Linux-Advocacy Digest #476, Volume #31 Mon, 15 Jan 01 03:13:07 EST
Contents:
Re: The real truth about NT's aggressive caching (The Ghost In The Machine)
Re: OS-X GUI on Linux? ("Erik Funkenbusch")
Re: Linux Mandrake 7.2 and the banana peel (David Utidjian)
Re: Windows 2000 ("Erik Funkenbusch")
Re: Windows 2000 ("Erik Funkenbusch")
Re: you dumb. and lazy. (The Ghost In The Machine)
Re: Windows 2000 ("Erik Funkenbusch")
Re: Does Linux envy Microsoft? (R.E.Ballard ( Rex Ballard ))
Re: Global Configuration tool (WAS: Re: linux does NOT suck (oh yes it does) )
("Ayende Rahien")
Re: Delphi Forums Downgrading from Windows 2000 to NT 4.0 ("Ayende Rahien")
Re: The Server Saga (Ketil Z Malde)
Re: The Server Saga (Ketil Z Malde)
----------------------------------------------------------------------------
From: [EMAIL PROTECTED] (The Ghost In The Machine)
Subject: Re: The real truth about NT's aggressive caching
Date: Mon, 15 Jan 2001 07:06:49 GMT
In comp.os.linux.advocacy, Aaron R. Kulkis
<[EMAIL PROTECTED]>
wrote
on Sun, 14 Jan 2001 19:04:26 -0500
<[EMAIL PROTECTED]>:
>The Ghost In The Machine wrote:
>>
>> In comp.os.linux.advocacy, Aaron R. Kulkis
>> <[EMAIL PROTECTED]>
>> wrote
>> on Thu, 11 Jan 2001 10:39:39 -0500
>> <[EMAIL PROTECTED]>:
>> >pip wrote:
>> >>
>> >> [EMAIL PROTECTED] wrote:
>> >>
>> >> > > I get 100% success with Windows 98 SE. What's your point?
>> >> >
>> >> > How much do you burn? My company burn's about 1000 CD's per month (all
>> >> > custom stuff for clients).
>> >>
>> >> Then get better CDR software :-)
>> >> What has this got to do with the OS?
>> >
>> >The CD writing process is susceptible to bottle-neck-induced errors.
>> >
>> >NT's "agressive caching" technique (i.e. caching when there's absolutely
>> >no fucking reason to cache), causes disk I/O bottlenecks...thereby
>> >increasing the likelihood of your CDR-burn to have errors.
>>
>> I'd like some details on this; in particular, does it preread sectors
>> of files open for read, thereby putting extra pages into the cache?
>
>whoops, I meant "aggressive swapping"...
>
>It was a much bally-hooed feature...that NT would start swapping BEFORE
>it had to.
Interesting. This means that it's effectively reserving pages
for the file cache (presumably, NT caches pages much the same way
Linux does, using an LRU algorithm -- although I wonder about that, too).
But that would create extraneous I/O, as well. IMO, stupid;
might as well do LRU on all of executable/DLL pages, process data pages,
and file pages. It's not perfect, but computer systems aren't
clairvoyant, either. :-)
If your phrasing is correct, it sounds like marketdroid speak:
"Buy Windows NT, now with new extra special Aggressive Paging
Capability!".
Yuck.
[.sigsnip]
--
[EMAIL PROTECTED] -- insert random marketdroid speak here
EAC code #191 3d:23h:49m actually running Linux.
Are you still here?
------------------------------
From: "Erik Funkenbusch" <[EMAIL PROTECTED]>
Subject: Re: OS-X GUI on Linux?
Date: Mon, 15 Jan 2001 01:14:37 -0600
"." <[EMAIL PROTECTED]> wrote in message
news:93tldr$3p5$[EMAIL PROTECTED]...
> Erik Funkenbusch <[EMAIL PROTECTED]> wrote:
> > "." <[EMAIL PROTECTED]> wrote in message
> > news:93t7rd$en0$[EMAIL PROTECTED]...
> >> Erik Funkenbusch <[EMAIL PROTECTED]> wrote:
> >> > "." <[EMAIL PROTECTED]> wrote in message
> >> > news:93t22l$gob$[EMAIL PROTECTED]...
> >> >> Erik Funkenbusch <[EMAIL PROTECTED]> wrote:
> >> >> > "Charlie Ebert" <[EMAIL PROTECTED]> wrote in message
> >> >> > news:[EMAIL PROTECTED]...
> >> >> >> Absolutely,
> >> >> >>
> >> >> >> OS-X on Linux.
> >> >> >>
> >> >> >> I'll try that.
> >> >>
> >> >> > What kind of a moron are you? OS-X is BSD. How could you run BSD
on
> >> > Linux?
> >> >>
> >> >> OSX is *not* BSD. You are quite highly misinformed.
> >>
> >> > It's BSD based.
> >>
> >> Loosely. WindowsME is dos based in just about the same way.
>
> > No matter how you look at it, OS-X can't run on Linux.
>
> You really are an idiot.
>
> 1. why would anyone want it to.
Ask Charlie Ebert, he made the comment. Scroll up, it's still there.
> 2. sure it can, it just doesnt because no one has written a translation
layer
> yet, and probably never will.
Well, there's VMWare, but that's not what we're talking about here. You'd
need a PPC emulator for Linux.
Still, that's not what the point was. Charlie seemed to think OSX was a
window manager or something.
> 3. this is hardly a linux shortcoming, it wont run under W2K either, and
> CANNOT do so.
I didn't say it was. This was all about Charlie Ebert making an inane and
stupid statement.
------------------------------
From: David Utidjian <[EMAIL PROTECTED]>
Subject: Re: Linux Mandrake 7.2 and the banana peel
Date: Mon, 15 Jan 2001 02:09:58 -0500
Pete Goodwin wrote:
> > Does MDK-7.2 have support for ReiserFS built in? I
> > don't think it does.
>
> Actually it does.
Thanks.... didn't know that. I think it must have been an option I
missed on a friends box I installed it on.
> > Your networking problems are another issue.... what does the the command
> > "/sbin/route -n" give you?
>
> Too late, it's a Windows Millenium machine now.
Gosh thats a shame... friends shouldn't let friends use Windows ;-)
> What I don't understand is why choosing slightly different settings during
> installation gets such wide results.
Well part of the reason could be that whatever system they use for
deciding which packages are installed how is not very well thought out.
I have had pretty unsatisfactory experiences with Mandrake and I have
tried installing it on several systems several times. It is nice for the
newbie but I don't really care for it. It has allways given me problems
with hardware of one kind or another and basic setup. I much prefer
RedHat, Debian and SuSE. Many people are perfectly happy with Mandrake
and it works fine for them. Nice thing about Linux... if a distro
doesn't work the way we like it... we can use another... or
roll-yer-own. Kinda the same thing with Windows I suppose... but much
more expensive to experiment with different versions.
-DU-...etc...
------------------------------
From: "Erik Funkenbusch" <[EMAIL PROTECTED]>
Subject: Re: Windows 2000
Date: Mon, 15 Jan 2001 01:18:23 -0600
"Matthias Warkus" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> > Word has traditionally stored binary data structures in it's file
format.
> > This means that, unless you always convert endianness when loading and
> > unloading documents, the file formats (even if otherwise identical) will
not
> > be the same for data content. More likely, Word only does endianness
> > conversion when using filters for a non-native file format.
>
> Is there a reason not to simply have all data on all platforms stored
> in network byte order and be done with it?
Not a good reason, just legacy.
------------------------------
From: "Erik Funkenbusch" <[EMAIL PROTECTED]>
Subject: Re: Windows 2000
Date: Mon, 15 Jan 2001 01:20:22 -0600
"Gary Hallock" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> > And your point is what? The early years of the PC industry were
entirely
> > Intel based. I'm not talking about Apple II's or Commoodre 64's here,
but
> > rather business machines bought by businesses. Back then, the mainframe
> > people (even inside IBM) thought PC's were a fad that would pass and put
> > nearly no effort into connectivity between systems. The only way to
> > communicate at the time for most people was BBS's, and even then people
> > didn't exchange documents.
>
> Well, I thought my point was quite obvious. The computer
> industry has a long history of different data formats. There
> have been some standards for data developed over time such as
> the 8-bit byte and IEEE floating point, but there are still many
> differenenes. Not taking this into account is very
> short-sighted.
The PC industry as a whole was short sighted. The IBM PC's original design
was highly short sighted, especially when you consider that 68000's were
available at the time (and had been for several years).
IBM controlled the PC, and they treated it like a throw away project with
little to no planning. The PC industry followed IBM's lead.
------------------------------
From: [EMAIL PROTECTED] (The Ghost In The Machine)
Crossposted-To: alt.linux.sux
Subject: Re: you dumb. and lazy.
Date: Mon, 15 Jan 2001 07:17:03 GMT
In comp.os.linux.advocacy, Aaron R. Kulkis
<[EMAIL PROTECTED]>
wrote
on Sat, 13 Jan 2001 15:39:18 -0500
<[EMAIL PROTECTED]>:
>ono wrote:
>>
>> > You moron....
>> > Linux doesn't have any "DLL Hell", ignorant twit.
>> > Linux libraries are properly versioned, and successive versions
>> > of the same library can be kept on the same system without conflict.
>> And thats why linux suffers from bloat! Because every program comes
>> with it's own statically linked version of the same shit. MS at least
>> tried to do something.
>
>Dumbshit. Statically linked libraries were abandoned years ago.
Pedant point: statically linked libraries are useful for some
applications, such as minimal functionality in case libc got trashed,
for example. Admittedly, this use is very limited.
Most developers -- myself included -- would use shared libraries.
And on Linux, they might even be true shared libraries, as opposed
to merely dynamically loadable. I'd have to dig, though, to be sure.
Amigas, for example, had true shared libraries; mind you, they
didn't have memory protection. Apollo DOMAIN systems running
Aegis had them, too -- for those who remember those workstations
(Apollo got swallowed up by HP, and some of the ideas might have
been incorporated to their PA RISC software or something).
Now I've got myself curious. Hmm....
[rest snipped]
--
[EMAIL PROTECTED] -- insert random misquote here
EAC code #191 3d:23h:58m actually running Linux.
This space for rent.
------------------------------
From: "Erik Funkenbusch" <[EMAIL PROTECTED]>
Subject: Re: Windows 2000
Date: Mon, 15 Jan 2001 01:24:24 -0600
"Tom Wilson" <[EMAIL PROTECTED]> wrote in message
news:4Hw86.153$[EMAIL PROTECTED]...
> "Erik Funkenbusch" <[EMAIL PROTECTED]> wrote in message
> news:6Uo86.96$[EMAIL PROTECTED]...
> > <[EMAIL PROTECTED]> wrote in message
> > news:[EMAIL PROTECTED]...
> > > >Yes, there was a long history of such in the
> scientific and perhaps even
> > > >banking industry, but not the *PC* industry.
> > >
> > > So? Are you trying to tell us that BillyBob was so
> incompetent
> > > and disinterested in his 'beefier' potential rivals
> that he
> > > was completely unaware of any of that?
> >
> > It's not like Bill Gates was personally writing the
> software.
>
> I was under the impression that, early on, he was very
> "hands-on" with the development end of things.
Yes, he was. But in an interview he gave a while back when he stepped down
as CEO, he stated that the last software he personally worked on was writing
the BIOS for the Tandy 100, which if I recall was around 1983.
------------------------------
From: R.E.Ballard ( Rex Ballard ) <[EMAIL PROTECTED]>
Subject: Re: Does Linux envy Microsoft?
Date: Mon, 15 Jan 2001 07:15:35 GMT
In article <92tmm8$ojo$[EMAIL PROTECTED]>,
hackerbabe <[EMAIL PROTECTED]> wrote:
> A quote from http://microsoft.aynrand.org/hate.html,
> referring to why
> Microsoft has been persecuted in the anti-trust trial:
>
> "There is only one fundamental reason why
> great businessmen [like Bill
> Gates] or great companies [like Microsoft] are hated, and it has
> nothing to do with so-called monopolies. [Microsoft is] hated . . .
> because [it is] good, that is, smarter, more visionary, more
> creative,
> more tenacious, more action-focused, more ambitious, and more
> successful than everyone else.
This is partly true. What makes Microsoft Unique in the world of business is
that it is run by 4 majority stockholders who can't be threatened by a proxy
war. As a result, Bill Gates was able to conceive, announce, and execute a
10 year plan for "World Domination" with relatively little interference.
In a business world where most managers are focused on the
next quarter, Microsoft was able to offer trivial discounts
in exchange for key strategic concessions. This, more than
anything is a tribute to Bill Gate's brilliance.
Very few men in all of history are able to do this. In
reverse order, you have Josef Stalin, Adolf Hitler, Napoleon,
Genghis Kahn, Julius Ceasar, Alexandar the Great, and Rameses
the Egyptian.
In many ways, Gates is equally ruthless, killing off competitors economically
in much the way Hitler exterminated Jewish towns. The Gates' approach is
cleaner, with fewer bodies to bury and burn. Killing businesses is much
"cleaner" than killing the owners of the business.
Microsoft frequently stripped company founders of their dreams, their
financial security, their intellectual property, and their dignity, by
putting the pressure to the VCs, customers, and suppliers. In many cases,
Microsoft was able to get strategic intellectual properties for pennies on
the dollar.
In any other time in history, and any other part of the world, Bill Gates
would have been killed by mobsters hired by competitors. But in this time
and place, the only defense is lawyers, who often turn against you when
offered a quick and easy settlement.
Microsoft has often lost in court, only to pull it out with a quick
settlement, a court ordered sealing of the records and transcripts, and a
trivial amount of cash.
As much as I enjoyed Antitrust, I know of no actual homicides ordered or
tracable directly to any Microsoft executive. There have been a number of
mysterious deaths and accidents involving Microsoft competitors.
Ironically, when Bill Gates announced his plan for world domination in
November of 1984, only a handful of people took him seriously. One of those
was Richard Stallman. It was Stallman and some of the other nameless members
of the Open Source movement who created the General Public License, and in
doing so, created the only effective means of limiting and undermining
Microsoft's plans. Because of Open Source, Microsoft is nearly 5 years
behind schedule. Microsoft was unable to create a Monopoly of digital
communications. They were unable to create a Monopoly of financial
transactions. They were unable to create a Monopoly of intellectual
property, mass media, and financial markets.
Ironically, Ballmer's accusations of "Linux as Communism" is ironic coming
from a company that thinks in binary, one or zero, Microsoft IS the
"Standard", totalitarian approach to business.
> Haters of the good [competing OSes
> and browsers] do not want the less
> able to be raised up to the level
> of the great producers (which is
> impossible); they want the great
> producers to be brought down.
Actually, all they really want is access to the market
on similar terms. Linux isn't demanding exclusive
disttribution on every PC, Microsoft is. Netscape
wasn't demanding that IE be excluded form PCs, Microsoft
demanded that Netscape be excluded. Linux actually supported
the ability to concurrently run Linux and Windows, but Microsoft
demanded that NO modifications be made to the boot sequence
(especially boot manager that could boot Linux).
The Open Source community has generated a very competitive market
in which hundreds of players thrive and grow. Open Source is the
absolute CORE of nearly ALL of the economic growth that has happened
in the last 20 years. Microsoft has stolen Open Source property and
exploited it with the intent of expanding their monopoly, which allowed
it to benefit from Open Source generated economic growth.
> They want to use government coercion
> to cripple the greatest minds so that
> lesser minds will not feel inferior."
If Michael Jordan walked out and started shooting
the entire New York knicks Basketball team before
shooting a series baskets that allowed the Bulls
to win by 200 to 0, would you respect the victory.
When businesses refuse to obey the law, and government
refuses to enforce those laws, the industry becomes lawless.
The criminalization of liquor, combined with the selective
enforcement of prohibition created the environment for
organized crime. When extortionists moved into the clubs,
the legal disputes and competition issues were settled with
Thompson submachine guns.
The same thing happened with the recreational drug
industry. Dealers began shooting other dealers,
customers killed dealers, dealers killed customers,
and dealers formed alliances with other dealers.
The problem wasn't that drugs were illegal, but rather
that because drugs were illegal, that none of the normal
laws applied. There was no FDA monitoring quality control,
no FTC preventing fraud, no tax revenue to fund regulation,
and no financial management since the drug money had to be
laundered.
Today, drug cartels own thousands of small businesses used for
money laundering (theaters, restaurants, video arcades, and so-on).
But the illegal part of the business still functions unchecked
because the best the DEA can do is get to 2nd or 3rd level dealers.
If the DOJ appears to place Microsoft "above the law", there may be
reprecussions. We have already seen at least 5 viruses that threatened
millions of machines. We've seen major exploits of Microosft's
vulnerabilities in Melissa, I love you, and resume.
If Microsoft is perceived to be above the law, there may be those
who take the law into their own hands.
Jessie James was considered to be a hero by most of the homesteaders and
farmers in Kansas and Missouri because the railroads and banks exploited the
homesteaders and appeared to be above the law. James would take a modest
amount of money, but he would also burn the mortgages.
The Dillinger gang and the Barrow gang were often protected by farmers and
local who had already been forced to give up their farms and become
sharecroppers because of the Great Depression.
We now live in a world where cash is little more than "government green
stamps", and the "Real Money" is the bits in the bank's corporate computers.
We live in a world where SWIFT moves more cash in a few minutes than the
government prints in an entire year.
> --------
> Is there any truth to this accusation of envy, or are there other
> reasons people dislike Microsoft?
>
> Sent via Deja.com
> http://www.deja.com/
>
--
Rex Ballard - Sr I/T Systems Architect
Linux Advocate, Internet Pioneer
http://www.open4success.com
Linux - 80 million satisfied users worldwide
and growing at over 9%/month! (recalibrated 01/14/00)
Sent via Deja.com
http://www.deja.com/
------------------------------
From: "Ayende Rahien" <[EMAIL PROTECTED]>
Crossposted-To: alt.linux.sux,alt.microsoft.sucks,alt.linux.sucks,alt.linux.slakware
Subject: Re: Global Configuration tool (WAS: Re: linux does NOT suck (oh yes it
does) )
Date: Mon, 15 Jan 2001 09:14:21 +0200
"Steve Mading" <[EMAIL PROTECTED]> wrote in message
news:93u4k5$9ce$[EMAIL PROTECTED]...
> In comp.os.linux.advocacy Jure Sah <[EMAIL PROTECTED]> wrote:
> : Steve Mading wrote:
> :>
> :> In comp.os.linux.advocacy The Ghost In The Machine
<[EMAIL PROTECTED]> wrote:
> :> : In comp.os.linux.advocacy, Steve Mading
> :> : <[EMAIL PROTECTED]>
> :> : wrote
> :> : on 9 Jan 2001 20:58:53 GMT
> :> : <93fu2d$ib4$[EMAIL PROTECTED]>:
> :> :>In comp.os.linux.advocacy Jure Sah <[EMAIL PROTECTED]> wrote:
> :> :>: Yup. At least until it gets a Linuxal Basic that is better than
Visual
> :> :>: Basic.
> :> :>That's already happened.
> :> : Just out of curiosity -- where?
> :> I was being somewhat sarcastic. ANY basic is "better than Visual
Basic".
>
> : You're an idiot. Mind showing me the complete list of all Basic computer
> : languages then?
>
> Hint: I wasn't being completely serious. I even *labelled* myself as
> being sarcastic, for crying out loud.
>
> If you want pendantic seriousness, then fine, here it is:
> There are many basics that I consider "better" than Visual Basic.
> There do exist some basics that are worse, but they aren't modern.
> Part of the reason for this is that I don't like embedded GUI
> toolkits and auto code generators, becuase they are a pain in the
> ass to maintain when you want to change what the tool generated,
> and yet still keep using the tool to edit it too. So, all the
> GUI bells and whistles are irrelevant to me. Once that's taken
> away, there isn't much left in Visual Basic that makes it good.
> In other words, the good stuff isn't technically part of the
> language at all. It's part of the programming environment. IMO,
> the programming tool should be independant of evaluating the
> language itself. They should be seperated very clearly into two
> layers such than the GUI programming tool is a pluggably replacable
> by a third party. This ensures that language and compiler design
> issues don't leak into the interface tool, and interface design issues
> don't leak into the compiler itself.
You need to check VBScript, then.
VB is the language and the GUI, VBScript is the language alone.
Well, there are some modifications, but it's almost the same.
------------------------------
From: "Ayende Rahien" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.ms-windows.nt.advocacy
Subject: Re: Delphi Forums Downgrading from Windows 2000 to NT 4.0
Date: Mon, 15 Jan 2001 09:32:07 +0200
"Adam Warner" <[EMAIL PROTECTED]> wrote in message
news:93ts5h$jdc$[EMAIL PROTECTED]...
> > #2) You also miss the points that NT4 is so stable for them that given
the
> > chance to change not once buy twice (once up to W2K and once again down
to
> > NT4) they continue to use MS NT. Obviously (and from experience) NT4 is
> very
> > stable and works great, great enough for Delphi.
>
> There would be a VERY big legacy issue in the
> OS/database/scripting/server/etc.
I don't think so, depending on what they have, of course.
But you can transfer ASP to PHP automatically (there is a tool that does it,
never used it, can't tell how good it is).
They would've to invest in a good DB, MySQL can't hold a candle to serious
players like SQL Server, Oracle, DB2 and so on.
Perhaps there is a free DB that can satisfy their needs, but you'll have to
be in their shoes in order to decide about it.
About the sever & OS, I would hold judgement until I can get their hardware
specs.
> > The NIC drivers caused
> > problems - ok, but to blame MS's product for another vendor's problems?
> How
> > about if I write some crappy drivers for Linux and when people use them
> and
> > linux crashes hard, can I say it's Linus' fault? See the point?
>
> If a linux network driver is 100% stable under 2.0 (or 2.2) and it's not
> under 2.2 (or 2.4) then you can probably blame the Linux kernel. However I
> realise the buck doesn't stop with Microsoft because they/you blame a
third
> party.
>
> This is a legacy issue. This was expensive vendor hardware which runs
stable
> under NT. It would probably have been on the HCL. Isn't it also
Microsoft's
> responsibility to see that critical legacy hardware remains compatible?
It's not.
NT & 2K are using totally different drivers schemes.
2K uses WDM, Win32(r) Driver Manager, NT use something else, it probably has
an acronym(sp?), but I can't recall it right now.
If the NIC has problems working with 2K, it's a driver issue.
In theory, you *could* use the NT drivers on 2K, but it would've to be an
idiot that would try that on a production machine.
I think that we can safely assume that there are 2K drivers for it, and that
they aren't very good.
NT drivers work on NT only, WDM works on 98/Me/2K.
Since the problem exist with 2k & the NIC cards, I would say that it's a
driver problem and go to the producer (anyone know who they are?) and ask
them for a better driver.
In server class equipment, that request can hardly be turned down. (Can you
imagine the damage that would be if they would announce that they are
degrading because XYZ didn't offer good drivers for 2K, after being
requested?)
> > #3) If they were smart they'd have changed NIC's instead of changing
OSes
> > twice. I mean, really... NICs are cheap in comparison to both the OS
cost
> > and the cost to convert (in time/energy) and in the cost to the end
users
> > opinion of Delphi's service. I mean, they'll think Delphi sucks, not
W2K.
>
> Couldn't they have been talking about 1Gb network cards? (probably not
> cheap).
There is this, and the fact that it's not that expensive to downgrade.
IIRC, 2K's EULA allows you to run previous versions of it. IE, if you buy 2K
license, you can run NT4 under it.
I know several companies that buy 2K licenses but run NT instead while they
test their needs on 2K.
> > Smart people know where to lay the blame and in this case Delphi is
taking
> > the easy way out and blaming the easy target - why don't they name the
NIC
> > vender?? HMmmm??! Why didn't they simply change NICs?
>
> Yes I would like to know who the vendor is.
Dito. And I would like to know why they don't release good drivers for 2K.
Any server-class manufacteror has the responsibility (if they want to keep
their reputation) to either release good drivers or release nothing at all.
I can accept beta drivers while they produce better drivers, but that is
probably not what happened.
------------------------------
Subject: Re: The Server Saga
From: Ketil Z Malde <[EMAIL PROTECTED]>
Date: Mon, 15 Jan 2001 07:42:07 GMT
Pete Goodwin <[EMAIL PROTECTED]> writes:
> Now, I can see the reactions of people already:
[...]
> What's my answer to any of these points?
I don't think so. The only thing you did to provoke anything like it,
would be to post in c.o.l.advocacy, instead of a more appropriate
group - if it's help you want, that is.
> Linux can boot and read this setup, so I tried to use it,
> through several installations:
What does that mean - you actually reinstalled everything?
> First Installation
> TCP/IP was broken. Everything was correct but I could not ping my other
> machine, and likewise from the other machine. I could write a CDR OK.
Did you check kernel messages to see if the driver was loaded? Did
you check that the network scripts were run, or try to run them
manually?
> Second Installation
> TCP/IP was working, however, Linuxconf could not configure my system for
> me, so I was unable to get TELNETD, NFSD or SMBD working. I could run SMB
> directly with /etc/rc.d/init.d/smb start but not at boot up.
Sounds like an incorrect runlevel. Check /etc/inittab to see that the
correct default runlevel is specified. Check also that the
appropriate symlinks from /etc/rc.d/rcN.d/ to /etc/rc.d/init.d (if
Mandrake uses an RH layout, that is).
> Third Installation
> I deliberately installed KDE/GNOME so I could run linuxconf on X. TCP/IP
> was working but a previously working script for Samba is now broken.
What's the error message?
> Fourth Installation
> The only snag with this setup is the lack of remote access to the PC. I
> have to access the desktop to do anything. Remote file access is possible,
> but I can reboot or run the CD burner remotely. I can live with that for
> now.
All kinds of remote access (telnet, rsh, rlogin, ssh)? What happens?
> Linux Mandrake is a very nice package for someone who doesn't want to get
> into the bowels of UNIX style configuration... but falls apart if you want
> to do something other than install everything.
I think you - if you wish to administrate systems - need to know how
things work under the hood. This is very true for Linux, and quite
true for the varieties of Windows as well.
It is theoretically possible for software and hardware to configure
itself, but I've yet to see any practical solution that actually
works.
-kzm
--
If I haven't seen further, it is by standing in the footprints of giants
------------------------------
Subject: Re: The Server Saga
From: Ketil Z Malde <[EMAIL PROTECTED]>
Date: Mon, 15 Jan 2001 07:46:12 GMT
Pete Goodwin <[EMAIL PROTECTED]> writes:
> On my workstation my host.allow and hosts.deny are both empty. Telnetting
> into it works just fine.
> On the server machine, telnetting to it resulted in "Permission
> denied".
"Permission denied"? What username? Did you check file permissions,
in particular on root, home and similar directories? Can you "su" to
the same user?
-kzm
--
If I haven't seen further, it is by standing in the footprints of giants
------------------------------
** 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 by posting to comp.os.linux.advocacy.
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-Advocacy Digest
******************************
