Linux-Advocacy Digest #569, Volume #34 Thu, 17 May 01 05:13:02 EDT
Contents:
Re: LILO no boot .. says "LIL-" then just hangs there ("Bernadette and Warren
McRobie")
Re: Linux beats Win2K (again) (Jeff Silverman)
Re: Oracle 8.1.6 on Solaris or Linux? ("Mart van de Wege")
Re: Win 9x is horrid ("Erik Funkenbusch")
Re: Linux posts #1 TPC-H result (W2K still better) ("Edward Rosten")
Re: Microsoft - WE DELETE YOU! ("Mart van de Wege")
Re: Why did Eazel shutdown? ("Erik Funkenbusch")
Re: The Economist and Open-Source ("Donal K. Fellows")
Re: Why did Eazel shutdown? ("Erik Funkenbusch")
Mandrake 8 sets the standard - for Desktop users anyway. ("fk10")
Re: The Economist and Open-Source ("Erik Funkenbusch")
----------------------------------------------------------------------------
From: "Bernadette and Warren McRobie" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.setup,dc.org.linux-users
Subject: Re: LILO no boot .. says "LIL-" then just hangs there
Date: Thu, 17 May 2001 20:13:38 +1200
I found that having the drive linux is installed on a slave with no master,
then this error occurs - yes I know it should be the master - but I did not
set up this machine hardware wise ;P
HTH
Warre
"rider web" <[EMAIL PROTECTED]> wrote in message
news:9ds2af$d5v$[EMAIL PROTECTED]...
> this might help
> http://www.redhat.com/support/docs/faqs/rhl_general_faq/FAQ-5.html
>
>
> "Counts" <[EMAIL PROTECTED]> wrote in message
> news:9dc3ut$[EMAIL PROTECTED]...
> > I installed everything successfully (Slackware 4.1), recompiled a
kernel,
> > rebooted, everything worked as expected. Then, not long afterwards on
> > another reboot, LILO just hangs. I have to boot using a floppy now.
Any
> > suggestions on how to fix this?
> >
> > Should I uninstall LILO and then re-install?
> >
> > Thanks for any suggestions.
> >
> > Bob
> >
> >
>
>
------------------------------
From: Jeff Silverman <[EMAIL PROTECTED]>
Subject: Re: Linux beats Win2K (again)
Date: 17 May 2001 08:15:37 GMT
Greg Copeland wrote:
> "Edward Rosten" <[EMAIL PROTECTED]> writes:
>
> > I'm bored of all these benchmarks, since benchmarks are just marks and
> > meaningless out of context. So Win2K served up an extra transcation per
> > second or Linux manages an extra web page per second? So what?
>
> I agree. At this point, it's pretty clear that Linux can keep up with
> Win's performance without problem. Likewise, Win is on the same page
> as Linux. Uptime is a different issue. A sustained, high average load
> is again, another issue.
>
I agree. Security is another issue.
> > To get a better idea, you need to look at the real world.
> >
> > If you look in the real world, you see Linux having several spots in the
> > top 100 fastest supercomputers. If Win2K/NT is so great and so scalable
> > and gives such a great price/performance ratio, then why is there not a
> > *single* Windows cluster in the top 100 supercomputers list?
>
> Though I do believe that this somewhat makes your point, it's not entirely
> fair. Win doesn't really have a high-end distributed computing solution
> like Linux does (if one exists, please point it out to me). Likewise,
> it doesn't really support clusters like MOSIX does. So, saying that Win
> doesn't show up on the list, doesn't mean that it can't because of simple
> performance and scalability issues, but rather, Win doesn't really have
> the technologies to compete (no distributed computing and no clusters
> [hot fail-over/redundancy makes a true cluster not]). That makes, I think,
> another good point, but it is clearly a different point unto its own.
Although I'm not sure how relevant it is for the mass market, scalability works both
upwards and
downwards. Which actually helps make your point. Consider, for example,
http://wearables.stanford.edu/hardware.html#pc . It is a a single-board AMD 486-SX
computer with a
66 MHz CPU, 16 MB RAM, and 16 MB flash ROM, big enough to hold a useful amount of
RedHat 5.2 Linux
including the HTTP daemon that runs the web server. It measures 2.8" x 1.8" for an
area of 5 square
inches. Being 0.2" thick this makes the volume one cubic inch (16 cc's). It weighs 3/4
oz (20
grams).
Now Microsoft is trying to promote Windows/CE for embedded systems. It's minimum
footprint, just
for the OS, is a 10 megabytes and it requires a pentium. I would argue that Windows
is too big for
an embedded environment, where smallness is a virtue.
However, we're all missing the point here. The point of Windows/2000 is to be a
reliable desktop
OS. It is what Windows really should have been. Is it useful in other environments?
I don't think
so. Is it the best solution for the desktop? I don't think so, either.
>
> --
> Greg Copeland, Principal Consultant
> Copeland Computer Consulting
> --------------------------------------------------
> PGP/GPG Key at http://www.keyserver.net
> DE5E 6F1D 0B51 6758 A5D7 7DFE D785 A386 BD11 4FCD
> --------------------------------------------------
Jeff
--
Jeff Silverman, PC guy, Linux wannabe, Java wannabe, Software engineer, husband,
father etc.
See my website: http://www.commercialventvac.com/~jeffs
[EMAIL PROTECTED]
------------------------------
From: "Mart van de Wege" <[EMAIL PROTECTED]>
Subject: Re: Oracle 8.1.6 on Solaris or Linux?
Date: Thu, 17 May 2001 10:17:45 +0200
In article
<[EMAIL PROTECTED]>,
"GreyCloud" <[EMAIL PROTECTED]> wrote:
>
> For the lone linux desktop one partition usually suffices. Why linux
> distros don't do it the way you suggest is beyond me, but it is a very
> good idea for doing O/S upgrades... unless you stay with that distro and
> like RH the upgrade is controlled by their script. So many different
> ways. :-)
<Grin>
I saw this and just had to reply. Here's my partition scheme:
/dev/hda2 on / type ext2 (rw,errors=remount-ro,errors=remount-ro)
/dev/hda5 on /opt type ext2 (rw)
/dev/hda6 on /var type ext2 (rw)
/dev/hdb1 on /usr/local type ext2 (rw)
/dev/hdb2 on /home type ext2 (rw)
/dev/hdb3 on /var/mirror type ext2 (rw)
(hda1 is my Windows partition. that's gonna go real soon now. When I buy
a new Nokia PDA this month, I'll boot one more time into Windows to
download my info from backup to my PDA, and then it is bye bye Windows)
I started with a simple scheme like Bob: the root, /opt, /home and
/usr/local on seperate partitions. In case of a reinstall or an upgrade,
I could keep my commercial applications (StarOffice and Railroad Tycoon),
my data, and the apps I compiled myself (Mozilla and the most recent
kernel among them).
Then I repartitioned /var so I could keep a local Debian mirror. If I
ever have to do a reinstall, I can install my favourite apps straight
from the HD, without downloading them again (that's what /var/mirror is
for).
I keep a seperate var, because apt puts packages there during download,
and it fills up mighty fast. I want to keep my root partition from
locking up due to running out of inodes, thank you very much (it happened
once before).
Mart
--
Gimme back my steel, gimme back my nerve
Gimme back my youth for the dead man's curve
For that icy feel when you start to swerve
John Hiatt - What Do We Do Now
------------------------------
From: "Erik Funkenbusch" <[EMAIL PROTECTED]>
Subject: Re: Win 9x is horrid
Date: Thu, 17 May 2001 03:26:26 -0500
"Mart van de Wege" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> In article <yyCM6.1126$[EMAIL PROTECTED]>, "Erik Funkenbusch"
> <[EMAIL PROTECTED]> wrote:
>
> > "Mart van de Wege" <[EMAIL PROTECTED]> wrote in message
> > news:[EMAIL PROTECTED]...
> >> In article <XlAM6.1104$[EMAIL PROTECTED]>, "Erik Funkenbusch"
> >> <[EMAIL PROTECTED]> wrote:
> >>
> >> > "Mart van de Wege" <[EMAIL PROTECTED]> wrote in message
> >> > news:[EMAIL PROTECTED]...
> >> >> In article <WrlM6.1009$[EMAIL PROTECTED]>, "Erik Funkenbusch"
> >> >> <[EMAIL PROTECTED]> wrote:
> >> >>
> >> >> > MS uses encrypted data so that activation can't be spoofed.
> >> >> > Otherwise, you could simply watch the data, and generate your own
> >> >> > "activation".
> >> >> >
> >> >> > "Mart van de Wege" <[EMAIL PROTECTED]> wrote in message
> >> >> >> Ok, from c't, issue #9 (23/4-6/5/2001), german edition:
> >> >> >>
> >> >> >> "Our attempt to use a man-in-the-middle attack to listen in on
> >> >> >> the HTTPS connection between Windows-XP and Clearing
> >> >> >> House...failed: not only does XP encrypt the data, but it
> >> >> >> receives new certificates used for further communication...
> >> >> >> Extremely questionable is why Microsoft would go to such lengths
> >> >> >> to simply exchange a few numbers, especially since the numbers
> >> >> >> are already tied to the PC hardware.
> >> >> >> ...The amount of data exchanged during activation leaves all
> >> >> >> possible options open in the light of the complexity of the
> >> >> >> process: It is possible that aside from the necessary
> >> >> >> data...other information is exchanged, it is also possible that
> >> >> >> the bloat in the data traffic is caused by the certificates
> >> >> >> alone. C't advises not to use the online activation until
> >> >> >> Microsoft makes the process more transparent. In the meantime
> >> >> >> you're better off using the telephone."
> >> >> Yes, Erik,
> >> >>
> >> >> But that wasn't the point. Read it again, *please*?
> >> >
> >> > I don't see how any other point can be derived from this.
> >> >
> >> > It's complaining because the data is encrypted, then wonders why MS
> >> > goes to such lengths to pass the data. The answer is obvious, and
> >> > the one I gave.
> >> >
> >> >
> >> >
> >> Uh no,
> >>
> >> Maybe because I have the German text beside me. They are complaining
> >> that a) MS is sending new certificates, why are they necessary? and b)
> >> the registration process is sending *too much* data to be the simple
> >> hash that MS is telling us it should be (they do note that the amount
> >> of data may be caused by the new certificates).
> >
> > Clearly they're sending certificates specific to activation, so that
> > other certificates can't be used to spoof the activation process. I
> > don't see the problem.
> >
> >
> >
> Now we are getting somewhere. The main beef c't has with the activation
> process, is that Microsoft does not tell what it does exactly. The amount
> of traffic is big enough to hide additional data in the activation
> request, and unless the exact method is known, we only have Microsofts
> word that that doesn't happen. You may trust them, c't apparently doesn't
> and they seem to be justified based on past experiences.
If you don't trust them, you shouldn't be using the OS at all, since if they
wanted to send data, they could do it in any of a million ways without you
ever being able to detect it. Why would they put such data in a high
visibility scheme like activation?
The fact of the matter is, if you trust them enough to use the OS at all,
you trust them enough to activate. There is no more or less danger or risk.
> Clearly the certificates aren't all that necessary if you can just pass
> along your hardware key over the phone if you choose that method of
> activation
The phone is much more trustworthy than the Internet, with less of a chance
for interception of the data. Otherwise, why aren't we using encrypted
telephone communications?
> Face it Erik, *nobody* trusts Microsoft on their word anymore. Well, you
> do, obviously, so I'm curious if you can give a reason for that? You seem
> to be too intelligent to trust anything on faith alone.
I don't trust them in the way you mean, but I can use common sense. Simply
using the OS gives them the opportunity to do whatever they like. If i'm
going to do that, there is no extra risk in activation.
------------------------------
From: "Edward Rosten" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.ms-windows.nt.advocacy
Subject: Re: Linux posts #1 TPC-H result (W2K still better)
Date: Thu, 17 May 2001 10:23:27 +0100
> So - if you want better performance, less heat, less maintenence (single
> vs cluster of servers), and save a bundle of money - go MS.
cite please. Evidence that R10000s (or whatever they were) make more heat
than PIII's.
-Ed
--
You can't go wrong with psycho-rats.
u 9 8 e j r (at) e c s . o x . a c . u k
------------------------------
From: "Mart van de Wege" <[EMAIL PROTECTED]>
Subject: Re: Microsoft - WE DELETE YOU!
Date: Thu, 17 May 2001 10:28:07 +0200
Crossposted-To:
alt.destroy.microsoft,comp.os.ms-windows.nt.advocacy,comp.os.ms-windows.advocacy
In article <[EMAIL PROTECTED]>, "Matthew Gardiner"
<[EMAIL PROTECTED]> wrote:
>> Why not?
>> Small servers spend most of their time idle, why shouldn't they double
>> as a workstation?
>> You save costs this way, you know.
>
> If they are so idle, why not setup an X-Server on them, and run dumb
> terminals off the server? at University, they had a Digital UNIX server
> up with X-Server that allowed people to remotely access applications on
> ther server using Win-X32.
>
> Matthew Gardiner
>
No even better: cluster them! This is what Progeny Debian is using as a
selling point: you can have your computers working as workstations, and
in periods where you need more computer horsepower, you can take their
idle cycles into a cluster. Boom! instant computing power increase!
Don't know how well this will work, but it sounds like a good idea,
especially outside office hours, when you have all these workstations
sitting idle anyway. Instead of turning them off, why not do something
useful with them?
Mart
--
Gimme back my steel, gimme back my nerve
Gimme back my youth for the dead man's curve
For that icy feel when you start to swerve
John Hiatt - What Do We Do Now
------------------------------
From: "Erik Funkenbusch" <[EMAIL PROTECTED]>
Subject: Re: Why did Eazel shutdown?
Date: Thu, 17 May 2001 03:29:52 -0500
"Matthew Gardiner" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> > > My business plan is to provide the processing and support required to
get
> > > companies connected to this service. Several EFTPOS vendors in NZ are
> > > interested, and my own bank is interested in the idea as well. All I
need
> > is
> > > backing to create the necessary infrastructure.
> >
> > That's not a business plan. That's an idea. A business plan is a
document
> > which details many things, including revenue and profit projections,
company
> > structure, company officers, detailed market analysis, etc..
>
> Even if I were to write up the business plan and submit it, the VC's in
the US
> are so ignorant, they wouldn't fund it, because its not in the US.
Well, can you blame a VC for wanting to keep a close eye on his investment?
Most VC is local, or at least has local monitoring.
But still, part of what makes a good business is the people involved. If
you don't have a good management team, you're not likely to get funded.
------------------------------
From: "Donal K. Fellows" <[EMAIL PROTECTED]>
Subject: Re: The Economist and Open-Source
Date: Thu, 17 May 2001 09:21:26 +0100
Erik Funkenbusch wrote:
> The part you are forgetting is that a component can pass all of it's tests
> and be working flawlessly, but still cause software that uses it to fail
> because that software relied on some previous undocumented functionality
> that you "fixed" or changed when fixing the bug. This is what happened with
> Lotus Notes and SP6, and the reason for the SP6a release to add the
> side-effects back in.
The part you are neglecting to mention is the fact that it was Notes that
was substantively in the wrong if it was designed to rely on a feature of
a particular implementation, and not what was in the API. Maintaining
"bug" compatability is usually wrong...
Donal.
--
Donal K. Fellows http://www.cs.man.ac.uk/~fellowsd/ [EMAIL PROTECTED]
-- Actually, come to think of it, I don't think your opponent, your audience,
or the metropolitan Tokyo area would be in much better shape.
-- Jeff Huo <[EMAIL PROTECTED]>
------------------------------
From: "Erik Funkenbusch" <[EMAIL PROTECTED]>
Subject: Re: Why did Eazel shutdown?
Date: Thu, 17 May 2001 03:31:02 -0500
"somebody" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> Anonymous <[EMAIL PROTECTED]> stands accused of saying:
> >And you only have to pay the star
> >programmer an amount of $100,000 per
> >year,
>
> lol, on what planet?
$100k is not an outrageous salary for a programmer, especially when you
factor in benefits, taxes, social security, etc.. (the company is required
by law to pay a portion of that).
------------------------------
From: "fk10" <[EMAIL PROTECTED]>
Subject: Mandrake 8 sets the standard - for Desktop users anyway.
Date: Thu, 17 May 2001 10:35:14 +0200
I was always a RedHat fan. Since 5.0 I sticked with RedHat and I always
thought the other distro's will be just the same.
Well, I finally took a deap breath and installed Mandrake 8.
W O W ! ! !
No fine tuning the X server, no playing with the sound configuration, no
need to read pages on pages of ppp howto's - everything works out of the
box.
Not even Windows is this easy!
My previous box was compromised, so this time I made sure security was top
on my priority list. Nice feature of Mandrake to give three levels of
security before you install the beast. I naturally selected PARANOID, and I
suddenly wish RedHat had this option...
Any way, everything works, and after about three days now I have no
complaints.
For those whining about headers, I'm writing this message from a VMWare
session running Win 95. I use a web service (www.news2web.com) at this
stage.
I think I will stick to Mandrake 8 for a while. The desktop environment
seems to be lightyears ahead of anything from the competition (distro's and
other OS's included), and I'm just so impressed, I can't sleep at night.
Anyway, nice chatting to you all again. See you!
Nico.
------------------------------
From: "Erik Funkenbusch" <[EMAIL PROTECTED]>
Subject: Re: The Economist and Open-Source
Date: Thu, 17 May 2001 04:02:05 -0500
"Donal K. Fellows" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> Erik Funkenbusch wrote:
> > The part you are forgetting is that a component can pass all of it's
tests
> > and be working flawlessly, but still cause software that uses it to fail
> > because that software relied on some previous undocumented functionality
> > that you "fixed" or changed when fixing the bug. This is what happened
with
> > Lotus Notes and SP6, and the reason for the SP6a release to add the
> > side-effects back in.
>
> The part you are neglecting to mention is the fact that it was Notes that
> was substantively in the wrong if it was designed to rely on a feature of
> a particular implementation, and not what was in the API. Maintaining
> "bug" compatability is usually wrong...
But MS does it. Check the "compatibility" section of the win.ini (or was it
system.ini, I forget) in Windows 9x/ME.
------------------------------
** 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
******************************
