Movatterモバイル変換

[Python-Dev] new Makefile.in

• Previous message:[Python-Dev] Weird use of hash() -- will this work?
• Next message:[Python-Dev] new Makefile.in
• Messages sorted by:[ date ][ thread ][ subject ][ author ]

Spurred on by comments made by Andrew, I spent some time lastnight overhauling the Python Makefiles.  I now have a toplevelnon-recursive Makefile.in that seems to work fairly well.  I'mpretty sure it still should be portable.  It doesn't use includesor any special GNU make features.  It is half the size of the oldMakefiles.  The build is faster and its now easier to follow ifsomething goes wrong.A question: is it possible to break the Python static library up?For example, instead of having libpython<version>.a haveParser/parser<version>.a, Objects/objects<version>.a, etc?  Therewould still only be one shared library.  This would speed upincremental builds and also help Andrew with PEP 229.  I'mthinking that the Makefile do something like this:    all: python$(EXE)    PYLIBS= Parser/parser.a Objects/objects.a ...  Modules/modules.a    python$(EXE): $(PYLIBS)        $(LINKCC) -o python$(EXE) $(PYLIBS) ...    Modules/modules.a: minpython$(EXE)        ./minpython$(EXE) setup.pyAFACT, the only thing affected by splitting up the static libraryis Misc/Makefile.pre.in.  Is this correct?  Neil

• Previous message:[Python-Dev] Weird use of hash() -- will this work?
• Next message:[Python-Dev] new Makefile.in
• Messages sorted by:[ date ][ thread ][ subject ][ author ]