Linux-Misc Digest #305, Volume #19 Thu, 4 Mar 99 22:13:13 EST
Contents:
Re: Microkernels are an abstraction inversion (Christopher Browne)
Re: Microkernels are an abstraction inversion (Christopher Browne)
Re: Overclocking (was: Re: K6-2 and Linux, Are there any Bug?) (Dang H. Nguyen)
Re: Passing Observation (Timo T.)
Re: Best Free Unix? (why FreeBSD?) (Richard Caley)
Caldera RPMs in RH? ("Aaron Dershem")
Re: Linux Collective ([EMAIL PROTECTED])
Re: Am I stupid or am I stupid. PPP. ALMOST!!! (Gernot Fink)
Re: Public license question (Barry Margolin)
Can I have a shared modem in Lunux ("Aaron Dershem")
Re: UMSDOS on FAT32? (F. Heitkamp)
Re: demand dialing vs. the other demand dialing (Scott Alfter)
Re: More bad news for NT ("deuce")
Re: More bad news for NT (Jim Richardson)
Re: Network Monitoring Script (Yan Seiner)
Re: Public license question (Geoffrey KEATING)
DNEWS4 News Server Mthly Pointer to FAQ (Stephen Pugmire)
----------------------------------------------------------------------------
From: [EMAIL PROTECTED] (Christopher Browne)
Crossposted-To: gnu.misc.discuss,comp.os.linux.advocacy
Subject: Re: Microkernels are an abstraction inversion
Reply-To: [EMAIL PROTECTED]
Date: Fri, 05 Mar 1999 01:01:32 GMT
On 03 Mar 1999 16:55:55 -0700, Craig Kelley <[EMAIL PROTECTED]> wrote:
>[EMAIL PROTECTED] (Stefaan A Eeckels) writes:
>> That's not such a wild idea - if you've got an interface that lets you
>> combine objects into a working program, the objects could be found
>> and made available by the OS, dynamically. They don't have to reside
>> on the local machine either.
>
>Just the act of doing it:
>
>(mail (uniq (sort (cut (file passwd)))))
>
>is simply a semantic one. I would wager that there is already a
>functional shell for UNIX somewhere.
There's not one, but at least *three* of em...
<http://www.ai.mit.edu/projects/su/scsh-projects.html>
<http://cargsun2.atmos.washington.edu/sys/software/e/es.html>
<http://esh.netpedia.net>
--
Coming Soon to a Mainframe Near You! Microsoft Windows NT 6.0, complete with
VISUAL JCL...
[EMAIL PROTECTED] <http://www.ntlug.org/~cbbrowne/unixshells.html>
------------------------------
From: [EMAIL PROTECTED] (Christopher Browne)
Crossposted-To: gnu.misc.discuss,comp.os.linux.advocacy
Subject: Re: Microkernels are an abstraction inversion
Reply-To: [EMAIL PROTECTED]
Date: Fri, 05 Mar 1999 01:01:33 GMT
On 4 Mar 1999 00:21:14 -0500, [EMAIL PROTECTED] <[EMAIL PROTECTED]> wrote:
>>>I'm always irritated when I hear people talk about OO design, when it
>>>is patently impossible to implement such a design on the current
>>>(mainstream) OSes, because these deal with processes and files, not
>>>with objects.
>
>even a turing machine needs to make a distinction between its data and its
>code. you cannot avoid it. so what is so bad about a process / file
>abstraction? it seems like the same thing.
They are two useful abstractions. They are not the only abstractions in
the world, as evidenced by the diverse set of data structures that
Common Lisp gives you.
Look at any computer language.
Whether FORTH, Lisp, C, Modula3, ML, Awk, or Prolog, they all provide
various sets of abstractions. Those being (arguably) amongst the
well-designed languages, they provide fairly consistent sets of
abstractions, unlike languages like C++ or Visual BASIC, where it
appears that bunches of abstractions were just "tossed in," which either
injures expressiveness or maintainability.
It's not a big leap to tie this notion of "sets of abstractions" to the
OS. After all, in order to run programs in the language, its
abstractions have to be mapped onto the abstractions that the OS
provides, and thereafter mapped onto the physical representations
provided by the hardware.
--
Coming Soon to a Mainframe Near You! Microsoft Windows NT 6.0, complete with
VISUAL JCL...
[EMAIL PROTECTED] <http://www.ntlug.org/~cbbrowne/langlisp.html>
------------------------------
From: [EMAIL PROTECTED] (Dang H. Nguyen)
Crossposted-To:
alt.os.linux.slackware,comp.os.linux.development.system,comp.os.linux.hardware,comp.os.linux.setup
Subject: Re: Overclocking (was: Re: K6-2 and Linux, Are there any Bug?)
Date: Thu, 04 Mar 1999 02:54:31 GMT
Reply-To: [EMAIL PROTECTED]
With the P-II you could block out pin21 on the pcb and unlock the
multiplier to overclock. Perhaps it would work with the P-III. Read
http://www.tomshardware.com/
The P-III is essentially a P-II with SSE. So it makes sense that
blocking out pin21 will work again.
On Wed, 3 Mar 1999 20:41:44 GMT, "Tuomo O. Vuolteenaho"
<[EMAIL PROTECTED]> wrote:
>I don't get this. Doesn't PIII have a fixed multiplier? Doesn't PIII450
>run at 4.5 times the bus speed, and 500 at 5.0 times? How can you get
>them to work? (An honest question.)
>
>On Wed, 3 Mar 1999, David wrote:
>
>> i've had a dual pIII 500 machine that was actualy a pIII 500 and a PIII 450
>> clocked up that ran fine. exactly the same results as a real dual 500.
>> didn't notice the 450 getting any hotter either.
>>
>> David
------------------------------
From: [EMAIL PROTECTED] (Timo T.)
Subject: Re: Passing Observation
Date: 4 Mar 1999 02:23:28 GMT
Reply-To: [EMAIL PROTECTED]
On Wed, 03 Mar 1999 06:46:42 GMT, bklimas <[EMAIL PROTECTED]> wrote:
>Could be. More likely, Windows 98 was designed by business people to make
>money while programers designed Linux for flexibilty, productivity and fun.
This certainly shows the fresh attidutes of all you linux freaks.
Is your idea of fun about compiling hours of programs, wondering why does
it crash, why does it segfault, etc?
>Anyway, I am a non-programer, use Linux and am happy with it.
Me too.
I'm happy with it, but I don't like the attidutes of some users, and
secondly, I think it unfriendly.
------------------------------
From: Richard Caley <[EMAIL PROTECTED]>
Crossposted-To: gnu.misc.discuss,comp.unix.advocacy,comp.unix.bsd.freebsd.misc
Subject: Re: Best Free Unix? (why FreeBSD?)
Date: 04 Mar 1999 12:22:50 +0000
In article <7blp98$jjs$[EMAIL PROTECTED]>, C Lamb (cl) writes:
cl> Roll Up, Roll Up! Pedants Corner is doing business!
^^^^^^
Vertex
--
Mail me as rjc not [EMAIL PROTECTED] _O_
|<
------------------------------
From: "Aaron Dershem" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux,comp.os.linux.setup
Subject: Caldera RPMs in RH?
Date: Wed, 3 Mar 1999 09:32:58 -0600
Can I install Caldera RPMs in Red Hat 5.2? I want to add netatalk, but RH
doesn't supply a package. Caldera does, but I'm not sure if it will work
correctly.
Aaron Dershem
------------------------------
From: [EMAIL PROTECTED]
Subject: Re: Linux Collective
Date: Wed, 03 Mar 1999 15:48:08 GMT
In article <[EMAIL PROTECTED]>,
Jose Urena <[EMAIL PROTECTED]> wrote:
>
> I hope that you joined out of your own free will and not because someone
> forced or told you to join ;)
>
> well, I am thinking that it should be the other way around, linux users are
> the rebelds & borgs that have reaquire their free will and are breaking
> free of the MS-Borg collective.
>
On the other hand, maybe the Linux collective is really Species 8642. MS-Borg
doesn't stand a chance.
============= Posted via Deja News, The Discussion Network ============
http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
------------------------------
From: [EMAIL PROTECTED] (Gernot Fink)
Crossposted-To: comp.os.linux.setup
Subject: Re: Am I stupid or am I stupid. PPP. ALMOST!!!
Date: 3 Mar 1999 15:37:31 GMT
enable debugging for the pppd
( insert "debug" in the optionsfile normally /etc/ppp/options)
make the output of your chatscript verbose (chat -v -f ...)
Then view /var/log/messages and /var/log/debug again
much luck
In article <7bil04$b6o$[EMAIL PROTECTED]>,
Alex Armstrong <[EMAIL PROTECTED]> writes:
> I'm having a similar program (pppd dies after connection is established and
> the login phase is complete) and it doesn't work either with the scripts
> I've written for pppd, not with wvdial and not with kppp, either.
> Any ideas?
>
> dizzy wrote:
>> Youre making this *way* more complicated than it has to be
>> Get a program called WvDial go here to read the readme, get the source
> and
>> simply follow the instructions. It dosent get any easier than this
>> http://www.connix.com/~dizzy73/WvDial.htm
>> you've got better things to loose sleep over ;-)
>> rob
>>
>>
>> Sergei Gerasenko wrote:
>>
>> > I've been working on this problem for 4 days already. I've had three
> nights
>> > when I went to bed at 8 a.m.! Here is the problem. I bought a RedHat
>> > distribution of Linux (kernel 2.2.32) a week ago and decided to install
> it
>> > on my notebook (NEC VERSA 4000). Of course, the first thing I wanted to
> work
>> > was PPP (I have a PCMCIA modem). My modem was successfully recognized
> and I
>> > didn't have any problems with that. But I was concerned about PPP
> support.
>> > At first I thought that it wasn't built in the kernel because the PPP
>> > registration lines wouldn't show up at boot up. So, I recompiled the
> kernel
>> > with the support built in. For some reason other features got messed up
> and
>> > I decided to switch back to the old kernel. (You must notice here that
> all
>> > that was going on during the Christmas time when everybody was drinking
>> > booze and having fun). Luckily, it turned out that the old kernel had
> PPP in
>> > modules. With great relief I edited all the scripts (PPP-on and
>> > ppp-on-dialer). Of course they didn't work right away, but ultimately I
>> > successfully passed the login-password phase. I'm sure that I send
> right
>> > responses to my ISP because /var/log/messages goes all the way to
> "serial
>> > connection established" and I also looked through the log line by line.
>> > Nothing suspicious. "Serial connection established" was about the only
>> > encouraging line in the whole log. Then comes the following:
>> >
>> > connect: ppp0 <--> /dev/ttyS2 (pretty lively too:-)
>> > channel ppp0 closing (Oh, oh!)
>> > hangup (SIGHUP)
>> > modem hangup
>> > connection terminated
>> > exit
>> >
>> > I have read a whole bunch of readme's on PPP. Nothing describes
> SIGHUP
>> > in a detailed way. There was a short description of that in PPP-HOWTO,
> but
>> > it didn't help. I have looked through old postings dated all the way
> back to
>> > 1997 and found nothing similar.
>> > The PPP service also starts O.K. according to the log. I have tried
>> > minicom. The same thing. I get to the point when I enter my
> credentials, get
>> > the garbage, exit the program without resetting the modem, type
>> > #pppd -d -detach /dev/ttyS2 38400 & and it goes no further. Something
>> > disconnects the modem all the time. My PPP is 2.2.0, so I'm safe there.
> No
>> > other errors. I'm in despair and about to get on an anti-depressant.
> Anybody
>> > who has a good suggestion, will get a bottle of virtual vodka and even
> drink
>> > it with me (virtually)!
>> >
>> > THANXXXXXXXXXX
>>
>
>
> ------------------ Posted via SearchLinux ------------------
> http://www.searchlinux.com
--
MFG G.Fink
------------------------------
From: Barry Margolin <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux,gnu.misc.discuss
Subject: Re: Public license question
Date: Fri, 05 Mar 1999 01:49:37 GMT
In article <m5FD2.69711$[EMAIL PROTECTED]>,
Christopher Seawood <[EMAIL PROTECTED]> wrote:
>So if copyright was amended in the way you state, then dynamically
>loaded modules cannot legally be considered a derivative work of the
>program that uses them and vice-versa unless code was actually copied.
>If that is true, then aren't RMS' beliefs and theories about the spirit
>of copyright law irrelevant? So unless he specifically adds a section
>to the GPL forbidding the separate distribution and later assembly of
>incompatible components A & B, programs can be dynamically linked against
>GPL'd code w/o fear of tainting. Am I still missing something?
It's not the loading into memory that's a violation, it's the distribution
of A and B, with the intent to have the customer link them together (either
statically or dynamically).
Whether RMS really has a legal leg to stand on, I can't say. The theory is
that the spirit of the GPL and copyright laws should not distinguish these
various ways of getting from X (vendor writes a proprietary program A that
needs to be linked with a GPLed module B) to Y (customer runs A+B, which is
a derivative work of B).
Note that this only applies when the only useful module B is GPLed. If
there's an alternate module B', which can be linked in place of B to
produce equivalent effect, there's no violation. The distribution of A is
not implicitly pulling B with it -- it's the customer's choice whether to
get B or B'. And the GPL doesn't restrict the customer's ability to link B
by themselves. It's only when the customer has no choice but to link in B
that the distribution of A is considered equivalent to distributing A+B.
--
Barry Margolin, [EMAIL PROTECTED]
GTE Internetworking, Powered by BBN, Burlington, MA
*** DON'T SEND TECHNICAL QUESTIONS DIRECTLY TO ME, post them to newsgroups.
Please DON'T copy followups to me -- I'll assume it wasn't posted to the group.
------------------------------
From: "Aaron Dershem" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.networking
Subject: Can I have a shared modem in Lunux
Date: Wed, 3 Mar 1999 09:48:35 -0600
Can I set up a home network and put the modem in the Linux server, but have
a Win98 machine use the modem on a dial-on-demand system? I'd like to call
my ISP using Win98, but have the modem on the Linux server. Does that make
sense? I'm using RH 5.2, if that helps any.
Thanks,
Aaron Dershem.
------------------------------
From: [EMAIL PROTECTED] (F. Heitkamp)
Subject: Re: UMSDOS on FAT32?
Reply-To: [EMAIL PROTECTED]
Date: 5 Mar 1999 01:54:57 GMT
In message <7bn9c8$sh7$[EMAIL PROTECTED]> - Kent Robotti
<[EMAIL PROTECTED]> writes:
>
>[EMAIL PROTECTED] wrote:
>> I currently have an UMSDOS partition on a win95b partition.
>> I've been thinking about converting the partion to fat32
>> so that I might gain a extra couple hundred MB of storage.
>> Does anyone know if you can put a UMSDOS partition on fat32?
Is there any more information available? And don't tell me it's
in the kernel sources. ;)
>
>Yes!
>
Fred
------------------------------
From: [EMAIL PROTECTED] (Scott Alfter)
Subject: Re: demand dialing vs. the other demand dialing
Date: 4 Mar 1999 17:33:57 -0800
In article <LlgD2.1042$[EMAIL PROTECTED]>,
Alan Curry <[EMAIL PROTECTED]> wrote:
>OK, here's what happens: at boot time, the loopback and ethernet devices are
>ifconfiged and routed. A blanket masq rule is added with ipfwadm. The local
>network works. Other local machines can be pinged. Then comes the
>"pppd demand defaultroute" (which lots of other options too but those are the
>big ones) and pings across the ethernet stop working. The local route is
>still in the routing table, but nothing makes it through. killing pppd causes
>the local net to start working again.
Are you using private IP addresses (192.168.x.y) for your LAN? If you're
not, it could cause some weird problems. I have my three machines set up as
192.168.100.1 (a K6-200 running Linux, ip-masq, and diald), 192.168.100.2 (a
K6-2-300 running Win98 connected by Fast Ethernet), and 192.168.100.3 (an
Apple IIGS connected by PPP to a serial port on the Linux box), and
communication between any of the three works whether the dial-up 56K link is
up or down. Traffic on your local network shouldn't be affected by the
dial-up link.
_/_
/ v \
(IIGS( Scott Alfter (salfter at (bitte keine Spam) delphi dot com)
\_^_/ http://people.delphi.com/salfter
------------------------------
From: "deuce" <[EMAIL PROTECTED]>
Crossposted-To: alt.destroy.microsoft,comp.os.linux.advocacy,alt.linux
Subject: Re: More bad news for NT
Date: Thu, 4 Mar 1999 19:55:35 -0600
Actually the beta format suffered because a consortium of companies
refused to licence the patented betamax format owned by Sony. This
resulted in the creation of the VHS format. The Beta system f(or eg.
beta hifi and beta ed) are really quite high end and where in many instances
used by television companies in their hayday.
------------------------------
From: [EMAIL PROTECTED] (Jim Richardson)
Crossposted-To: alt.destroy.microsoft,comp.os.linux.advocacy,alt.linux
Subject: Re: More bad news for NT
Date: 4 Mar 1999 02:59:54 GMT
Reply-To: [EMAIL PROTECTED]
On Tue, 02 Mar 1999 11:28:31 -0600,
Tim Kelley, in the persona of <[EMAIL PROTECTED]>,
brought forth the following words...:
>
>
>Jon Wiest wrote:
>>
>
>> As for Netscape, it does lock up Linux. I have locked up Linux, not with
>> any server apps, yet though. Couple of DOOM sessions tells me Linux is
>> pointless as a DESKTOP.
>
>Never seen Netscape lock up "linux" - did you try just typing "killall
>-9 netscape", or using "ps" to get the process id and killing it that
>way?
>
>Are you familiar with the kill command?
>
>One particularly annoying thing about Netscape is that it is always
>contacting my DNS server for no apparent reson. When I'm not connected
>I can't use netscape because it will take forever to start ...
>
I have this problem also, the simplest solution (for me, with the laptops)
is to turn off the eth0 interface, then dns helper aborts right away and gives
me back the netscape window. (not a very elegant solution, and no good if you
need that interface for something else. But...) I think that NS4.5 ignores
the resolv.conf, since even with hosts first, it does this when I try to access
a machine in my hosts file.
------------------------------
From: Yan Seiner <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.networking
Subject: Re: Network Monitoring Script
Date: Wed, 03 Mar 1999 21:51:49 -0500
Why not just set your firewall to log? Or do you need additinal info?
Yan
Brian Ferris wrote:
> I am using a RedHat 5.2 server to masquerade an 80+ user office
> to provide Internet access. A major sticking point with my boss is
> that the machine must be able to log Internet usage among employees.
> I have tackled this problem from a number of angles and I finally
> decided that I would write a shell script to handle the job (I mainly
> chose this because I wanted to learn shell programming). The script
> is a follows (it is run out of my home directory <bferris> right now
> so files are based there)...
>
> #!/bin/sh
> # This script monitors Internet usage by monitoring the
> /proc/net/ip_masquerade file. This file
> # stores info in the following format...
> # type (TCP,UDP,etc.) source_ip:port dest_ip:port masq_port
> # Ip addresses are in hex format. Linux uses this file to keep track
> of which ports it masqued
> #addresses to.
>
> tcp_only="TCP"
>
> # I use a temporary file <addition> to store the results of the
> script. The first thing added to this
> # file is the date.
> echo $(date) > /home/bferris/addition
>
> #I use a read to pull the type, source_ip, and dest_ip from
> /proc/net/ip_masquerade
> while read type source dest rest; do
>
> # I check to see if the type is TCP since I only want this
> type
> if [ $type = $tcp_only ]; then
>
> # I wrote a quick and dirty cpp prog to convert the
> hex IP address into the regular x.x.x.x format
> source=$(/usr/local/bin/hextodec $source)
> dest=$(/usr/local/bin/hextodec $dest)
>
> # To avoid having to make a nslookup call for every
> IP, I created a list of addresses and
> # their host names in the host_list file. If the
> address hasn't been added to the list yet...
> if dest_name=$(grep $dest /home/bferris/host_list) ;
> then
> adder="$source $dest_name"
> else
> # ... I make a nslookup to find the server
> name.
> dest_name=$(nslookup $dest | grep Na | cut -c
> 7-)
> # If the lookup doesn't come up with anything,
> the name becomes unknown
> if [ -z $dest_name ]; then
> dest_name="Unknown"
> fi
> adder="$source $dest $dest_name"
>
> # Since this server wasn't in the host_list
> file, I add it
> echo $dest $dest_name >> host_list
> fi
>
> # I check to see if I have added the $adder to the
> addition file yet. I do this because
> # a website can be masq as many as ten times to the
> same IP in a normal ip_masquerade
> # listing. There is no point in adding it ten times
> so I just add it once.
> if ! grep -q "$adder" /home/bferris/addition ; then
> echo $adder >> /home/bferris/addition
> fi
>
> fi
> done < /proc/net/ip_masquerade
>
> # When I'm all done, I have a time stamped file that lists who's been
> using the net. This file
> # can be appended to the final log file which I will then let my boss
> look at so he can deal out
> # sweet justice.
> more /home/bferris/addition
>
> <EOF>
>
> First, let me say that all I know about shell scripting came from
> this project. This script works but I have a feeling that there might
> be a better, more efficient way of getting the same results. This
> script would currently have to be run by cron every once in a while to
> properly log. With that in mind, I will now ask all my questions
> -Is this script the answer or am I barking up the wrong tree?
> -Should I do something with CPP that would run 100x faster?
> -Should I just bite the bullet and install a proxy like squid
> which does logging too?
> -Should I do something crazy like install my own local DNS server
> and do logging here (I want to set up a DNS for fun anway...)?
>
> I know that is a lot of questions so any thoughts the world has on
> the subject is much appreciated. Besides, the whole thing can
> definitely be chocked up as a learning experience if nothing comes of
> it.
>
> Thanks,
> Brian Ferris
>
>
>
>
------------------------------
Crossposted-To: gnu.misc.discuss
Subject: Re: Public license question
From: Geoffrey KEATING <[EMAIL PROTECTED]>
Date: 04 Mar 1999 15:41:42 +1100
Mark Mokryn <[EMAIL PROTECTED]> writes:
> Okay, guys... as the originator of this thread, I must protest - this
> discussion is not what I intended. I assume that at least the vast majority
> of us, if not all of us, are engineers, and not (thank god) lawyers. So let's
> cut the legalese bullshit, since that's exactly what it is. Please, let's
> talk in more concrete terms we can all understand.
Um. You want to talk about copyright law, but without using any legal
technical terms? That's like a copyright lawyer saying "can we please
stop talking about this computer stuff and get back to the law?"
> My original question pertained to the intent of the GPL. Can one indeed make
> money on a project based on the Linux platform? And I do not mean to make
> money the Red Hat way, i.e. open up the source code and sell support. For
> argument's sake, let's say I intend to build a proprietary box (e.g.
> super-duper new video server) running on some OS, possibly Linux. Of course,
> an open-source OS can really help, and I can promise to release any
> modifications to the existing OS platform. I would most certainly not want to
> release the codebase to my project, however.
Well, if you plan to release your box as a whole thing, and you
include GPLed software, then you must include source to the whole
thing. That's the point of the GPL.
> Worrisome to me is the following, clipped from Section 2 of the GPL: ...
> In other words, if I sell my super-duper video server box, I may be forced to
> release the code to the modules I wrote entirely by myself? Is this the
> intent of the GPL, to force the little guy to seek another platform?
No. The intent of the GPL is to encourage people to make their
software free. If you don't want to make your software free, then the
best the GPL can do is try to exert more pressure and hope you will
change your mind.
> Why doesn't the FSF provide simple, CLEAR, for god's sake,
> guidelines, so I may know what I am getting into? The ambiguity
> hurts us, and not the big guys.
The FSF does provide a clear guideline. They want you to make your
code free. If you don't want to make your code free, there is no
reason for them to help you do this (arguably it's even against their
charter). But they are very happy to help you make your code free, and
will do whatever they can.
If you really want to make sure someone can't reverse-engineer your
incredibly clever code and produce a proprietary product based on it,
why not apply for a software patent on it, and licence the patent to
anyone who releases their code under the GPL? [I can't believe I just
recommended applying for a software patent :-s.]
This whole "derived work" stuff is really very simple.
Here's how to make a derived work: Take some substantial code from
GCC, or from the linux kernel, or whatever. Copy-and-paste it into
your program. You now have a derived work.
Here's how to not make a derived work: Think hard. Make your own
solution to the problem. Type it into your program.
You see the difference? It's nothing to do with the thing that's
produced; they may turn out the same. It's how they were made. Of
course, it's much easier to prove that wasn't copied if they are
substantially different, but that's not the essential thing.
--
Geoff Keating <[EMAIL PROTECTED]>
------------------------------
From: [EMAIL PROTECTED] (Stephen Pugmire)
Crossposted-To:
news.software.nntp,comp.os.ms-windows.nt.admin.networking,comp.infosystems.www.servers.ms-windows,comp.infosystems.www.servers.misc
Subject: DNEWS4 News Server Mthly Pointer to FAQ
Date: Fri, 05 Mar 1999 02:26:23 GMT
This message is a brief description of the DNEWS 4 news server
features and a pointer to the FAQ and binary areas.
DNEWS is a high performance NNTP News Server. It is suitable
for large and small ISP's and Corporate Intranets. DNEWS is
available for all common operating systems, installs easily and
is designed to run unattended.
Design :
Database : DNEWS incorporates a database specifically
designed for News, enabling DNEWs to handle full news feeds
more efficiently and faster than traditional news server
software. This is important as news volumes
continue to grow rapidly. DNEWS can provide significant
performance improvements particularly where resources are
limited.
Dynamic Sucking Feed Option: In addition to traditional news
feed options DNEWS also provides the unique dynamic feed
option, which for business and smaller sites delivers
enormous savings in Network bandwidth and diskspace.
By dynamically reflecting users reading habits the server
automatically only gets and maintains those news
groups that users actually want to read. In this way users
retains all the advantages of the full news feed while
enormous saving in network bandwidth and diskspace are
possible.
The process is completely automatic so there is no
ongoing administration .
News Reader Efficiency: The advanced design significantly improves
news server efficiency. On similar Hardware DNEWS will
typically support 2 times as many news readers per server
with the same fast interactive response time.
Scalability: Designed for high scalability DNEWS design will
handle 0-60 Gig news spools most efficiently. Can be
configured with 2-10 Multiple Server processes on single
server each handling 50-200 concurrent users.
DNEWS can be configured to take multiple full news feeds
and up to 16 dynamic suck feeds and can efficiently send
multiple full feeds with built-in Live feed technology.
Easy Installation and Management: Designed to run unattended
DNEWS is easy to install with virtually no ongoing management.
Key Features:
Easy Installation - totally configurable without recompiling.
Low Maintenance - designed to run unattended.
Support for Multiple IHAVE Feeds
Support for Multiple Dynamic Suck Feeds
Built in Live Feeds Technology for sending full news feeds to other
servers efficiently.
News to Web Gateway , for putting news onto web pages , with full
text searching .
New to Mail Gateways (both ways)
PGP authentication (optional)
Flexible Access and Security options for reading and posting.
Flexible Expire Options allowing intelligently response to number of
items and available diskspace.
Advanced 'SPAM' detection and filtering mechanisms.
Email confirmation of posts.
XOVER extensions implemented in software removing the need for a
separate ' XOVER' database
Streaming extensions are built in speeding up feed throughput
Full Documentation
Commercially supported
DNEWS is a not FREEWARE. Registration is free to state schools and
universities for non profit purposes.
Web page: http://netwinsite.com/
FTP Binaries ftp://netwinsite.com/dnews
FAQ ftp://netwinsite.com/dnews/install.txt
Mirror Site ftp://ftp.std.com/ftp/vendors/netwin/dnews/
Email: [EMAIL PROTECTED]
------------------------------
** FOR YOUR REFERENCE **
The service address, to which questions about the list itself and requests
to be added to or deleted from it should be directed, is:
Internet: [EMAIL PROTECTED]
You can send mail to the entire list (and comp.os.linux.misc) via:
Internet: [EMAIL PROTECTED]
Linux may be obtained via one of these FTP sites:
ftp.funet.fi pub/Linux
tsx-11.mit.edu pub/linux
sunsite.unc.edu pub/Linux
End of Linux-Misc Digest
******************************
