme, you put a pointer to the
661 object itself in the cdr of the cell, and the tag in the car of the
662 cell. The code looks like this:
668 return(SCM_NIMP(obj) && SCM_CAR(obj) == Foo_class::Foo_smob_tag);
671 // given a Scheme representation, return, a C++ representation
675 return((Foo_class*)SCM_CDR(obj));
678 // the SCM representation of the object is stored inside itself
679 // this will become useful when destructor gets called
681 foo_box(Foo_class *foo)
684 SCM_NEWCELL(foo->obj);
685 SCM_SETCAR(foo->obj, Foo_class::Foo_smob_tag);
686 SCM_SETCDR(foo->obj, (SCM)foo);
691 > C++ determines life time, so what is the strategy?
693 What happens now when the destructor gets called? Lets set the cdr to
694 NULL, and then check for that in all of the primitive functions that
695 use your smob. We'll call this notion 'live'; a scheme object is
696 'live' if the C++ representation still exists, it's dead if the C++
697 object is gone. You can still have references to dead objects in
698 scheme code; it's just an error to use them. This idea is stolen from
699 Emacs's handling of buffers ('buffer-live-p' and friends).
701 Add another function, foo_live_p:
706 return(foo_p(obj) && SCM_CDR(obj) != NULL);
709 In you destructor, you need to do:
713 SCM_CDR(this->obj, NULL);
716 When writing new primitives, now just check foo_live_p().
718 I hope the above helps,
724 (defvar outline-font-lock-keywords
725 '(;; Highlight headings according to the level.
726 ;; must change this to allout node layout
727 ;;("^\\(\\*+\\)[ \t]*\\(.+\\)?[ \t]*$"
728 ("^\\([.]*[ ]*[*]\\)[ \t]*\\(.*\\)$"
729 (1 font-lock-string-face)
730 (2 (let ((len (- (match-end 1) (match-beginning 1))))
731 (or (cdr (assq len '((1 . font-lock-function-name-face)
732 (2 . font-lock-keyword-face)
733 (3 . font-lock-comment-face))))
734 font-lock-variable-name-face))
736 ;; Highlight citations of the form [1] and [Mar94].
737 ("\\[\\([A-Z][A-Za-z]+\\)*[0-9]+\\]" . font-lock-type-face))
738 "Additional expressions to highlight in Outline mode.")
740 (add-hook 'outline-mode-hook
742 (setq font-lock-maximum-decoration t)
743 (setq font-lock-maximum-decoration t)
744 (make-local-variable 'font-lock-defaults)
745 (setq font-lock-defaults '(outline-font-lock-keywords t))
746 (font-lock-mode global-font-lock-mode)