ntp-4.0.96 is available
Harlan Stenn
ftp://ftp.udel.edu/pub/ntp/testing/
The only pending items in the queue are:
- a multiprocessor NT patch that should be here soon
- a couple weeks' stability
From the NEWS file:
(4.0.96)
* -lnsl, -lsocket, -lgen configuration patches
* Y2K patches from AT&T
* Linux portability cruft
From the ChangeLog:
1999-08-09 Harlan Stenn <st...@whimsy.udel.edu>
* configure.in: 4.0.96
* parseutil/dcfd.c (main): Replace SA_ONSTACK and SV_ONSTACK with
HAVE_SIGACTION and HAVE_SIGVEC, respectively. HP-UX provides both
of the former but only one of the latter...
1999-08-08 Harlan Stenn <st...@whimsy.udel.edu>
* configure.in: Better tests for -lnsl and -lsocket
From: Albert Chin-A-Young <ch...@thewrittenword.com>
Works for me - handle openlog() and -lgen the same way.
* Makefile.am (EXTRA_DIST): Add in the y2k notes
* parseutil/dcfd.c: Renamed drift_comp to accum_drift
* configure.in: Added MAKE_CHECK_Y2K support; check_y2k needs libparse.
* ntpd/Makefile.am (check_PROGRAMS): Use MAKE_CHECK_Y2K
* ntpd/Makefile.am (check-local): Added.
* parseutil/Makefile.am (check-local): Added.
* include/ntp.h: Y2KFixes
* libparse/parse.c: Ditto
* ntpd/Makefile.am (check_PROGRAMS): Ditto
* ntpd/refclock_acts.c: Ditto
* ntpd/refclock_arc.c (arc_receive): Ditto
* ntpd/refclock_heath.c: Ditto
* ntpd/refclock_hpgps.c: Ditto
* parseutil/Makefile.am (check-local): Ditto
* parseutil/dcfd.c (check_y2k): Ditto
* NOTES.y2kfixes: Ditto
* readme.y2kfixes: Ditto
* results.y2kfixes: Ditto
* ntpd/check_y2k.c: Ditto
From: y...@y2k.labs.att.com
1999-08-07 Harlan Stenn <st...@whimsy.udel.edu>
* configure.in: Look for sys/ppstime.h.
1999-07-31 Harlan Stenn <st...@whimsy.udel.edu>
* ntpd/ntp_io.c (create_sockets): Typo.
From: Doug Wells <d...@contek.com>
1999-07-29 Harlan Stenn <st...@whimsy.udel.edu>
* configure.in (ac_cv_struct_ntptimeval): Explicitly look for
struct ntptimeval.
(ac_cv_var_kernel_pll): Require struct ntptimeval.
Linux. Grrr.
Reported by: Ronald Kuetemeier <ron...@smginc.com>
Carl Byington
NT binaries available at
http://www.five-ten-sg.com/util/ntp4093a.zip
http://www.five-ten-sg.com/util/ntp4095.zip
http://www.five-ten-sg.com/util/ntp4096.zip
However, the 95 and 96 versions don't work well for me. I am still
running 93a.
The later versions make my clocks slowly diverge, but it would be
interesting to see if that is true on other systems as well.
-----BEGIN PGP SIGNATURE-----
Version: 4.5
iQCVAgUBN6+LK9ZjPoeWO7BhAQE4oAP/YEo56hmQ4DEwPjI+BlFSZLBoHpl+OpZU
tbVhButVMAns8QNDgmuuy32AMqMcVlPpwCCkwVSEc0rKYkVsbyaEDCsbc0HauznL
s4nbIz2jtZDc/qiJNkWZIU7TUUidDOHXJamI1qUpatvMgbM7KoIP/cnkKa1e2ZHp
SFOOqlW+AkM=
=Mmte
-----END PGP SIGNATURE-----
Harlan Stenn
A possible fix is being tested.
H
Carl Byington
In article <o4emhb2...@whimsy.udel.edu>, st...@whimsy.udel.edu says...
>
>Apparently, the clock diversion problem is related to the NT port only.
>
>A possible fix is being tested.
And it seems to work - *very* well actually. New 4.0.96 binaries at
<http://www.five-ten-sg.com/util/ntp4096.zip> built on the morning of
August 10.
- --
PGP key available from the key servers.
Key fingerprint 95 F4 D3 94 66 BA 92 4E 06 1E 95 F8 74 A8 2F A0
-----BEGIN PGP SIGNATURE-----
Version: 4.5
iQCVAgUBN7IggtZjPoeWO7BhAQFUlwP/W6JdxoPZ9uGHC69sq9Iv6akvVbLgbwrR
7sFpQMUvcw/msT+hPx4ODEUfbKLpeCqNdRKjNoWpZgL9Iow1v3npEiIG+3ZG1Xk+
YxPT8/WH2mWPRJkIMCfLB/k9yMiw2e1hrPL4WwMHx8r1ukvjJzH/HIVOstHYpVjx
R+4ejU3frhA=
=HGGK
-----END PGP SIGNATURE-----
Andrej Borsenkow
Carl Byington <ca...@five-ten-sg.com> wrote in message
news:7ot7ah$ijv$1...@la-mail4.digilink.net...
>
> And it seems to work - *very* well actually. New 4.0.96 binaries at
> <http://www.five-ten-sg.com/util/ntp4096.zip> built on the morning of
> August 10.
>
Two questions.
1. Which authentication is supported? I tried keys of type A and M - ntpd
reports "invalid key type for key number 1, entry ignored ".
2. I get the following messages when starting ntpd (I did not have them with
4.0.93):
Unusual Clock increment 156250: Multiprocessor?
(Yes, it is dual processor system)
Wraparound 1225108046 -> 475681The operation completed successfully.
(only once so far)
Is there something to worry about?
regards
/andrej
Carl Byington
In article <7pbpp6$dir$1...@news.mch.sbs.de>, Andrej.B...@mow.siemens.ru
says...
>
>1. Which authentication is supported? I tried keys of type A and M - ntpd
>reports "invalid key type for key number 1, entry ignored ".
I don't think any authentication is supported in this port since I don't
have a license for the RSAREF library. All the encryption stuff was pulled
from the distribution due to crappy US export policy.
>2. I get the following messages when starting ntpd (I did not have them
with
>4.0.93):
>
>Unusual Clock increment 156250: Multiprocessor?
>(Yes, it is dual processor system)
>
>Wraparound 1225108046 -> 475681The operation completed successfully.
>(only once so far)
Not sure. The folks at Trimble did the mods to use the performance counters
to interpolate between NT 10ms ticks. These seem to be debug messages from
that code - which is still being worked on.
- --
PGP key available from the key servers.
Key fingerprint 95 F4 D3 94 66 BA 92 4E 06 1E 95 F8 74 A8 2F A0
-----BEGIN PGP SIGNATURE-----
Version: 4.5
iQCVAgUBN7nyjNZjPoeWO7BhAQElngP7BlKiXT1bzCR+Le2Q9S6d499tWMY8X+rt
hPvb0nOttGKxHYhzKkaSo6ey30h/5PVY0s+QhRtRhj5kWIrjeNCfqNyhxpslKHQp
r0w1UbjL+37ZTxEHFISNTjbXIdz33pSxpUMMyp3P1ayd3ttYAzZM+kIaptbYLghx
HtkuCk+Unb8=
=5Njy
-----END PGP SIGNATURE-----
David Schwartz
>
> -----BEGIN PGP SIGNED MESSAGE-----
>
> In article <7pbpp6$dir$1...@news.mch.sbs.de>, Andrej.B...@mow.siemens.ru
> says...
> >
> >1. Which authentication is supported? I tried keys of type A and M - ntpd
> >reports "invalid key type for key number 1, entry ignored ".
>
> I don't think any authentication is supported in this port since I don't
> have a license for the RSAREF library. All the encryption stuff was pulled
> from the distribution due to crappy US export policy.
Odd. I thought that:
1) MD5 was fine for export, and
2) If it's only used for authentication, it isn't considered encryption
under US law.
DS
David Schwartz
> Yeah, sort of what I thought. But maybe UDel is just playing it real safe.
> Of course, you could use MD5 to build a fairly secure stream cipher.
Perhaps then they should simply apply for a license. It's really not
that complicated, and I'm 99% certain that it would be granted.
DS
Carl Byington
In article <37B9FF24...@webmaster.com>, dav...@webmaster.com says...
>
> Odd. I thought that:
>
> 1) MD5 was fine for export, and
>
> 2) If it's only used for authentication, it isn't considered
encryption
>under US law.
>
Yeah, sort of what I thought. But maybe UDel is just playing it real safe.
Of course, you could use MD5 to build a fairly secure stream cipher.
- --
PGP key available from the key servers.
Key fingerprint 95 F4 D3 94 66 BA 92 4E 06 1E 95 F8 74 A8 2F A0
-----BEGIN PGP SIGNATURE-----
Version: 4.5
iQCVAgUBN7opH9ZjPoeWO7BhAQFu/QP/RmGCEoCmsD5rGW8H91kNA4UU+mFsDZjS
keXg98+AlQ/Wy0geXTnO3ZZveTl8+h4h1pqBEkBakZ/T8l2WVixsB8U7xcYxC5Lf
zt+7MyAQvAq1Gpq2kxZ+IzYbAIevCKKe60JuWI5MxX+iHR+rEGf6Ux8yNDCv4J9q
Qr1xhLFLvww=
=mjZM
-----END PGP SIGNATURE-----
Harlan Stenn
H