Yamano-uchi added support for SFNT-wrapped CID bitmap font files (with
Updating unix/ftconfig.in to recent config/ftconfig.h changes.
Added little know adx offset to t1operator_seac() -- accent positioning now seems correct.
Forgot to run autoconf...
Fixed Z1_Parse_CharStrings() to work correctly with "open" paths (ie. missing closepath operator)
fixed t1_tofixed() to handle floats of the form .001 and -.001 correctly.
Fixed dec-osf compiler switches. From Tom.
Removing FT_MAKE_OPTION_SINGLE_LIBRARY_OBJECT. It has never worked.
Added #ifdef's for C++ to all header files.
Added a `distclean' target to remove files which are produced by the UNIX
fixed two bugs related to T2 real numbers: the "skipping" code didn't behave correctly (leaving a bad stack ptr), and the actual parsing code also parsed the code (30) as two nibbles.
Some simple patches that allow synthetic fonts to be rendered. *However*: although the FontMatrix is read correctly, it does _not_ seem to be used! I don't understand this...
Use the ANSI offsetof() macro instead of something home-brewn.
Added autoconf tests FT_MUNMAP_DECL and FT_MUNMAP_PARAM to fix some devils.
Simplifying the FIELD_* and FRAME_* macros. Before calling these macros,
Fixing gcc warnings on Solaris.
Making path of $INSTALL always absolute.
Update to latest versions.
Applied a patch from Tom Kacvinsky (taken from the curren libtool CVS) to
FT_Done_Stream(): do stream->close = 0 after calling stream->close(); this makes it safe to call FT_Done_Stream() more than once.
Adapted to new external stream semantics.
Oops, I had exaggerated the warning levels a weensy bit...
Updated to reflect the new Mac FOND support setup.
Obsolete; Mac FOND support now lives in src/base/ftmac.c.
Put FT_New_Face() in an #ifndef macintosh switch; ftmac.c provides an extended implementation.
Removed Mac fond driver #ifdefs
Brand new Mac support -- no more muching with changing drivers and reinitializing streams, but a clean layer on top of FT_Open_Face(): an extended version of FT_New_Face().
Header for one Mac-specific additional API call.
re-adding a "unix-dev.mk". Debugging libtool output
*slight* formatting only. Congratulations, David! :-)
small fix to the auto-hinter:
Adapting changes for Unix to other makefiles.
Removing compiler warnings about unused portions of code.
Updating to libtool 1.3.5.
small compilation fix on 64-bits machines..
Use `patsubst' instead of `subst' in cleaning rule to avoid problems with
now that "type1z" is renamed to "type1"; revert to hardcoding "type1"..
- fixed the incorrect SO extension for Visual C++
Use the "type1z" driver instead of the old one.
Changed parse_subrs() so that it'll work correctly both with wrongly sorted subr indices *and* non-contiguous subr arrays. Thanks to Tom K. for helping to isolate the problem.
`make install' is available now.
Adding $(SO) and $(SA), denoting objects and library for a static build.
removed duplicate macro (FT_MODULE_IS_DRIVER was defined twice)
Hardcode loading the Mac FOND driver here, as config/ftmodule.h sometimes gets regenerated.
updating the Mac stuff to the current situaton
For Unix platforms, `configure' will be always called now.
Adding $(CFG) variable to pass configure script parameters.
some welcome fixes related to the auto-hinter:
Boris's patch regarding glyph index 0
added auto-hinter module. Note that the code has been
- introduced FT_Get_Glyph_Name (see freetype.h)
Renaming license.txt -> LICENSE.TXT for orthogonality.
Boris patches for ascender/descender/text height
more patches from Tom (to fix its former patches :-)
removed some warnings with Visual C++
Formatting/Synchronizing.
fixed a bug that caused invalid clipping of monochrome bitmaps
fixed a really tupid bug in the Multiple Master support
removed some stupid #defines in <freetype/config/ftconfig.h>
Removed `demos' tree. Use `ft2demos' module instead.
ttdebug will compile again...
the font driver now supports CEF fonts and needs a few
fixes from Tom related to 64-bitness and strange
fixed a rather stupid bug in the last patch (one of those with
removed the "config" directory, now replaced by "builds"
vast clean-up of the sources in order to allow flat
renaming "freetype2/config" to "freetype2/builds"
removed a compiler warning..
- fixed a leak in the Type 1 driver