Return to DEVELOPMENT CVS log

Up to [GnuMach] / Gnu-Mach

Annotation of Gnu-Mach/DEVELOPMENT, revision 1.1

1.1     ! root        1: -*- Text -*-
        !             2: 
        !             3: If you're ``just'' looking for something to work on, have a look at the
        !             4:   * bug list, <http://savannah.gnu.org/bugs/?group=hurd> and
        !             5:   * task list, <http://savannah.gnu.org/task/?group=hurd>.
        !             6: 
        !             7: HOW TO CONTRIBUTE LARGER CHUNKS
        !             8: ===============================
        !             9: 
        !            10: If you want to help the maintainers to be quickly able to evaluate and
        !            11: check in your contribution, please try to follow these suggestions:
        !            12: 
        !            13: Try to mark in the code stuff (i.e. whole functions, parts of header
        !            14: files) that you've just copied (and then perhaps modified; also note that
        !            15: briefly) from somewhere else and stuff that you've actually written
        !            16: yourself.  Either do that by simply writing a ChangeLog in parallel (an
        !            17: informal one is fine as well) or put notes in the modified / imported
        !            18: files.  The one who will be checking in your patches will then probably
        !            19: remove most of these notes, as soon as everything is written down in the
        !            20: real ChangeLog.  Logging your changes right from the beginning makes it
        !            21: much easier for the maintainers to track down where which chunk is coming
        !            22: from, so that they can be handled appropriately.
        !            23: 
        !            24: HISTORY
        !            25: =======
        !            26: 
        !            27: `gnumach-1-branch-before_removing_unused_and_unsupported_code' was tagged
        !            28: on 2006-02-20.  After creating that tag, code for unused and unsupported
        !            29: device driver for ISA cards and a good deal of i386 dependent, also
        !            30: unused and unsupported code was removed.
        !            31: 
        !            32: On 2006-03-19, support was removed for FIPC, which only ever was used
        !            33: within the native Mach NE2000 NIC device driver, see
        !            34: <http://www.cs.utah.edu/flux/mach4-i386/html/mach4-UK22.html#FIPC>.
        !            35: <http://lists.gnu.org/archive/html/bug-hurd/2006-01/msg00162.html>.
        !            36: 
        !            37: Support for NORMA was removed on 2006-03-20.
        !            38: <http://lists.gnu.org/archive/html/bug-hurd/2006-03/msg00007.html>.
        !            39: 
        !            40: Support for PS2, i860, iPSC 386/860 and MB1/MB2/EXL was removed on
        !            41: 2006-11-05.
        !            42: <http://lists.gnu.org/archive/html/bug-hurd/2006-11/msg00001.html>.
        !            43: 
        !            44: Support for the old ipc interface, MACH_IPC_COMPAT, was removed on 2006-12-03.
        !            45: <http://savannah.gnu.org/patch/?5017>.
        !            46: 
        !            47: Support for building without CONTINUATIONS was removed on 2006-12-03.
        !            48: <http://savannah.gnu.org/patch/?5019>.
        !            49: 
        !            50: Support for FP emulation was removed on 2006-12-13.
        !            51: <http://lists.gnu.org/archive/html/bug-hurd/2006-12/msg00031.html>.
        !            52: 
        !            53: Support for Olivetti XP7 & XP9 was removed on 2007-01-02.
        !            54: <http://lists.gnu.org/archive/html/bug-hurd/2006-12/msg00107.html>.
        !            55: 
        !            56: Support for the `iopl' device and some i/o emulation code (that might be useful
        !            57: for DOSEMU) was removed on 2007-04-02.
        !            58: <http://lists.gnu.org/archive/html/bug-hurd/2007-04/msg00002.html>.
        !            59: 
        !            60: 
        !            61: Be sure to check the ChangeLog and have a look at the repository at that states
        !            62: if you want to work on those parts of GNU Mach.
        !            63: 
        !            64: LAYOUT OF THE SOURCE TREE (very incomplete)
        !            65: 
        !            66:   * include/ 
        !            67: 
        !            68: [TODO: Check.]
        !            69: 
        !            70:     ... is mainly for installed header and definition files, but it also holds
        !            71:     pseudo-clones of C library headers, which don't get installed because the C
        !            72:     library has better versions.  In that category are <mach/error.h>,
        !            73:     <mach/mach_traps.h>, <mach/mig_support.h>, <sys/ioctl.h>, <sys/reboot.h>,
        !            74:     <sys/time.h>, <sys/types.h>, <alloca.h> and <string.h>.  By putting such
        !            75:     headers into there, the relevant kernel code is easier to understand,
        !            76:     because the user will expect that the file named <alloca.h> or
        !            77:     <sys/types.h> does more or less what the normal C library file does, and
        !            78:     calling those <kern/alloca.h> or <kern/types.h> would make the reader have
        !            79:     to wonder or remember what they are.  The directory is, essentially, a
        !            80:     special `/usr/include' for use by the kernel itself when compiling.  It
        !            81:     only should get things which belong in `/usr/include'.  The reason for
        !            82:     <alloca.h> and <sys/types.h> is because those are files found in
        !            83:     `/usr/include', even if on an actual installed system the versions in
        !            84:     `/usr/include' are provided by a different package.

This archive runs on limited infrastructure. Preserving old code on modern bandwidth. Automated agents are requested to crawl responsibly.