Home | History | Annotate | Line # | Download | only in sets
README revision 1.9
      1 # $NetBSD: README,v 1.9 2009/09/07 02:27:27 jnemeth Exp $
      2 
      3 To: "NetBSD Port Maintainers"
      4 Subject: distribution sets, etc...
      5 Date: Mon, 18 Jul 1994 05:02:11 -0400
      6 From: "Chris G. Demetriou"
      7 
      8 [ this is also going in a README in the 'sets' directory... ]
      9 
     10 well, i've finally sat down and started doing distribution set
     11 stuff in a "nice, machine-independent way."  Should have done
     12 it months ago, but too much other stuff going on then...
     13 
     14 anyway, grab ~cgd/sets.tar.gz -- it's three scripts, and a bunch
     15 of files.
     16 
     17 the scripts should be run from the directory where they reside.
     18 
     19 makeflist:	output the list of files that should be in a
     20 		distribution, according to the contents of the
     21 		'lists' directory.
     22 
     23 checkflist:	check the file list (as internally generated
     24 		by makeflist) against the tree living in $DESTDIR.
     25 		(that tree should be made with 'make distribution'.)
     26 
     27 maketars:	make tarballs of the various sets in the distribution,
     28 		based on the contents of the lists, the tree in
     29 		$DESTDIR, and put the tarballs in $RELEASEDIR.
     30 		Note that this script _doesn't_ create the 'secr'
     31 		distribution, because (for now) it requires
     32 		manual intervention to get the binaries right...
     33 		(i'll add another script to create that dist, later.)
     34 
     35 what's in 'lists':
     36 
     37 lists describing file sets.  There are two sets of lists per file
     38 set: machine dependent and machine-independent files. (there's
     39 also another file in the 'man' dir, which is used by the 'man'
     40 and 'misc' sets, but that's explained later.)
     41 
     42 There is one machine-independent file, named "mi".  There are
     43 N machine-dependent files (one per architecture), named "md.${ARCH}".
     44 
     45 XXX - Temporarily (while USE_NEW_TOOLCHAIN is in effect) there are also tc.mi
     46 and tc.old files in some of the set lists. tc.mi lists files in the new 
     47 toolchain that didn't exist in the old toolchain. tc.old lists files removed
     48 and aren't present in the new toolchain. All other toolchain information was
     49 left for now in the various mi/md files. The makeflist script will check
     50 for USE_NEW_TOOLCHAIN and pull the appropriate one into place. Once all
     51 ports are converted to the new toolchain the entire toolchain should be pulled
     52 into tc.mi and/or tc.${MACHINE} and everything in tc.old moved to obsolete.
     53 
     54 the sets are as follows:
     55 
     56 	base:	the base binary set.  excludes everything described
     57 		below.
     58 
     59 	comp:	compiler tools.  All of the tools relating to C, C++,
     60 		and FORTRAN (yes, there are two!) that are in the
     61 		tree.  This includes includes, the linker, tool chain,
     62 		and the .a versions of the libraries.  (obviously,
     63 		base includes ldd, ld.so, and the shared versions.
     64 		base also includes 'cpp', because that's used by X11.)
     65 		includes the man pages for all the binaries contained
     66 		within.  Also, includes all library and system call
     67 		manual pages.
     68 
     69 	etc:	/etc, and associated files (/var/cron/tabs, /root,
     70 		etc.).  things that shouldn't be blindly reinstalled
     71 		on an upgrade.
     72 
     73 	games:	the games and their man pages.
     74 
     75 	man:	all of the man pages for the system, except those
     76 		listed elsewhere (e.g. in comp, games, misc, text).
     77 		Includes machine-dependent man pages for this CPU.
     78 
     79 	misc:	share/dict, share/doc, and the machine-dependent
     80 		man pages for other CPUs which happen to always
     81 		be installed.
     82 
     83 	modules:	stand/${MACHINE}/${OSRELEASE}/modules kernel modules
     84 
     85 	tests:	unit, regression, integration and stress tests for the
     86 		whole system.
     87 
     88 	text:	text processing tools.  groff and all of its friends.
     89 		includes man pages for all bins contained within.
     90 
     91 
     92 as noted, in addition to the "standard" files in each dir, there's
     93 a file called 'md_share' in lists/man.  it's the list of man pages
     94 that are installed from /usr/src/share, which are machine-dependent.
     95 (note that ones that are installed from elsewhere, and thus are
     96 installed on only one architecture, are listed in the md.${ARCH}
     97 file.)  basically, it's grepped through, to see which of the
     98 machine-dependent man pages that are always installed should go
     99 in the 'man' set, and which should go into the 'misc' set.
    100 
    101 
    102 I've made attempts to get the md files for the amiga, i386, hp300, and
    103 sparc done.  I've not attempted the rest (though there are empty files
    104 for all architectures except the da30, in the various directories).  I
    105 think i got the amiga, i386, and sparc done correctly, but i know that
    106 (because of weirdnesses in how their last snapshots were built) i didn't
    107 get hp300 done right.
    108 
    109 what you folks need to do:
    110 
    111 run 'checkflist' on a tree that you've created with 'make
    112 distribution'.  send me the output, and i'll take care of putting
    113 stuff on the right lists for you.  In the future, you'll get
    114 to do it yourselves, as you add programs and include files, but
    115 i'd prefer to do it the first time (unless you really want to do it
    116 yourself -- in that case, please ask me to check your suggestions
    117 re: the appropriate sets for things...)  There should be a minimal
    118 list output by checkflist -- it should all be things that
    119 _should_ be on your lsits, but aren't, i.e. machine-dependent
    120 binaries, man pages, and include files.
    121 
    122 Also, please look through the lists, and tell me if you think
    123 i've put anything in the wrong place.  i think i've
    124 taken care of all of the machine-dependence bugaboos, by
    125 doing it for the N architectures, but i'm not _sure_...  8-)
    126 
    127 
    128 thanks,
    129 
    130 chris
    131