Skip to content
Commits on Apr 10, 2011
  1. @danstowell
Commits on Apr 4, 2011
  1. @joshpar

    update version number

    joshpar committed Apr 4, 2011
  2. @timblechmann

    rats... I missed two others of the same

    James Harkins committed with timblechmann Jun 27, 2010
  3. @timblechmann
Commits on Apr 3, 2011
  1. @timblechmann

    scons build system: libsclang build fix

    based on a patch by James Harkins <jamshark70@gmail.com>
    
    Signed-off-by: Tim Blechmann <tim@klingt.org>
    timblechmann committed Apr 3, 2011
  2. @timblechmann

    Incorrectly used dataptr instead of fileptr in previous commit on thi…

    …s file
    James Harkins committed with timblechmann Apr 3, 2011
  3. @timblechmann
  4. @timblechmann

    Fix asPseg bug for short curves array (which should wrap, not kill th…

    …e stream)
    James Harkins committed with timblechmann Apr 3, 2011
  5. @timblechmann

    fix corner case in ClassBrowser

    sometimes prevMethod would be nil, so prevMethod.name would break
    James Harkins committed with timblechmann Apr 3, 2011
  6. @timblechmann

    fix two bugs in NotificationCenter registerOneShot:

    - it needs to pass the args to the wrapped function
    - it needs to return a NotificationRegistration, like *register
    James Harkins committed with timblechmann Apr 3, 2011
  7. @timblechmann

    Fix nowExecutingPath bug in scel (never backported?)

    James Harkins committed with timblechmann Apr 3, 2011
Commits on Mar 30, 2011
  1. @timblechmann

    two commits: (1) simplify access to the superclass tree in Class. (2)…

    … when looking for a code file (openCodeFile) or cmd-J, it is now enough to select a full line, instead of having to select both words around the colon.
    
    backported from master in order to fix 3.4.2 issue
    
    Conflicts:
    
    	common/build/ChangeLog
    timblechmann committed Mar 30, 2011
Commits on Mar 28, 2011
  1. @danstowell

    SC 3.4 set correct SOVERSION 1.0.0 for libs, and install more properly.

    (Changes ported from downstream debian packaging.)
    danstowell committed Mar 28, 2011
Commits on Mar 24, 2011
  1. @timblechmann

    scsynth: set ftz flag on osx

    scsynth on osx/x86 is prone to denormal issues, because the denormal
    handling flags of the sse unit are not set correctly.
    
    Signed-off-by: Tim Blechmann <tim@klingt.org>
    Reported-by: Mattias Petersson <emp@wickerwork.net>
    timblechmann committed Feb 11, 2011
Commits on Mar 15, 2011
  1. @danstowell

    Changelogging for 3.4.2

    danstowell committed Mar 15, 2011
  2. @danstowell

    gitignore for 3.4 branch

    danstowell committed Mar 15, 2011
Commits on Mar 11, 2011
  1. @rkuivila @danstowell
Commits on Mar 9, 2011
  1. @danstowell

    SCVim: avoid generating scvim help cache if not currently in scvim

    Backport 9080107 master->3.4,
    for purposes of debian packaging
    danstowell committed Mar 9, 2011
Commits on Mar 2, 2011
  1. @timblechmann

    sclang: ensure minimum stack size

    the sclang interpreter does not do any stack overflow checks. the stack
    size of sclang threads is defaulted to 64, which is quite low. especially
    it caused a gc corruption when running certain code in routines.
    
    this patch increases the default stack size and adds a lower bound to the
    stack size
    
    Reported-by: Jonatan Liljedahl <lijon@kymatica.com>
    Signed-off-by: Tim Blechmann <tim@klingt.org>
    timblechmann committed Feb 27, 2011
Commits on Jan 14, 2011
  1. @timblechmann

    sclang - prString_FindRegexp

    ensure correct size of results array during gc calls
    
    backport 0e77375 from master
    
    Signed-off-by: Tim Blechmann <tim@klingt.org>
    timblechmann committed Jan 14, 2011
Commits on Jan 8, 2011
  1. @timblechmann

    sclang: makeIntrinsicClass - correct bounds for memcpy

    Signed-off-by: Tim Blechmann <tim@klingt.org>
    timblechmann committed Jan 7, 2011
  2. @timblechmann

    sclang: prString_FindRegexp - fill array after allocating objects

    ensures that size and data of the array match.
    backport from 2343cfc
    
    Signed-off-by: Tim Blechmann <tim@klingt.org>
    timblechmann committed Jan 8, 2011
  3. @cappelnord @timblechmann

    array primitives: fix allTuples and unlace

    Signed-off-by: Tim Blechmann <tim@klingt.org>
    cappelnord committed with timblechmann Jan 6, 2011
Commits on Jan 7, 2011
Commits on Jan 4, 2011
  1. @danstowell
  2. @danstowell

    changelog for 3.4.2

    danstowell committed Jan 4, 2011
  3. @danstowell

    Update package script for 3.4 branch (git-aware rather than svn); cha…

    …nges are like those from master but a bit different
    danstowell committed Jan 4, 2011
  4. @danstowell
  5. @danstowell
Commits on Jan 2, 2011
  1. @timblechmann

    unit generators: LagControl - dynamically allocate buffer for filter …

    …states
    
    issue reported by James Harkins.
    
    Signed-off-by: Tim Blechmann <tim@klingt.org>
    timblechmann committed Jan 2, 2011
  2. @timblechmann

    unit generators: LagControl - fix initialization order

    patch by James Harkins.
    
    Signed-off-by: Tim Blechmann <tim@klingt.org>
    timblechmann committed Jan 2, 2011
Commits on Dec 30, 2010
  1. @cappelnord @timblechmann

    fix String:rotate

    cappelnord committed with timblechmann Dec 30, 2010
  2. @timblechmann

    Pipe does not remove closed pipes from openFiles

    Found this in 3.4 -- I don't know if it's fixed in master.
    
    There's an object leak in Pipe. When you close the pipe, the Pipe object
    remains in the UnixFILE.openFiles array. It's properly closed in the
    backend -- fileptr is nil -- but the object will never get garbage
    collected.
    
    One fix is to replicate the way File does it. The other way (maybe better,
    but it might have some unintended consequences I can't think of at the
    moment) would be to promote close and prClose up to UnixFILE, and let File
    and Pipe inherit those methods. That would at least reduce the risk of
    somebody adding something clever to one and not adding it to the other.
    
    Signed-off-by: Tim Blechmann <tim@klingt.org>
    James Harkins committed with timblechmann Dec 30, 2010
Commits on Dec 17, 2010
  1. @timblechmann

    fix Latch first sample output bug: if trigger > 0 initially, latch sh…

    …ould not output 0
    
    git-svn-id: https://supercollider.svn.sourceforge.net/svnroot/supercollider/trunk@10143 a380766d-ff14-0410-b294-a243070f3f08
    jamshark70 committed with timblechmann Jun 5, 2010
Commits on Dec 13, 2010
  1. @timblechmann
Something went wrong with that request. Please try again.