more nfo conversion

copy trunk/freewrt r3264 into a new branch for making the nfo system
because upgrading the old adk branch would be a hell more work

upgrade to a prerelease version of mksh R30, which is even smaller in the
“extra small” (default in FreeWRT) mode – at the cost of vi editing mode
support – and has a lot of bug fixes and some new functions (even some of
the GNU bash extensions are supported)

first of a series of commits I'd suggest to not test until I did:

remove busybox ash and the other shells (hush, lash, msh); always
use mksh by default, as /bin/sh, /bin/ash symlink, /bin/mksh native,
and default admin login shell unless bash or zsh are selected

mksh is a superset of ash (I didn't find anything busybox ash can
do that mksh can't), and the upcoming mksh R30 will be even smaller
in -DMKSH_SMALL mode (but that one will lack vi editing mode)

ok wbx@

we planned this for 1.2, but it will greatly help the nfo conversion
and users request shells they can actually use if they don't have a
webinterface handy

for those who have 8224 bytes¹ (on nokernel-i386 platform) too much,
here's an option to *not* build an extra small mksh

suggested by wbx@

¹) in mksh R30, the savings from -DMKSH_SMALL will be much larger,
   alone the vi editing mode is taking up about 17 KiB on i386 BSD…

austriancoder suggested that I pack tsort(1) into tools/ for these
whose tsort doesn't do -r

Now, GNU tsort doesn't do -r, and the only other tsort I know comes
bundled with mirmake (and depends too much on libmirmake and its other
inner workings to separate it from it), and mirmake has a dependency
on guess what…

This is untested, but seems to work.

move arc4random(3) and, while here, libnotimpl into uClibc
this should make a few executables even smaller, most notably
mksh, rdate, openntpd, everything with OGG Vorbis

the treebreaker ☺

overhaul the fetch'n'checksum system to be more bsd-like
• make fetch, make checksum supported
• future expansion possibilities already prepared:
  – more distfiles than just one
  – check sizes of downloaded files (size matters!!!!!111)
  – more hashes than just MD5
  – recursive fetch/checksum targets
  – PERMIT_{SOURCES,DISTFILES}=Yes/«reason» – licence issues
• download.pl dies

quirks:
• GNU make only has a poor excuse for a .for loop
• some things (esp. toolchain and target/linux) now have
  a ${WRKDIR} where they didn't have it before
• ${WRKDIR}/.checksum_done – no comment… it was necessary
  because otherwise everything will get rebuilt every time
  (did I mention that GNU make sucks?)
• I had troubles tracking the dependency problems down
  because GNU make, unlike BSD make, doesn't have good debugging
• DISTFILES -> FULLDISTFILES
• DL_DIR -> DISTDIR
• PKG_SOURCE -> DISTFILES
• PKG_SOURCE_URL -> MASTER_SITES
• no @SF and @GNU any more, use BSD network.template syntax
  (feel free to add more master sites)
• GNU make doesn't really strip whitespaces from variables…
  (did I mention how much better BSD make is?)
• echo 'LOCAL_DISTFILES=/path/to/distfiles' >>${TOPDIR}/prefs.mk
  will save the need to download 'em, Closes: #102
• MASTER_SITES now *must* have the trailing slashes (there may be
  a few that don't and a few with double ones, please fix them)

I've tested that a normal build (unchanged .config from the default) works
on GNU/Linux. There's definitively room for improvement, but most of that
will come with the NFO system only. Fixes appreciated.

This work sponsored by AurISP, part of the NFO conversion project ☻

mksh R29f, fixes a distressing regression

fix build warnings; speed up build

add arc4random(3) from package/rdate distfile

conffiles were undocumented, but now I got it. add.

shouldn't commit only half of it.

mksh R29e
dedicated to David Ramsey and Coverity, Inc. for spotting all these bugs

• ed: the distfile is now at the MirBSD master site, too
• ed, mksh: take maintainership, bump dashver
• rdate: new package (can do rdate _and_ SNTP, is very small, fast and secure)
• busybox: default built-in rdate to “no”, use EITHER that OR new rdate package

upgrade to R29d

• mksh: update to R29b
• scan-tools.sh: if the host compiler is a gcc subject to gcc PR#30785
  work around this bug by adding -fno-tree-vrp to HOSTCFLAGS; code ta-
  ken from mksh R29b Build.sh; bug found by http://blog.fefe.de/?ts=bb2654d4

mksh R29, released today, even smaller than the previous versions

tonight's updates from the mir* front - I have a paxmirabilis package
ready, but as long as we don't have an "alternatives" system, i.e. "I
want the cpio and tar from busybox / paxmirabilis / no", it won't get
us much, so I don't add it…

fix the 'hd' alias for hexdumping to not go beyond 80c screen width
from mksh-current cvs

update
* bug fixes
* size improvements (32 bytes less)

busybox ls(1) doesn't have -o

merged from branches/common-adk 1310:1685

the following packages are not converted yet:
* alsa
* asterisk
* pmacct
* zaptel
* elinks
* libgd
* lua
* mypackage
* openser
* php*
* snort*
* tmsnc

additionally, freeradius doesn't compile.

happy fixing :P

we do determine the hostname slightly differently from the rest

major package Makefile cleanup. remove mostly-clean, add licence header, always use RSTRIP varibale for stripping, rename some init scripts, indent

update to mksh R28

* adding mksh to /etc/shells is no longer necessary as of r441
* while here, instead of patching /etc/profile on mksh postinstall
  time, add a hook for /etc/mkshrc reading to standard /etc/profile
  and create that; bump mksh patchlevel
  note: the next _upstream_ mksh version will already come with a
  sample ~/.mkshrc which we will install (patched appropriately) as
  /etc/mkshrc (read from within our /etc/profile, not by mksh default)
  then, to get rid of the postinstall script as well

let the user decide which of the available shells he
wants as /bin/sh - not only these included with busybox

(ash and mksh are probably the only ones which will
work, but like this I won't affect any existing con-
figuration)

this is what you people get for using := ...

also fix some indentation while here

instead of setting WRKBUILD, set WRKSRC and let WRKBUILD?=WRKSRC like mirports
(WRKSRC isn't used anywhere though)

* whitespace cleanup, while here
* move WRKDIR and DISTFILES definition into PKG_template
* define WRKDIR via parent BUILD_DIR instead of going up from WRKSRC
* use a sub-directory as WRKDIR, like BSD
* rename PKG_BUILD_DIR into WRKBUILD
* define WRKBUILD relative via WRKDIR
* rename PKG_INSTALL_DIR into WRKINST
* fix a few dependencies
* make use of common cases in the definition of WRKBUILD and
  WRKINST, removing many redundant lines
* use WRKDIR instead of BUILD_DIR in places where the latter
  was used previously
* while here, remove PKG_CAT stuff

this only affects packages, for now

oops, pasto during all the recent changes
fix hostname determination

save 2 bytes in /etc/profile

* save a few bytes in /etc/profile and the environment
* we use USER, so define it if unset
* export HOSTNAME and USER if defined by us
* bump patchlevel

mksh R26e, and sync with MirOS src/etc/profile,v 1.23
and ports/shells/mksh/pkg/DESCR,v 1.2 (for $PS1) thus
saving a few bytes

Update to mksh R27d (first OS which includes it!),
switch to upstream CPIO distfile.

* Add more compatibility for cross-builds in the build system
* ~/mkshrc is not executed for non-interactive shells

Not yet tested, but the IPKG builds. Will test ASAP,
I'm under FrOSCon strain.

bump package version, remove 'bash-like feeling for $PS1'(*)
because if HOME=/ it looks stupid

tested with this package being built _on_ mirbsd! yay! :)

(*) http://cvsweb.allbsd.org/cvsweb.cgi/src/etc/profile.diff?r1=1.18&r2=1.19&cvsroot=mirbsd

make mksh cross-buildable and reduce executable size

add mksh package

ok wbx@ "unusual large shit to /etc/profile warning"