Linux-Setup Digest #478, Volume #19 Sat, 26 Aug 00 04:13:08 EDT
Contents:
Re: Linux, XML, and assalting Windows (Christopher Browne)
Re: Linux, XML, and assalting Windows (Christopher Browne)
Re: Do I Need a Firewall (Eiki Martinson)
Re: first time user (Eiki Martinson)
Monitor ip masq? ("Devon Harding")
Re: do I have to reboot to change network settings ([EMAIL PROTECTED])
Re: Partition Size Advice (hac)
Help!!! Cant get Linux to install, tried 4 distros ("SOME1")
Re: Linux, XML, and assalting Windows (Ian Pulsford)
Re: Linux, XML, and assalting Windows (Ian Pulsford)
Get rid of LILO in the MBR ("Stephan J")
Re: Linux, XML, and assalting Windows
Re: Linux, XML, and assalting Windows (Ian Pulsford)
Re: Monitor ip masq? (John Hovell)
Stupid Question - 1 ("GreatFree")
----------------------------------------------------------------------------
From: [EMAIL PROTECTED] (Christopher Browne)
Crossposted-To: alt.os.linux,comp.text.xml,comp.os.linux.misc,comp.os.linux.advocacy
Subject: Re: Linux, XML, and assalting Windows
Reply-To: [EMAIL PROTECTED]
Date: Sat, 26 Aug 2000 03:45:27 GMT
Centuries ago, Nostradamus foresaw a time when The Ghost In The Machine
would say:
>In comp.os.linux.advocacy, mlw
><[EMAIL PROTECTED]>
>>That's all that XML is, nothing more. It can not replace programs, it is
>>not a new concept in operating systems.
>
>It might replace programs (programs are interpreted data in their
>own right, after all -- to the right interpreter, such as an x86
>micro, a JVM, or even a BASIC environment), but it sure looks
>hard to manage, although not too hard to generate.
It only "replaces" programs if it can express programs itself.
Note that providing the ability to _embed_ programs is not that;
that merely replaces one language with another.
>But why can't we use a schema/data approach? Something like:
>
>first 8 bytes - magic signature number, just because
>byte - endianity
>byte - user-defined version ID
>2 bytes - number of fields
>field descriptor byte: 0=short, 1=long, 2=float, 3=double,
> 4=zero-terminated string
>field name: zero-terminated string
>field descriptor byte:
>field name:
>...
>
>(The floats would be in IEEE format, which is the one 680x0 and
>80x86 micros use -- and possibly a large number of other computer
>systems.)
>
>Surely somebody out there's thought of a standard for this.
There's not one; there's several.
Leaping to mind are:
a) IIOP - the Internet protocol defined for CORBA that does
essentially what you describe, albeit a _little_ differently;
b) Casbah's LDO (Lightweight Distributed Objects)
>Or one can use a chunky format, something a la Amiga's IFF,
>where data is in chunks, understood by each program. Chunks
>could even have DTD-like structures if necessary.
>
>But nooooooo....we get to clutter up what is essentially a
>data-centric stream with a lot of framing clutter. Unless
>I'm missing something in the DTD spec which allows for the
>specification in binary of all of this data...?
I think WAP provides some such mapping...
--
[EMAIL PROTECTED] - <http://www.ntlug.org/~cbbrowne/corba.html>
"What did we agree about a leader??"
"We agreed we wouldn't have one."
"Good. Now shut up and do as I say..."
------------------------------
From: [EMAIL PROTECTED] (Christopher Browne)
Crossposted-To: alt.os.linux,comp.text.xml,comp.os.linux.misc
Subject: Re: Linux, XML, and assalting Windows
Reply-To: [EMAIL PROTECTED]
Date: Sat, 26 Aug 2000 03:45:32 GMT
Centuries ago, Nostradamus foresaw a time when Bob Hauck would say:
>On Fri, 25 Aug 2000 22:40:10 GMT, [EMAIL PROTECTED]
><[EMAIL PROTECTED]> wrote:
>
>[snip miracle-working]
Indeed. All the problems of interoperability, complex systems, and
programming, resolved into a simple matter of a little XML, all in
a couple hundred lines.
>>Let�s try to discuss what it is going to take to make the above
>>happen.
>
>Oh, you'll be having a talk about witchcraft then?
>
>[snip a few hundred lines of handwaving]
Handwaving that is at least vigorous enough to keep someone off the
ground for a couple of seconds, given a trampoline and a 50 foot cliff...
--
[EMAIL PROTECTED] - <http://www.hex.net/~cbbrowne/>
"sic transit discus mundi"
-- From the System Administrator's Guide, by Lars Wirzenius
------------------------------
From: Eiki Martinson <[EMAIL PROTECTED]>
Subject: Re: Do I Need a Firewall
Date: Fri, 25 Aug 2000 23:58:09 -0400
Reply-To: [EMAIL PROTECTED]
Cory Phillips wrote:
> I've been using Linux for almost 2 years now. I mainly use it as a
> development platform on my home desktop computer. I use a dial-in ppp
> connection to the Net and my IP address is dynamic. I know my system
> is accissible to anyone on the Net via the IP address. I'm able to
> telnet to my desktop computer via the dynamic IP from another computer
> on the Internet.
>
> Should I use a firewall to help protect my system? If so, any good
> recomendations?
>
> Thanks,
>
> Cory
An additional recommendation:
Do you need to be able to login as root remotely? If not then disable
such access. Cut out all services you don't need.
--
Eiki Martinson - [EMAIL PROTECTED]
------------------------------
From: Eiki Martinson <[EMAIL PROTECTED]>
Subject: Re: first time user
Date: Fri, 25 Aug 2000 23:50:43 -0400
Reply-To: [EMAIL PROTECTED]
�e�J@Office wrote:
> It's not the first time I use Linux, but is the first time I install
> Linux and also with a computer build on scratch. I'm wondering if
> Redhat 6.2 support "Promise Ultra66 PCI Controller Card" and also Voodoo
> 3 2000 (or 2200, I forgot the number).
>
> Also, when I want to install Linux on a blank hard drive, should I use
> the Linux boot disk created from the CD to boot it? Can the hard drive
> be connected to the Promise card when I install Linux, since nothing is
> on the hard drive, so no driver for the card.
>
> --
> Anson
To the best of my knowledge (possibly wrong), no hard drive interface
should require drivers. Consider this: When you boot up your computer, the
driver is not loaded yet. That means you have to get it off of the hard
drive. But the hard drive is inaccessible if the interface doesn't already
work! Therefore the card should work without drivers.
Pretty much all Voodoo cards are supported now; certainly you can get an X
server for it. I don't really know about using the acceleration for OpenGL
or such, but I think that too can be done.
Use the boot disk you created from the CD to boot it up for installation.
Afterwards, you can continue using other boot disks, or install LILO, which
is a MBR program on your hard drive which lets you boot linux without a
disk.
Eiki Martinson - [EMAIL PROTECTED]
------------------------------
From: "Devon Harding" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.misc,comp.os.linux.networking
Subject: Monitor ip masq?
Date: Sat, 26 Aug 2000 00:28:03 -0400
How can I monitor IP Masq. traffic, such as which web site each user is
going to?
-Devon
------------------------------
From: [EMAIL PROTECTED]
Subject: Re: do I have to reboot to change network settings
Date: Sat, 26 Aug 2000 04:44:08 GMT
In article <8o65in$1k5$[EMAIL PROTECTED]>,
[EMAIL PROTECTED] (Peter Bismuti) wrote:
>
> Right now I boot my machine, run netconf, and then reboot, is there a
faster way?
> I tried booting into single user, running netconf, and then running
init 3, but this
> does not work.
>
> Thx
>
>
try running the following scripts
# /etc/rc.d/init.d/network stop
# /etc/rc.d/init.d/network start
Sent via Deja.com http://www.deja.com/
Before you buy.
------------------------------
From: hac <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.hardware,linux.redhat.install
Subject: Re: Partition Size Advice
Date: Sat, 26 Aug 2000 05:02:27 GMT
John Beardmore wrote:
>
> In article <[EMAIL PROTECTED]>, hac <[EMAIL PROTECTED]>
> writes
> >John Beardmore wrote:
> >>
> >> So is this analogous to cluster size / directory entry issues in FAT
> >> file systems ?
> >>
> >
> >Unlike FAT, it has nothing to do with partition size limits.
> >
> >It's more of a performance issue. A larger block size, particularly
> >one related to the processor's 'natural' block size, can give you
> >faster reads and writes.
>
> OK, but what's a natural block size ?
>
A block size that the hardware handles with minimum fuss. For a PC,
disk access is most efficient using DMA. But PC DMA accesses physical
memory[1], not virtual memory. So the memory management hardware gets
involved in the mapping of physical to virtual. The X86 VM page size
is 4096 bytes. Making the filesystem block size the same as the VM
page size is common practice for Unix systems.
Unlike Windows, Linux runs on other processor types than just X86.
Those other processors - Alpha, SPARC, PowerPC, etc. - may have
different natural block sizes. I know very little about X86 memory
management, and even less about other processors. Don't ask me what
the best block sizes are for other platforms.
The executive summary: The hardware works better with certain block
sizes. The size depends on the hardware. For X86, it's 4k.
> >The percentage of space reserved for root can be adjusted from the
> >default 5%. After an ordinary user fills the disk, root still has
> >some space to work with to clean up. The side-effect, and an
> >important one, is that fragmentation goes up markedly as the disk
> >approaches full. The reserved space helps keep the fragmentation
> >down.
>
> Are there any defragmentation tools for Linux ?
Yes. But you don't need them. You'd have to work quite hard to get
an ext2 filesystem fragmented enough to notice the slowdown. Cleaning
the lint out of your mouse is a bigger problem than fragmentation in
Linux filesystems.
Just because Windows does some really stupid things is no reason to
expect any other OS to do them. Robust, fast filesystems have been a
solved problem for decades. Microsoft chose to ignore the widely
published solutions and research, and has been slapping Band-Aids on
the DOS filesystem without ever replacing it. They do know better;
NTFS is decent.
[1] Just to further complicate matters, the PC DMA hardware can only
access the lowest 16MB of memory. The ISA bus can only access the
lowest 1MB. This is reason enough to rid of the ISA bus - bad
hardware design causes software complications. The X86 is bad enough
without crippling it further.
--
Howard Christeller Irvine, CA [EMAIL PROTECTED]
------------------------------
From: "SOME1" <[EMAIL PROTECTED]>
Subject: Help!!! Cant get Linux to install, tried 4 distros
Date: Fri, 25 Aug 2000 22:21:19 -0700
I have a:
1) 486/66 (Genuine Intel)
2) 24 Ram (SIMM)
3) a 2 gb hd, a .5 gb hd
4) Cirrus Logic generic 2 mb video card
No SCSI
My Problem: With Redhat 6.2, my boot floppies absoulutely refused to work,
With Slackware 4.0, I couldn't get the hard drive to
partition, I would set the table, everything would be going fine, but it
wouldn't actually format the partitions.
With OpenBSD 2.3(I think, 2 lazy to go look) It was
the same thing (Could they have made that install any less user-friendly
; ))
With Storm 2000, It would actually start installing,
but at some point it would just stop. Not freeze, I could still move the
windows, mouse etc, but the install stopped.
I have access to a Mandrake 6.1, Corel and some European version if any
thinks I should just try another distro.
Any advice anyone has would be greatly appreciated.
Also, I feel kind of stupid asking this, but how do u use the C compiler,
the MAKE program and all those other installer type things. Im new to linux
and come from a Microsoft-infected background (in fact, my parents wont even
let me set our 650 mhz box up to dual boot) so I really, reallllllllly need
help.
Thanks,
D-503
"How could I have found this so difficult before? It is clear
now what I must do. The only explanation is my former sickness (the soul)."
-D-503, after the operation which removed
his
imagination, thus making him a slave to society.
from "We" by Yvegnev Zamyatin
------------------------------
Date: Sat, 26 Aug 2000 15:34:36 +1000
From: Ian Pulsford <[EMAIL PROTECTED]>
Crossposted-To: alt.os.linux,comp.text.xml,comp.os.linux.misc,comp.os.linux.advocacy
Subject: Re: Linux, XML, and assalting Windows
[EMAIL PROTECTED] wrote:
> Joseph T. Adams <[EMAIL PROTECTED]> wrote in message
> news:8o6lm6$f4j$[EMAIL PROTECTED]...
>
> > : <MSWORDDATA>
> > : sdlkfjsda;lfkjsda f;lsdkf 2340985r32j23lkr2j r23o978xdf0sdjalfkwj 32
> > : </MSWORDDATA>
> >
> > : (where that contains some encoded table)
> >
> >
> > That will parse just fine.
> >
> > It won't be very useful to anyone but M$. But XML never promised to
> > make all data useful.
>
> If the data cannot be interpreted, then what is the use of parsing it? If a
> propertary data file were converted to text as a Radix-64 or uuencoded data
> stream and is framed by valid tags it is still no more useful that what we
> have today. If in fact it would be less useful, processing it would be
> slower and it would consume more storage and other resources than otherwise
> needed.
Well you can do whatever you like with the data; you might want to encrypt it
because it contains your customer's credit card details, but the format of the
data file is still standard and that is what matters.
IanP
------------------------------
Date: Sat, 26 Aug 2000 15:40:54 +1000
From: Ian Pulsford <[EMAIL PROTECTED]>
Crossposted-To: alt.os.linux,comp.text.xml,comp.os.linux.misc,comp.os.linux.advocacy
Subject: Re: Linux, XML, and assalting Windows
>
> Well you can do whatever you like with the data; you might want to encrypt it
> because it contains your customer's credit card details, but the format of the
> data file is still standard and that is what matters.
[Maybe I shouldn't have posted so quickly] You probably wouldn't encrypt
non-sensitive data, but everyone tries to hide their sensitive data one way or
another. I don't think the issue is whether or not someone else's obscure or
encypted data is readable by all.
IanP
------------------------------
From: "Stephan J" <[EMAIL PROTECTED]>
Subject: Get rid of LILO in the MBR
Date: Sat, 26 Aug 2000 15:57:18 +1000
Hi all,
I have installed LILO and now wish to install Windows 98 over it, I don't
know why? How can I get rid of LILO? I have created new partitions but
still get the LI when my new operating system boots.
--
Stephan J
[EMAIL PROTECTED]
------------------------------
From: <[EMAIL PROTECTED]>
Crossposted-To: alt.os.linux,comp.text.xml,comp.os.linux.misc,comp.os.linux.advocacy
Subject: Re: Linux, XML, and assalting Windows
Date: Fri, 25 Aug 2000 23:38:30 -0700
Reply-To: <[EMAIL PROTECTED]>
Ian Pulsford <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> >
> > Well you can do whatever you like with the data; you might want to
encrypt it
> > because it contains your customer's credit card details, but the format
of the
> > data file is still standard and that is what matters.
>
> [Maybe I shouldn't have posted so quickly] You probably wouldn't encrypt
> non-sensitive data, but everyone tries to hide their sensitive data one
way or
> another. I don't think the issue is whether or not someone else's obscure
or
> encypted data is readable by all.
I am not debating on the end user need for privacy and encryption.
I am refering the the files data being encrypted for the benefit of the
software's developer alone inorder to prevent that data being handled by any
program that were not written by that devloper to lock the user into their
products.
In such a case what good does it do if the is in a "standard format" or that
it can be parsed so long as not a single datum can be interpreted?
Let's reexamine the long suffering example again. You are using an office
productivity application. You are working on a document that consumes 5 Meg
stored in a non-XML file. You are given the option of saving it in either
XML or the program's native format. You choose the XML file format. When
you examine the file you have just saved you find the XML the following XML
"tokens" the the start of the file.
<?xml version="1.0" encoding="iso-8859-1" ?>
<!DOCTYPE RST "http://localhost/fubar.dtd>
<RST>
<R ID="0" >
<F0>
Followed by 6.5Meg of data of the follow kind
alahasdfnaxvc9qweafva8712345lkf0asdf
Followed by the closing tags:
</F0>
</R>
</RST>
What have you gained?
------------------------------
Date: Sat, 26 Aug 2000 17:02:56 +1000
From: Ian Pulsford <[EMAIL PROTECTED]>
Crossposted-To: alt.os.linux,comp.text.xml,comp.os.linux.misc,comp.os.linux.advocacy
Subject: Re: Linux, XML, and assalting Windows
[EMAIL PROTECTED] wrote:
> Ian Pulsford <[EMAIL PROTECTED]> wrote in message
> news:[EMAIL PROTECTED]...
> > >
> > > Well you can do whatever you like with the data; you might want to
> encrypt it
> > > because it contains your customer's credit card details, but the format
> of the
> > > data file is still standard and that is what matters.
> >
> > [Maybe I shouldn't have posted so quickly] You probably wouldn't encrypt
> > non-sensitive data, but everyone tries to hide their sensitive data one
> way or
> > another. I don't think the issue is whether or not someone else's obscure
> or
> > encypted data is readable by all.
>
> I am not debating on the end user need for privacy and encryption.
>
> I am refering the the files data being encrypted for the benefit of the
> software's developer alone inorder to prevent that data being handled by any
> program that were not written by that devloper to lock the user into their
> products.
>
> In such a case what good does it do if the is in a "standard format" or that
> it can be parsed so long as not a single datum can be interpreted?
>
> Let's reexamine the long suffering example again. You are using an office
> productivity application. You are working on a document that consumes 5 Meg
> stored in a non-XML file. You are given the option of saving it in either
> XML or the program's native format. You choose the XML file format. When
> you examine the file you have just saved you find the XML the following XML
> "tokens" the the start of the file.
>
> <?xml version="1.0" encoding="iso-8859-1" ?>
> <!DOCTYPE RST "http://localhost/fubar.dtd>
> <RST>
> <R ID="0" >
> <F0>
>
> Followed by 6.5Meg of data of the follow kind
>
> alahasdfnaxvc9qweafva8712345lkf0asdf
>
> Followed by the closing tags:
>
> </F0>
> </R>
> </RST>
>
> What have you gained?
Who's me? The casual web browser, the company vice president, the
programmer?
Answer: nothing unless you know what the hell it is.
If two companies wanted to share the data then they would agree on a
method of encryption. If someone wanted to share it with the whole world
then they would make it nice simple english (swahili whatever).
IanP
------------------------------
From: John Hovell <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.misc,comp.os.linux.networking
Subject: Re: Monitor ip masq?
Date: Sat, 26 Aug 2000 07:10:38 GMT
Devon --
Well, this can be done with IPChains... the "-l" option will log all
packets. So, if you are interested in logging packets to from your network
to port 80, I would try:
/sbin/ipchains -N weblog
/sbin/ipchains -I forward 1 -j weblog
/sbin/ipchains -A weblog -s $internal_net --dport 80 -j RETURN -l
/sbin/ipchains -A weblog -s $internal_net --dport 443 -j RETURN -l # for
https
This will output to kernel logs. Unfortunately it doens't resolve names or
make stuff very pretty, but many frontends for ipchains logs can be found
on the 'Net and many of them even email the output to you in pretty colored
HTML tables.
I use logcheck but it blows. Look for something better at freshmeat.net or
your favorite Linux software site.
Of course, getting a proxy might serve this better, as this will (as your
question requested) give web *sites* but not URLs or content details. This
would be easier to do using a proxy, which is made for many reasons
including this. Or, maybe tcpdump could be used in conjunction with perl
-- a do-it-yourself job, to give you nice URLs and maybe even link to
IPchains to do content filtering.
There's probably software out there that already does this though.
Cheers,
John
Devon Harding wrote:
> How can I monitor IP Masq. traffic, such as which web site each user is
> going to?
>
> -Devon
------------------------------
From: "GreatFree" <[EMAIL PROTECTED]>
Subject: Stupid Question - 1
Date: Sat, 26 Aug 2000 00:32:55 -0700
Dear All,
I have a large hard disk :-). so I plan to install both Windows & Linux. Now
Windows is installed. But I don't know how to install Linux on my hard disk.
I have a F: disk that is more than 9G. Can I install Linux on it? How can I
do it?
Thanks,
Bing
------------------------------
** 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.setup) 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-Setup Digest
******************************
