[Cs-dev] Final Draft of 6.04 release notes
Date | 2014-11-10 14:39 |
From | jpff |
Subject | [Cs-dev] Final Draft of 6.04 release notes |
Comments etc welcome, but soon as we are very close to a release ==John ffitch ------------------------------------------------------------------------ ============================ CSOUND VERSION 6.04 RELEASE NOTES VERSION 6.04 ============================ This new version has many extensions and fixes; many new opcodes and significant numbers of internal reworking. There is a new frontend and iOS and Android version have seen many improvements. As ever we track bugs and requests for enhancements via the github issues system. Already proposals for the next release are being made but the volume of changes require a release now. The Developers USER-LEVEL CHANGES ================== New opcodes: o pinker generates high quality pink noise o power opcode ^ now works with array arguments o exciter opcode, modelled on the calf plugin o vactrol opcode simulate an analog envelope follower o family of hdf5 opcodes to handle hdf5 format files o (experimental undocumented) buchla opcode models the lowgate filter of Buchla o New k-rate opcodes acting on arrays: - transforms: rfft, rifft, fft, fftinv - complex product: complxprod - polar - rectangular conversion: rect2pol, pol2rect, mags, phs, - real - complex: r2c, c2r - windowing: window - cepstrum: pvscpes, iceps, ceps - column / row access: getrow, getcol, setrow, setcol - a-rate data - k-array copy: shiftin, shiftout - phase unwraping: unwrap New Gen and Macros: Orchestra: o Line numbers corrected in instr statements o New control operation, while, for looping o A long-standing bug with macros which use the same name for an argument has been corrected o Redefinition of an instrument in a single call to compile is flagged as an error o ID3 header skip for mp3 files now properly implemented. o Errors induced by not defining the location of STK's raw wave files has been removed o bug fixed where UDO's could not read strings from pfields Score: Options: Modified Opcodes and Gens: o stackops opcodes deprecated o lenarray extended to handle multi-dimensional arrays o ftgenonce accepts string arguments correctly and multiple string arguments o max and min now have initialisation-time versions o gen23 improved regarding comments and reporting problems o in OSCsend the port is now a k-rate value o socksend now works at k-rate o a number of envelope-generating opcodes are now correct in sample-accurate mode o faust compilation is now lock-protected o mp3 fixed to allow reinit to be used with it. o In remote opcode the name of the network can be set via the environment variable CS_NETWORK. Defaults to en0 (OSX) or eth0. Utilities: Frontends: icsound: New frontend icsound is now ready for general use. icsound is a python interface for interactive work in the ipython notebook. csound~: Emscripten: csdebugger: A number of changes and improvements have been made, like stepping through active instruments, better line number use General usage: Jack module now does not stop Csound if autoconnect fails Bugs fixed: o atsinnoi fixed o ftsavek fixed o sprintf fixed o gen27 fixed, especially with extended arguments, as well as fixed a number of errors in extended score arguments. o Physem opcodes (guiro cabasa, sekere) fixed so second call works o flooper fixed in mode 2 o OSCsend multiple fixes o UDO fix for case of local ksmps of 1 o More changes/fixes to dssi code o xscanu and scanu fixed o temposcal and mincer fixed o crash in ftload fixed ==================== SYSTEM LEVEL CHANGES ==================== System changes: o In server mode exit is now clean o Fixes to rtalsa module o Pulseaudio rt module fixes o Fix to remove fluidEngine entries for csound instance (prevents crash on moduleDestroy) o Opcodes called through function calls that returned arrays did not correctly synthesize args as array types due to not converting the arg specifier to the internal format o fixed crashing issue during note initialization for tied notes due to goto skipping over code o fixed incorrect initialization of pfields when note's pfields length were less than instrument expected (off-by-one) Internal changes: * Added Runtime Type Identification for instrument variables; removed use of XINCODE/XOUTCO * fix malloc length in negative number parsing, and improved handling of negative numbers * writing to circularBuffer is now atomic * a number of memory leaks and potential dangerous code have been fixed * type-inference has been extensively reworked, as have a few parsing areas API === * Added API function for retrieving GEN parameters used for creating a table Platform Specific ================= iOS --- * API Refactored for clearer method names and abstraction names (i.e. CsoundBinding instead of CsoundValueCacheable) * Updated to remove deprecated code * A significant amount of reworking has been done on the code Android ------- * API Refactored for clearer method names and abstraction names (i.e. CsoundBinding instead of CsoundValueCacheable) * Changes to enable HTML 5 with JavaScript and it is to be hoped WebGL in the Csound6 Android app. * Enabled change of screen orientation in the Csound6 app without forcing a restart of the app. * Enabled local storage (useful for saving and restoring widget values, etc.). Windows ------- * fixed pointer arithmetic that caused crashing on Windows * pyexec changed to use python's file opening functions to prevent crash on Windows OSX --- * CsoundAC now compiles Linux ----- * threadlocks bug fix on linux. ======================================================================== ------------------------------------------------------------------------------ _______________________________________________ Csound-devel mailing list Csound-devel@lists.sourceforge.net |
Date | 2014-11-10 15:53 |
From | Michael Gogins |
Subject | Re: [Cs-dev] Final Draft of 6.04 release notes |
Attachments | None None |
Looks good! Mike ----------------------------------------------------- Michael GoginsIrreducible Productions http://michaelgogins.tumblr.com Michael dot Gogins at gmail dot com On Mon, Nov 10, 2014 at 9:39 AM, jpff <jpff@codemist.co.uk> wrote: Comments etc welcome, but soon as we are very close to a release |