1.5.17.jcn6 - aka: `pgwit! Ah, dacht dat-i zo wel aardig was.' =========== * Some more hacking at midi2ly.py: - options and stuff: moved to library for ly2dvi, midi2ly, mup2ly, update-lily - handle keys and note names - simple quantising - handle tuplets and silly durations - guess clef - don't include empty staffs in score - print relative pitches by default - don't repeat duration by default - bugfix: allow 8th notes too - include new version of input/test/midi-scales.ly - added barchecks + fix - bugfix for relative mode * Shorter MIDI creation texts. * Bugfix: Key_change_req::transpose: don't deliver reversed list. * Bugfix: Key_performer::create_audio_elements: transpose list to `do', before determining tonality. 1.4.8.moh1 ========== * new property "end-alignment" to control non-centered lyric alignment 1.5.17.hwn1 =========== * Be quicker in Grob::handle_broken_dependencies(). * --strict option for abc2ly, exit if an error is found. * GUILE 1.3.4 fix. 1.5.17 ====== 1.5.16.hjj2 =========== * Emacs-mode: Inserting tags * Enable python2.1/Python.h 1.5.16.hwn1 =========== * Add unfold-repeats function (Rune Zedeler!) to standard init SCM file. * Bugfix: spacing is no longer confused by coupled clefs, where one of the clefs is loose, and the other not. * Robustness fix for Slur. Don't crash if attachment not set. * Arpeggios can now have arrows on the top or bottom to determine their direction. (MF code by Chris Jackson) * Symbol cache bugfix in system-start-delimiter.cc -- may switch symbol cache off for gcc 2.96? * MikTeX PDF detection bugfix (Mats Bengtsson) * Some more random hacking at midi2ly.py. (jcn1) * Inline unsmob_XXX functions. Speedup of 6% on wtk1-fugue2. 1.5.15.jcn1 =========== * Fixes for latest guile cvs. * Add --enable-optimising flag to configure (Huh?). * Only inline string utlis when optimising. * Compile fix. 1.5.15.rz1 ========== * added property autoReminders, automatically creating reminder accidentals. - major changes to accidental-engraver. 1.5.15.hwn1 =========== * Junk translator-description.scm * Make Repeated_music use 'element and 'elements in stead of 'body and 'alternatives. Associated changes in parser and music-sequence.cc. * Change Part_combine_music and Lyric_combine_music to use 'elements music property. * Porrectus patch (Juergen Reuter) - Rewrote code for vaticana style porrectus grob; the solid shape is now drawn as a single bezier sandwich rather than composed from a couple of misused slurs. Looks now much nicer. - Added code to forbid line-breaking inbetween a porrectus (preliminary; to be moved to ligature engraver framework). - Added auto-properties property for automatic determination of grob properties add-stem and stem-direction from musical context. - Small clean-ups. 1.5.14.jcn6 =========== * Some more random hacking at midi2ly.py. * Slightly better doco for r, s and \skip, with examples. * Bugfixes: add-html-footer.py and @MAILADDRESS@. * Some website related fixes (thanks Tiggr). * Ugly hack in add-html-footer for disabling tutorial and refman links in sidebar. * Guile > 1.4 compilation fixes. * Website bugfix: mailto: (thanks David Boersma). 1.5.14.hwn1 =========== * Rename: Local_key_engraver to Accidental_engraver * Add documentation for accidentals grob property. * Don't make double accidentals even if two of the same notes are specified. (Thanks Rune) * Search all staff-bars for non-empty barline to determine type. This fixes span-bars with the lowest staff hara-kiried. (Thanks Rune) 1.5.14.rz1 ========== * Less buggy beam-split * Added subdivideBeams, subdividing beams on beats 1.5.14 ====== 1.5.13.hwn2 =========== * Bugfix: revert font-relative-size for grace Accidentals * Bugfix: hara kiri'd piano staffs doesn't take space anymore. * Bugfix: reinstate rest-collisions. * Bugfix: key items have the c0 position again. * Make \property Voice.GrobName = \turnOff work again. * More intelligent symbol-cache, in less code. Turned off by default, doesn't work with -O2 yet. * Documentation fixes: change documentation of Moment and make-moment, add Translator::{name,description} * Junk STL implementation of Scheme_hash_table in favor of GUILE hashtables with automatic resizing. 1.5.12.jcn2 =========== * half-baken --with-kpathsea configuration fix. * ly2dvi: add `.' to TEXINPUTS to catch broken TEXINPUTS settings. * guile-1.4 compile fix. 1.5.12.hwn1 =========== * Move engraver descriptions into the C code. They are accessible from Scheme. Grand rehacking to make stuff work again. * Precompute which engravers are eligible to acknowledge a certain type of grob. Speedup: approx 10 % on interpretation phase. 1.5.12 ====== 1.5.11.jcn1 =========== * Add -p flag to package-diff.py's diff invocation, to show changed function. * Add configure checks for python headers. * Make midi module. 1.5.11.hwn1 =========== * Allow breaks at measures starting with grace notes. * Bugfix: don't give up when there is one entirely unspaceable measure/line. * Midi module: many bugfixes. * midi2ly python blurble. 1.5.11 ====== 1.5.10.jcn3 =========== * include python midi parser. * stepmake updates. * add .cvsignore patterns for making patches the standard, manual way. * midi2ly: support chords, duration conversion rewrite. * parser.yy: Guile > 1.4 compilation fix; scm_unprotect_object is deprecated. 1.5.10.hwn1 =========== * Rational bugfix. * Cleaned up definitions of SystemStartDelimiter * WTK fix. * Porrectus patch (Juergen Reuter): - Syntax change: \porrectus -> \~ - New porrectus properties: porrectus-width, line-thickness. - Bugfix: left<->right confusion in porrectus-engraver.cc. - Bugfix: Box dimensions for horizontal slope molecule in porrectus.cc. - Enhanced mensural style porrectus shape, considering optical illusion in vertical endings due to slope. Simplified drawing. - Code clean-up. 1.5.10 ====== 1.5.9.hjj2 ========== * Emacs-mode: Keyboard shortcuts, separated identifiers from keywords and spacing not necessary in font-locking. 1.5.9.jcn3 ========== * add-html-footer.py: use html
. * Added some template .ly's for evaluation. * Make blot work without -fps too. * Use blot diameter for postscript beams and boxes. * Guile-1.5.0-cvs updates and Guile-1.4.x compatibility. * Beam knee fix. * Now really included lexer-gcc-3.0.patch. * Make Voice context accept Thread again; this fixes MIDI output for music that uses Thread contexts. * Removed debug printing from chord-name.scm. 1.5.0 ===== 1.5.0.jcn1 ========== * Added feature to ly2dvi to find feta pfa font files used in a postscript file to make printing of direct postscript a bit easier: lilypond -fps input/trip.ly cat $(ly2dvi -f trip.ps) trip.ps | lpr * Reincluded 1.4.3.jcn2. * Website fixes. * Ly2dvi now leaves .tex and .latex output if latex fails, and cleans the temp dir if anything fails. * Fixes for windows scripts. 1.5.0.uu1 ========= * Hara kiri and span-bar. * Junk old grace stuff. * Don't crash on grace notes in MIDI output. 1.5.0 ===== * experimental grace notes: - add grace timing to Moment - hack to Sequential_music_iterator to use the correct timing for the note before a grace - change the implementation of \grace construct - property set/unset (eg. font size) is done through startGraceMusic and stopGraceMusic - various engraver fixes to deal with the fact that the main-time-moment may remain the same during several steps. - Change column creation logic in score-engraver. * Bugfix: don't crash when \name not set in \translator block. * rename ly/*.ly to ly/*-init.ly (except language files) to avoid name clashes * deprecate Property_engraver, add Font_size_engraver to Staff, Voice and Thread. * Change mailing list address: *gnu-music*@gnu.org becomes *lilypond*@gnu.org * Put span-bar lines only between the other barlines, not on top of them (thanks to Juergen Reuter) 1.4.4.jcn3 ========== TODO: look at other filledbox'es (stafflines and barlines don't match up). * Testing code off by default. On: -e "(define ps-testing 1)" -fps * Output via TeX/dvips still gives problems. Use direct postscript output for real testing. * Beam and stem fixes: - Stems drawn in PostScript (should do all filledbox'es in PS?). - Fixes for draw_box and draw_beam PostScript routines. - Stems reach to top (or bottom) of beam. - Beam uses correct stem thickness (only right beams, left TODO). - testing code in effect: only outlines are drawn.