This is being typed in the middle of a thunderstorm, and,
interestingly enough, the Suns and the NeXTs on the network have
stayed up during the power glitches, but the PCs have died a death. 
That's got to be some kind of advert for UPSs, or, not living in
England.

Not much new to report this week, except that to make things compile
(this is the 2nd September tar_files), you'll have to define YYDEBUG
as part of the extra CFLAGS in the usr.bin/xlint/lint1 Makefile, and
you'd be better updating your usr.bin/pagesize/Makefile to something
which looks like it'll allow you to do a make cleandir.

Interesting events this week:

1.  Postgres95 has been released - with a Berkeley style-copyright
replacing the old
no-commercial-use-without-getting-in-touch-with-the-Berkeley-office
one.  Then, a day later, the new copyright has been replaced with the
old one again, and they're asking that if you grabbed a copy of the
Berkeley-style one, that you grab the one that's on the ftp-server
now.  Confused?  Yup.  Anyway, the bottom line is that, if you want a
database running on NetBSD (and FreeBSD) which is SQL-92 (mostly)
compliant, you'll need flex 2.5.2, a fairly recent GNU make, shared
memory in your kernel, and the sources from:

ftp://s2k-ftp.CS.Berkeley.edu/pub/postgres95/postgres95-1.0.tar.gz

2.  No takers for last week's CVS question - how do you remove a
directory from the CVS repository, so a much easier one this week -
say I'd made a change to a file and committed it - that version is now
the HEAD.  The same file is updated by the vendor to include the
changes I'd made locally, how do I make the new vendor-file the HEAD? 
I presume by some checkout switch, although I'm damned if I can find
it.  Can someone point me to the cvs mailing list?  I feel like making
a prat of myself there.

3.  The integration of 4.4lite2 has started, with the first contenders
being usr.bin, and lately, bin.  Keep going, folks...

4.  What's the collective noun for CD-ROMs?  A `platter'?  It looks
like there's about to be a platter of NetBSD-related CD-ROMs about to
hit the caddies out there.  There's one from the folks at wasabi, and
I'm involved with two that will be released on September 23rd 1995.
For more information, mail me at agc@westley.demon.co.uk, but remember
I go on vacation on Saturday 9th.

5. To go back to yacc only defining yydebug if YYDEBUG is set, I've
had problems compiling perl 4.036, which expected yydebug, as did our
friend xlint - see above.  What's the justification for this change,
guys, because if it starts breaking perl and other things, it's
probably a contender for backout-of-the-week?

6.  There are plans afoot to use ncurses as a replacement for the
Berkeley curses that are used at the moment - it's just that ncurses
seems to have gone through major changes just recently, and, I gather,
will be imported as soon as it's stabilised.

7.  The latest snapshot of wine sounds as though it's doing very well,
and minor apps sound like they're working.  Check it out, if you're
using the i386 port.

8.  If anyone's still awake, some of the CVS commits for the September
2nd tar_files follow:

8.1 distrib/pc532
New distribution "floppies" from Matthias Pfaller.
8.2 bin/mt mt.c 
Sync exit codes with man page (from Scott Burns).
8.3 etc rc 
Configure ccd devices before the filesystem checks, as suggested by
Thomas Eberhardt in PR #1408.
8.4 sys/compat/linux linux_file.c linux_ioctl.c linux_signal.h syscalls.master 
Add a hook for MD ioctl calls, plus a couple of more changes to make
svgalib binaries work on the i386
8.5 sys/arch/atari/conf files.atari 
Detect type and precence of an FPU in the kernel instead of relying on GEM.
8.6 gnu/usr.bin/gdb/gdb/arch/ns32k
        nm.h ns32k-nat.c ns32k-opcode.h ns32k-tdep.c tm-umax.h tm.h 
Changes from Matthias Pfaller to get gdb to work.
8.7 sys/arch/pmax/dev cfb.c 
Work around a NetBSd/pmax kernel bug inherited from 4.4BSD: the DEC
TURBOChannel cfb has a vertical-retrace interupt that cannot be disabled
in software. 4.4BSD and NetBSD don't supply an interrupt hander and
simply neverenable interrupts from slots with a cfb.  This has never
ever worked with a cfb in a 3MIN (Decstation 5000/1xx), where a TURBOChannel
card in slot 0 (or 1 or 2) interrupts at the same IPL as spl0() (or spl1
or spl2), and there's also never been support for selectively enabling or
disabling those interrupts on a 3MIN, in either NetBSD or 4.4BSD.
This  revision add an interrupt handler for the cfb, used only on 3MINs,
so that a 3MIN can boot with a cfb as console, and enable/disable of TC
slot interrupts can be debugged. (Serial consoles via ioasic SCCs still
don't quite work,  and simply clearing the relevant bits in the R3000 cause
register is apparently over-ridden by a subsequent spl0() somewhere else.)
[This week's Tolstoy award for lengthiest commit message - agc]
8.8 sys/arch/sparc/dev espreg.h 
define reserved bits in cfg2 & cfg3 registers correctly.
8.9 sys/arch/atari/stand/loadbsd loadbsd.c loader.h 
Remove FPU detection and report on stdout instead of stderr.
8.10 sys/arch/pc532/stand
        Makefile boot.c conf.c cons.c devopen.c scn.c sd.c srt0.s 
        inflate.c rd.c 
Changes for new INSTALL generic kernel and compressed kernel boot.
8.11 sys/dev ccd.c 
Wrap dk stats stuff in `#ifdef WORKING_DISK_STATISTICS'.  Takes care
of PR #1388 until such time as dk stats works on all ports.  XXX !!
8.12 sys/arch/i386/isa/pcvt
        pcvt_drv.c pcvt_ext.c pcvt_hdr.h pcvt_kbd.c pcvt_out.c 
        pcvt_sup.c pcvt_vtf.c 
* Don't scroll by setting the CRT start address when in graphics mode, and
  make sure to use the right buffer (fixed in the same way as pcvt 3.30)
* Maintain keyboard state locally per VT. This makes sense, enables switches
  between VTs with different kbd states, and aids Linux VT emulation.
8.13 lib/libc/gen syslog.c 
Don't overflow fixed sized buffers.  Parts of this patch come from
Perry Metzger, Charles Hannum and Paul Kranenburg.
8.14 sys/arch/pc532/dev scn.c 
add option CON_BRK_PANIC to be able to force a panic by 3 consecutive
breaks without any input on the console port.
8.15 gnu/usr.bin/send-pr
        ChangeLog Makefile Makefile.in aclocal.m4 configure 
        configure.in fields.texi send-pr-el.in send-pr.sh send-pr.texi 
Upgraded to send-pr 3.95, with our local modifications
8.16 usr.bin most files Sync with 4.4lite2
8.17 sys/arch/sparc/sparc disksubr.c 
Enable writing of disklabels (Many thanks to Jason Thorpe, Gordon Ross
and Matthew Green).
8.18 sys/ufs/mfs mfs_vfsops.c 
Do any pending I/O before trying to unmount, per John Kohl.
8.19 sys/arch/i386/i386 machdep.c svr4_machdep.c linux_machdep.c 
SA_ONSTACK --> SS_ONSTACK
8.20 usr.sbin/rarpd Makefile rarpd.8 rarpd.c 
Make the "/tftpboot/<client-ip-address>" requirement conditional on
-DREQUIRE_TFTPBOOT and disable it by default.
8.21 share/zoneinfo/datfiles updated to tzdata95g

9. Did anyone try my patches from last week to get i386s with a
1542CF installed boot correctly with a GENERIC kernel? The changes work
fine for me, and I haven't heard anything from anyone else...

See you in two weeks, folks.

Alistair
--
Alistair G. Crooks (agc@uts.amdahl.com)                    +44 125 234 6377
Amdahl European HQ, Dogmersfield Park, Hartley Wintney, Hants RG27 8TE, UK.
[These are only my opinions, and certainly not those of Amdahl Corporation]