Linux-Advocacy Digest #224, Volume #26 Sat, 22 Apr 00 22:13:04 EDT
Contents:
Re: Sell Me On Linux (Mike Marion)
Re: on installing software on linux. a worst broken system. (Sierra Tigris)
Re: which OS is best? ([EMAIL PROTECTED])
Re: Corel Linux Office 2000 dubious at best? (Stuart Krivis)
Re: Grasping perspective... (was Re: Forget buying drestin UNIX...) (Bob Hauck)
Re: on installing software on linux. a worst broken system. ("Dan J. Smeski")
Re: Sell Me On Linux (Mike Marion)
Re: Corel Linux Office 2000 and Win32 Emulator Making Progress (Damien)
Re: Sell Me On Linux (Mike Marion)
Re: Sell Me On Linux (Mike Marion)
Re: Sell Me On Linux (Mike Marion)
Re: on installing software on linux. a worst broken system. (abraxas)
Re: on installing software on linux. a worst broken system. (abraxas)
Re: on installing software on linux. a worst broken system. (The Cat)
----------------------------------------------------------------------------
From: Mike Marion <[EMAIL PROTECTED]>
Subject: Re: Sell Me On Linux
Date: Sun, 23 Apr 2000 00:18:39 GMT
SeaDragon wrote:
> Linuxconf? I hope you're joking. That is the slowest, most unreliable, most
OK, on that poing I agree. I don't like Linuxconf either.
> software I've ever used in my life. As for telnet et al, Windows gives you
> the same thing - full access to the machine remotely. What more can you
> want? How can you possibly claim that Linux is superior when the solution
> for both OS'es (primarily) is to simply open a session on a remote machine?
So if your Windows server gets so bogged down that the TCP/IP stack eats it, or
it just won't respond, you can still get into the box, perhaps edit the
registry.. i.e. fix the box. Without being on the computer's console?
With a Sun running Solaris or Linux, or even a PC with Linux and a properly
configured serial console setup.. I can. I can even boot the machines into the
OS without using the GUI and/or bringing up networking. If there's a
configuration error, I can use vi (or any other text editor I want) and fix it.
Can you boot NT/2K that way and edit something buried in the registry?
BTW, opening a session on a remote machine is a very good solution for many
problems. For instance.. the engineers I support would all need massive compute
servers in their office to run what they run. Instead they can open a session on
another box to run large simuations, or submit a job via lsf. This of course
isn't just a linux thing (we actually use Solaris) but a Unix thing. MS is just
barely getting there with WTS. Even that isn't as robust as Unix solutions.
Although I admit that disconnecting from a session and reconnecting is nice...
which of course can be done with the right setup on Unix boxes too.
--
Mike Marion - Unix SysAdmin/Engineer, Qualcomm Inc.
What's nice about GUI is that you see what you manipulate.
What's bad about GUI is that you can only manipulate what you see.
------------------------------
From: Sierra Tigris <[EMAIL PROTECTED]>
Subject: Re: on installing software on linux. a worst broken system.
Date: Sun, 23 Apr 2000 00:39:37 GMT
Christopher Browne posted Apr 22 re: Re: on installing software on linux. a...
|> For example, lets say I run a program that doesn't behave exactly
|>like I want it to. When I'll have enough knowledge of programming I'll be
|>able to tweak the source code and recompile. Can you do that with Ports?
|
|
|OF COURSE YOU CAN!!! The ability to tweak the source code and recompile
|is pretty much the _point_ to the exercise of downloading the source code.
|
|It would be an _unbelievably stupid_ implementation of Ports that would
|not permit this.
|
|[Sigh... There are Linux partisans that choose to be ignorant of what
|is possible on *BSD, just as there are *BSD partisans that choose to be
|ignorant of Linux ways, just as there are WinVocates that choose to
|flame Linux in ignorance of how it works...]
I really don't appreciate you putting words in *my* mouth. I asked
an honest question and expected, and deserved, an honest answer without the
lecture. I am _not_ a Linux partisan nor have I ever stated anything about
BSD, for you to tag that last paragraph as a reply to my post is somewhat
insulting.
You had posted that Ports automatically downloade, compiled and
installed the applications. I asked if you could tweak it _before_ it got
compiled. All you needed to say was, "Yes you can." No need for this offensive
lecture about partisans. Perhaps you are a victim of the OS wars, put please
don't assume that all who post here are involved.
--
Da Katt
[This space for rent]
------------------------------
From: [EMAIL PROTECTED]
Crossposted-To:
comp.os.ms-windows.advocacy,comp.sys.mac.advocacy,comp.os.ms-windows.nt.advocacy,alt.flame.macintosh
Subject: Re: which OS is best?
Date: Sat, 22 Apr 2000 19:44:37 -0500
On Sun, 23 Apr 2000 00:37:12 +0100, Rodger Etz-Brown
<[EMAIL PROTECTED]> wrote:
>David Corn wrote (besides other stuff):
>
>> With NT's default security, if I have physical access to the SAM
>> database (which usually means I'd either A)be an administrator on the
>> box or B) have physical access to the machine and break into it) the
>> passwords can be brute-force cracked pretty easily if users choose
>> simple passwords (ie English words). If syskey is turned on,
>> forgetaboutit. You won't be able to get 'in'.
>David, you are mistaken about the above. Use a tool like LOphtCrack and
>use the SMB packet capture fascility. You don't need administrative
>rights in the domain to do this.
The original author spoke of file access - getting access to the
registry and cracking from there. That was what I responded to.
Syskey won't change the passwords sent over the wire.
>You are also wrong on the brute-force statement. It is easy to 'crack'
>'simple' passwords with a dictionary attack. Brute Force attacks use a
>character set and will try every possible combination of characters of
>that particular set. I am sure that you can imagine the success rate of
>this method.
I can. Now try it using LOphtCrack (with just a pure dump from the
registry) with syskey turned on. I can crack a typical SAM in a few
minutes with LOphtCrack and a fast machine; with syskey on (granted, I
only tried for a few hours with a P2/350 or so) I didn't crack
anything. Granted, that was before PWDUMP2. Again, to clarify, I'm
talking about attacks directly on the SAM database file - in other
words, somehow you bribe an administrator on the NT domain, get your
hands on this file, take it home, and start attacking it.
PWDUMP2 is new to me; from a quick glance on their site it looks like
it will dump a syskey-encrypted database, but their words on the page:
"NT Administrators can now enjoy the additional protection of SYSKEY,
while still being able to check for weak users' passwords"
doesn't really clarify if syskey gives any additional protection from
pwdump2 or not.
So, I'll need to retract my statement - syskey no longer gives the
protection it once did before PWDUMP2 was developed.
>Security is an issue on every operating system. How secure a system is
>depends mainly on the administrator looking after it, as well as the
>operating system (and other factors, i.e. physical access, etc ;-) ).
Agreed.
------------------------------
From: [EMAIL PROTECTED] (Stuart Krivis)
Subject: Re: Corel Linux Office 2000 dubious at best?
Date: Sat, 22 Apr 2000 06:29:41 -0400
Reply-To: [EMAIL PROTECTED]
On 13 Apr 2000 13:40:45 -0600, Craig Kelley <[EMAIL PROTECTED]> wrote:
>Itchy <[EMAIL PROTECTED]> writes:
>
>> This is a great opportunity for the folks at Applix.
>> They make a nice product, although I only used it briefly about a year
>> ago. I liked it better than StarOffice though. Much better.
>
>It really is shame that Appliware hasn't recieved the attention that
>it deserves.
>
They're their own worst enemy. I don't think applix could market their way
out of a paper bag. :-)
We tried to become an Applix reseller, and even placed an order for Applix
Anyware (we had 20 Javastations sitting and waiting). They never followed
through. Reseller agreements sent in vanished into a black hole, as did
inquiries.
Applix's ads seem to be pretty non-existent too. And to think that they had
the market to themselves at one time, yet failed to capitalize on it. It
reminds me of IBM and OS/2.
--
Stuart Krivis
*** Remove "mongo" in headers for valid reply hostname
------------------------------
From: [EMAIL PROTECTED] (Bob Hauck)
Crossposted-To: comp.os.ms-windows.nt.advocacy
Subject: Re: Grasping perspective... (was Re: Forget buying drestin UNIX...)
Reply-To: bobh{at}haucks{dot}org
Date: Sun, 23 Apr 2000 01:20:02 GMT
On Sat, 22 Apr 2000 21:00:45 GMT, SeaDragon <[EMAIL PROTECTED]> wrote:
>Facts why Linux sucks? How about a complete lack of asynchronous I/O
>for starters. A completely lack of application-level exception handling.
Async IO is partially implemented. What does "application-level exception
handling" mean in the context of an OS discussion? That would seem to be
a language issue rather than an OS issue, unless you are suggesting that
the OS should supply some sort of standard exception mechanism that works
with all languages. Which sounds pretty hard to me. NT's structured
exceptions seem to principally be useful for creating conflicts with C++
exception handling.
>A horrbibly inefficeint and slow string implementation.
I presume you are referring to C's string implementation. It is not
required to implement Unix programs in C.
>A window system which requires a context switch (!) whenever you move the
>mouse.
I really can't think of many ways to avoid that context switch. I suppose
you could put the GUI in the kernel, but that doesn't seem like such a
good idea (and a context switch might still be required if the kernel is
multithreaded). You could have the application do all the GUI processing,
but that seems to be wrong-headed as well. Perhaps you can elighten us as
to how you avoid a context switch when processing GUI events?
BTW, the fans of the microkernel would have us make even more context
switches than the typical Unix system does. Perhaps you mean that all
OS's should be single-threaded?
>I have met Linux users who are so new to computers, so unknowledgable
>about how computers work, that they couldn't give an intelligent discussion
[on smp performance, splay trees, heap allocators, virtual memory]
Are we talking about users or developers? I daresay that users of most
every OS don't know about those things either. Or are you in favor of
some kind of priesthood where only the few can touch the computer?
>I find that most Linux advocates are completeley technically ignorant of
>any serious technical issues, but instead harp on on legal and economic
I find that you know a lot of buzzwords.
>.. Why don't you try comparising your OS to something serious such as
>TOPS-20 or MVS for a change?
Last I checked, TOPS-20 died before it got a windowing system, and MVS
still doesn't have one either unless you count client terminals running
their own windowing system independent of the host. I guess that's the
"proper" way to resolve the issue of context switches when you move the
mouse, eh? Just don't have a window system.
--
-| Bob Hauck
-| To Whom You Are Speaking
-| http://www.bobh.org/
------------------------------
From: "Dan J. Smeski" <[EMAIL PROTECTED]>
Subject: Re: on installing software on linux. a worst broken system.
Date: Sat, 22 Apr 2000 20:19:29 -0500
Not in windows 2000...
"Matthias Warkus" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> It was the 22 Apr 2000 04:29:12 -0700...
> ...and test@myhome <test@myhome> wrote:
> > I wanted to install some rpm package
> > to try some application. ok, i do
>
> > it tells me it needs 5 others packages
>
> > now this one tells me that i need 3 other packages.
>
> > now this tells me i am missing 2 packages.
>
> > is this really the modern way of installing sw?
> >
> > we make fun of MS, yet, on windows, i never had to do this sort of
thing.
> > double click on setup.exe and all is done.
>
> Yeah. That's because on Windows, software ships in huge packages,
> complete with all the dependencies, which will mercilessly overwrite
> already installed libraries and dump DLLs all over the system anyway.
>
> Can you get a mail client, a text editor or a file manager for Windows
> in a package that's less than a megabyte in size?
>
> mawa
> --
> Blümchenpflücker!
> Bonsaigärtner!
> Beinrasierer!
> Beischlafbettler!
------------------------------
From: Mike Marion <[EMAIL PROTECTED]>
Subject: Re: Sell Me On Linux
Date: Sun, 23 Apr 2000 01:25:40 GMT
Bart Oldeman wrote:
> As for fp performance, we all know that Alpha's are the top of the bill,
> and that the Athlon is closing the gap a bit. But if you're on a certain
No arguments there..
> budget, it may be that you get more flops/bucks for an x86
> (esp. Athlon) than for an Alpha or Sparc.
For budget yes, the CPUs are getting there.. but when it comes to the whole
system including scalable I/O, etc... you can't beat Suns, HPs, etc. Of course
they cost more. :)
> Just did a test on the fp-intensive research program I run all the time,
> of course the optimizations were on in the compilations:
>
> (compiled with g77 from egcs 1.1.2, which is not the best Fortran compiler
> around, but good enough for me)
>
> Pentium II 400:
> real 0m34.797s
> user 0m33.580s
> sys 0m0.170s
>
> (compiled with f77: SC4.0 18 Oct 1995 FORTRAN 77 4.0)
>
> Dual (Ultra)Sparcv9 167:
> real 1:26.6
> user 1:22.6
> sys 2.4
Uh, you use one of the latest g77s vs f77 from 10/95?!? That f77s optimisation
is going to suck comparitively.
> Although we have lies, damned lies and benchmarks, it shows that I'm quite
> happy running this stuff on the PC instead of the shared Sparc facility
> :-).
Of course if it works for you, then that's the best solution. Of course I'd be
interested in seeing how the machines handled about 50000 spawned copies of the
same programs on each box... after being compiled with similar optimisations
too.
--
Mike Marion - Unix SysAdmin/Engineer, Qualcomm Inc.
Black holes are where God divided by zero.
------------------------------
From: [EMAIL PROTECTED] (Damien)
Crossposted-To: comp.os.ms-windows.nt.advocacy,alt.destroy.microsoft
Subject: Re: Corel Linux Office 2000 and Win32 Emulator Making Progress
Reply-To: [EMAIL PROTECTED]
Date: 23 Apr 2000 01:32:10 GMT
On Sat, 22 Apr 2000 23:17:46 GMT, in alt.destroy.microsoft,
Roger <roger@.> wrote:
| On Sat, 22 Apr 2000 15:59:23 -0400, someone claiming to be T. Max
| Devlin wrote:
|
| >This is the thing that interests me. What are the issues and disputes that
| >could arise from running MS Office on Wine? Wouldn't this be a violation of
| >the EULA (I seem to recall one of those outrageously excessive clauses I was
| >forced to agree to saying something about "you can only run this on the os
| >which we allow you to", that being, of course, Microsoft (c) (tm) (r) (pat.
| >pend.).
|
| This was the copy that they held the gun to your head to make you
| install on that Gateway notebook which doesn't exist?
|
| There are no such clauses in the EULA for OFfice 2000 nor Office 97
| SR2, those being the only versions I have access to currently to
| check.
There is no reason to believe that the EULA is the same for all
shipped copies of MS Office, nor that the EULA will change in the
future. I would not be surprised if MS changed the EULA for all their
Windows software to stop users from running them in non-Windows
environments. In fact I hereby predict that it will happen when (if)
a large corporation starts using WINE with MS apps on a moderate scale.
------------------------------
From: Mike Marion <[EMAIL PROTECTED]>
Subject: Re: Sell Me On Linux
Date: Sun, 23 Apr 2000 01:40:37 GMT
SeaDragon wrote:
> Incorrect. For example, the EV6 has roughly double the performance of
> EV5 in integer performance at the same clock speed. Most benchmarks
> increase linearly with clock speed. Most normal work that most people
> do is integer work.
You think all the graphical computations needed to display icons, smooth fonts,
etc on that screen is pure integer? It might not max fp usage, but fp is
definately needed.
> I/O, yes, floating point, huh? You think all of the corporate customers
> you run web sites, run word processors, do programming need floating
> point? What are you talking about? It's only application is in simulation
> which is in the scientific and engineering fields.
Really? Shit, even today's games do serious fp work. So when I'm playing a
game I'm running sims for "scientific and engineering fields.." cool. :)
> Proof? FYI, Willamette will have a 400 MHz bus which will be the
> fastest in the industry. K7 already has the same bus frequency
> of the EV6.
Will have... and you think every other architecture is just going to stop
developing?
> The EV6 has a 200 MHz FSB. Same as K7! So how is the K7's bus "very slow
> when compared to other systems"? I am extremely interested in hearing
> this, especially which systems you are talking about.
There still aren't mobos with make full use of even Athlon systems. I know..
when they come out with the mobos that do, hopefully with DDR RAM, I'll be
buying one. :)
However, my current Athlon 650 is pretty sweet for my use.
> No shit sherlock? You mean all of those 1 GHz machines should be using
> 1 NS memory? Where do I buy it? Man, get a clue. Have you ever heard of
> a thing known as "cache"?
As if cache is hit 100% of the time.
> But in practice all execution is done in cache. The only time bus
> speed matters is for an L2 cache miss. That's why EV6 has 4 MB
> of L2!
And most Sun CPUs have 4Meg and even 8 now... x86 boxes on the other hand...
> All good in theory, except for the fact that X86's do indeed have equal
> (and in the future, higher) bus bandwidth than the RISC chips (c.f. K7
> and EV6).
Bzz wrong. Right now Suns UPA beats all x86 offerings, and have had levels for
a couple years that beat the best x86 offerings of today. E450s have done
1.6Gb/s via UPA and 1Gb/s to PCI.. much better then current x86 offerings. The
massive E10ks do 12.8Gb/s across the backplane. It's not like Sun is going to
use the same bus forever.. they'll keep improving (and staying ahead of x86) too
I'm sure.
> ??? What the fuck does PCI have to do with the CPU? This is all handled
> by the chipset and has absolutely nothing to do with the CPU. Are you
> arguing systems, or CPU's?
We're arguing architechtures... which are comprised of much more then just the
CPU.
--
Mike Marion - Unix SysAdmin/Engineer, Qualcomm Inc.
If at first you don't succeed, you must be using Windows.
------------------------------
From: Mike Marion <[EMAIL PROTECTED]>
Subject: Re: Sell Me On Linux
Date: Sun, 23 Apr 2000 01:46:42 GMT
2:1 wrote:
> You have no clue. You deleted /etc/passwd ???
> Besides, it is not normally huge.
Actually they do get huge. Ours is over 11,000 lines long. Not only that but
we use NIS, so lookups go over the network. Yet even with this horrible design
as he puts it.. that ls -alR I do, on a directory with tons of files, comes back
almost the instant I hit return. Funny that parsing of the file doesn't make it
take forever eh? :)
Oh yeah, that's even on directories mounted via NFS.. which means more reads
across the network for each file too.
--
Mike Marion - Unix SysAdmin/Engineer, Qualcomm Inc.
ZMODEM: Big bits, Soft blocks, Tighter ASCII...
------------------------------
From: Mike Marion <[EMAIL PROTECTED]>
Subject: Re: Sell Me On Linux
Date: Sun, 23 Apr 2000 01:52:34 GMT
Christopher Browne wrote:
> Cache doesn't matter if it's getting flushed continually due to there
> being disk I/O that pushes 100MB of data through. You can't afford a
> 100MB cache...
Hell that wouldn't do it either. We have engineers running development apps
that use 100's and even Gigs of RAM all the time. Not to mention how much disk
I/O comes into play.
> /etc/passwd is one of the few config files where growth to large sizes
> tends to occur, and thus, there is an alternative mechanism whereby the
> file gets "compiled" into a DBM file so that getpwent() becomes an O(1)
> operation.
For instance, NIS doesn't do a line by line search. It does a match on the
database held on the NIS slave you're attached to.
BTW, to show the improvement (using NIS anyway) just do a few 'ypcat passwd |
grep user' vs ypmatch user passwd and look at the difference.
--
Mike Marion - Unix SysAdmin/Engineer, Qualcomm Inc.
Well, what else is important to them? As far as stimulants go, both of our
generations know the feeling of jonesing for product from Columbia; it's just
that their product is coffee. -- Dennis Miller on GenX
------------------------------
From: [EMAIL PROTECTED] (abraxas)
Subject: Re: on installing software on linux. a worst broken system.
Date: 23 Apr 2000 01:56:43 GMT
test@myhome wrote:
*snip* an excellent argument about how much RPM sucks ass.
Your linux alternative is a healthy debian install. The deb
package system is exactly what youre looking for.
Your other alternative is to chuck linux altogether and
use freebsd. :)
=====yttrx
------------------------------
From: [EMAIL PROTECTED] (abraxas)
Subject: Re: on installing software on linux. a worst broken system.
Date: 23 Apr 2000 01:58:20 GMT
test@myhome wrote:
> see, I wanted to install etherape:
>>rpm -Uhv etherape-0.5.3-1.i386.rpm
> error: failed dependencies:
> gnome-libs >= 1.0.0 is needed by etherape-0.5.3-1
> libglade >= 0.11 is needed by etherape-0.5.3-1
If you didnt get gnome-libs or libglade with your SuSe install,
then you did not install the entire distribution.
Besides that, I believe SuSe uses the deb package manager as well.
You may want to look into that.
=====yttrx
------------------------------
From: The Cat <[EMAIL PROTECTED]>
Subject: Re: on installing software on linux. a worst broken system.
Date: Sun, 23 Apr 2000 02:02:51 GMT
Looks like Gnome libs were not installed for some reason despite you
doing a full install.
TheCat
On 22 Apr 2000 04:29:12 -0700, test@myhome wrote:
>lets talk a little about the broken way of installing software on linux.
>
>it is most certinaly is a broken system now.
>
>a simple example. I wanted to install some rpm package
>to try some application. ok, i do
>
> rpm -Uhv foo.rpm
>
"Agent under Wine and powered by Mandrake 7.0"
------------------------------
** 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.advocacy) 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-Advocacy Digest
******************************
