Linux-Advocacy Digest #999, Volume #29 Thu, 2 Nov 00 05:13:04 EST
Contents:
Re: Ms employees begging for food (Peter Desnoyers)
Re: Windoze 2000 - just as shitty as ever ("Ayende Rahien")
Re: so REALLY, what's the matter with Microsoft? ("JS/PL")
Re: Windoze 2000 - just as shitty as ever ("Ayende Rahien")
Re: so REALLY, what's the matter with Microsoft? ("Ayende Rahien")
Re: 2.4 Kernel Delays. ("Ayende Rahien")
Re: to all who responded (Terry Porter)
Re: IBM to BUY MICROSOFT!!!! ("Ayende Rahien")
Re: Linux (Terry Porter)
Re: A Microsoft exodus! ("Lennart Gahm")
----------------------------------------------------------------------------
From: [EMAIL PROTECTED] (Peter Desnoyers)
Crossposted-To: comp.os.ms-windows.nt.advocacy,comp.arch,comp.os.netware.misc
Subject: Re: Ms employees begging for food
Date: Wed, 01 Nov 2000 21:12:12 GMT
>In article <[EMAIL PROTECTED]>,
>T. Max Devlin <[EMAIL PROTECTED]> wrote:
>>
>>I'm afraid that's a somewhat dubious value. An "ethernet" can run at
>>100% utilization, or 99%, but the question is how much of that
>>utilization is 'collisions', and even "resends" due to higher layer
>>connectivity issues. A non-switched, shared media ethernet with a
>>nominal 20 transceivers (potentially including both a router and a
>>server, to characterize the logical and software topologies) would be
>>hard pressed to maintain the "rule of thumb" 30% utilization *and do so
>>while providing sufficient throughput*. This presumes, of course, we
>>don't take the easy way out by defining that 18 of those transceivers
>>will rarely transmit anything at all.
>>T. Max Devlin
30% (or actually 1/e) is the throughput of slotted aloha, a purely
fictitious protocol which exists only as the basis of proofs which can
be extended to more realistic protocols.
To extend on this more than T. Max is likely to want to (or be willing
to) hear I'll give my media access protocol throughput stump speech.
If you want more detail, you can check out Bertsekas and Gallager,
chapter 4, for the full gory details.
First, the simplifying assumptions. We assume an infinite number of
stations on the network - it turns out that this gives the most
conservative results. Each station tries to transmit at an equal
(infinitessimally small) offered load, and actually transmits at a
real rate less than or equal to that. We'll call the total offered
load equal to L, defined to include all the bits in the packet, so
that if L=100% and we had a completely efficient protocol we could fit
every bit on the wire with no room to spare. Finally, we'll assume
that all packets are the same size, N bits. If we want to know the
max throughput, we set N to 1500bytes. If we want to know throughput
for a realistic packet mix, it's close to what the equations will give
us for the average packet length. Finally, we measure time in bits,
because it's easy that way.
Protocol #1: slotted Aloha. Divide time into slots of N bits, with
all transmitters exactly synchronized to these slots through telepathy
or some such method. If you want to send a packet, wait to the
beginning of the next slot, transmit it, and pray that there isn't a
collision. When L=100%, the probability that a packet doesn't collide
and makes it through OK is 1/e, and we get a throughput of 36%.
(ignore what happens when you get a big backlog because 100%>36%,
which increases the effective offered load, driving it into congestion
collapse...)
Protocol #2: unslotted Aloha. Don't divide time into slices, so each
packet can start at an arbitrary time. Now a packet can collide with
any packet starting within a window of 2N bits (instead of N for #1),
so the peak throughput is 1/2e or 18%.
Protocol #3: unslotted Aloha with carrier sense and collision
detection. In this case a transmitter can detect any packet which
started transmission more than S bits ago, and can detect a collision
within S bit times, where S < N. In the case of 10baseT, S is 512
bits, and is the maximum round-trip time of the largest legal network.
In this case you can think of each packet transmission taking a
collision round (2e slot times, or 5.4 S) followed by N-S bits of
unmolested packet transmission.
For 10BaseT with 1500 byte packets, N=12000 and S=512, so max
throughput is 12000/(12000 + 4.4 * 512) or 84%. But you still get
into collapse if your offered load ever exceeds that figure, because
you get into a mode where excessive collisions cause throughput to
decline.
Protocol #4: protocol #3 plus binary exponential backoff. (this is
easier to analyze if you apply it to protocol #2 or #1, actually...)
Each transmitter keeps a timer value T which it initializes to a small
number when it starts to transmit a packet. If it fails to transmit
due to a collision, it picks a number X randomly from 0..T, and waits
for X*S bit times. It also multiplies T by 2.
The result is that if the offered load exceeds the stable operating
load, more stations will spend more time waiting in backoff rather
than transmitting, and the ethernet will continue to operate in a
stable fashion.
............................................................................
Peter Desnoyers (781) 392-2216
InfoLibria Inc.
411 Waverly Oaks Rd., Waltham MA 02154 mailto: [EMAIL PROTECTED]
------------------------------
From: "Ayende Rahien" <[EMAIL PROTECTED]>
Crossposted-To: alt.destroy.microsoft,comp.os.ms-windows.advocacy
Subject: Re: Windoze 2000 - just as shitty as ever
Date: Thu, 2 Nov 2000 10:16:35 +0200
"Mark Lindner" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> Ayende Rahien wrote:
>
> > I would say that you're a rare case. Practically everyone that I've
talked
> > to agrees that 2000 is (much) more stable than NT.
>
> In response to my grips about problems, crashes, incompatibilities,
lockups,
> etc., Windows advocates always tell me I'm the rare case. I guess that
makes
> me the "common rare case." Somehow, I find it hard to believe that I am
> somehow jinxed and that Windows fails more miserably for me than for
anyone
> else. Far more likely that I just have higher expectations from an OS than
the
> typical Windows user does. ;-)
No, please read what I said.
I didn't say that windows doesn't has
problems/crash/incompatibilities/lockups and so on, what I said is that
windows 2000 has by far less problems than NT.
> > Windows wins hands down.
> >
> > When you are talking about pleasent of use, windows GUI is by far
superior
> > to any thing that linux or unix produced so far.
> > If you are talking about other aspects of win vs *nix, then it's
different
> > answer for every aspect.
>
> This is an aesthetic issue rather than an architectural issue. Personally
I
> think the Windows GUI looks like puke, but then again, I wasn't weaned on
> Windoze from day 1, so I don't have the "It's what I've always used so I
like
> it more than anything else" syndrome (a common ailment among Windows users
as
> well as COBOL developers, it seems)...in fact the first time I touched
Windows
> was in my first job out of college, but within a few months I'd dumped the
> Pee-Cee in favor of an Indigo workstation, because coming from a Solaris
> background, I found that trying to develop code on a Windoze box was like
> eating soup with a fork...no command-line tools, shit for a shell, no
> scripting support (sorry, Batch files don't count)...no way to automate
> anything for that matter...poor implementations of telnet, ftp, etc....
FTP & Telnet are hard to find on a non 2000 windows machine, very easy if
you've them.
Scripting support? Vbscript, JScript, WSH, batch files (why doesn't they
count?)
Automation? Scheduled tasks, you can get that in 95 if you install IE 4.
In win2000 you get this as weel as AT, which is a cli tool that gives you
even more power.
The lack of command line tools is not because most people like GUI much
better.
> With Linux, you have a choice of GUIs, at least one of which is capable
of
> emulating the Windoze interface, with all its drab grey buttons and
windows.
> However, IMHO, a fully and properly-configured desktop that uses
Enlightenment
> + GNOME makes Windoze, in comparison, look like the uninspired, drab piece
of
> vomit that it really is.
Talk about bloatware.
If it's customization that you want, you can make windows look like anything
that you want with a little work.
If you don't know how to change the windows colors then you really need some
lessons in windows.
> There may be a case for GUI consistency in Windows, but hopefully the
> consistency of application UIs under UNIX will improve as Gnome/GTK+
become
> standard (and assuming developers discipline themselves).
Oh, really?
The best GUI for linux that I've found was the one that came with Corel
Linux 1.2, it was KDE, and felt very much like windows.
As for GUI consistency, it goes against OSS principals, isn't it?
KDE, Gnome, Heliz, Enlightment, AfterStep - on the top of my head, GUI for
linux.
I can find you people that will swear in the names of either of them.
How do you make them choose a standard?
When programming for windows, I can make my own GUI, which is tiresome task
most often, or can use windows own GUI, giving the user familiar feel.
Pick up any programming book and read the part about UI, you will find
consistency mentioned time and time again. OSS never head about it.
------------------------------
From: "JS/PL" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.ms-windows.advocacy,comp.os.ms-windows.nt.advocacy
Subject: Re: so REALLY, what's the matter with Microsoft?
Date: Thu, 2 Nov 2000 03:14:32 -0500
"mlw" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> JS/PL wrote:
> >
> > "Andy Newman" <[EMAIL PROTECTED]> wrote in message
> > news:[EMAIL PROTECTED]...
> > > In article <8tk12f$614$[EMAIL PROTECTED]>, Christopher Smith
wrote:
> > > >
> > > >Which "fixes" are present in Win98 you can't download for free ?
> > > >
> > > >Similarly with NT, perhaps you've heard of service packs ?
> > > >
> > >
> > > You didn't get it did you. All versions of Win are fixes
> > > to the previous one. It's only recently with Win2K that
> > > they're getting the complete set of functions together in
> > > a package that's half-well implemented. It's taken long
> > > enough.
> >
> > Win2k isn't just a fix of NT4.
>
> True, this time they broke more than they fixed ;-)
What are you talking about? What's broken? I've been using it since the
final betas and havent had a single system crash. What is broken in
Windows2000?
>
> > Windows.Net (Whistler) due out in the spring isn't a fix for WinME or
Win9x.
> > So you are wrong. All versions of Windows aren't fixes to the previous
one,
>
> No, but often the only fix to a Windows problem is to purchase an
> upgrade to a new version.
Give an example of a fix that was only available by purchasing an upgrade,
c'mon back up you fud with a fact.
>
> --
> http://www.mohawksoft.com
------------------------------
From: "Ayende Rahien" <[EMAIL PROTECTED]>
Crossposted-To: alt.destroy.microsoft,comp.os.ms-windows.advocacy
Subject: Re: Windoze 2000 - just as shitty as ever
Date: Thu, 2 Nov 2000 10:24:07 +0200
"." <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> > Please check pre-win95 days, there wasn't a registery then, see how much
fun
> > they had those days.
>
> You might have missed the 'protected directory' part of the comment. So
> many problems stemmed from any app being able to write to any part of the
> disk it felt like, adding and removing stuff haphazardly from a myriad of
> INI files.
In 9x, I believe that they still can.
> > No-GUI, this I know that they are doing.
>
> About goddamned time.
Agreed
> > I don't know why CLI shell from *nix is better than CMD.EXE, except
maybe
> > the lack of applications to it.
>
> The power of the CLI (with proper tools) is simply amazing. I can't
> explain it well enough to ensure you understand the full potential!
>
> kill -9 `ps aux | grep httpd | cut -b 5-10`
CMD.exe doesn't have tools,. big part of the problem.
And the philosophy that allows this on *nix is very different than the one
that built DOS
> All 'fancy features' of little or dubious use.
They have their use, user comport, they just forgot that most people would
like to mutter about ASCII art than shout about "where did my term paper,
due tomorrow, has vanished?"
------------------------------
From: "Ayende Rahien" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.ms-windows.advocacy,comp.os.ms-windows.nt.advocacy
Subject: Re: so REALLY, what's the matter with Microsoft?
Date: Thu, 2 Nov 2000 10:28:18 +0200
"JS/PL" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
>
> "mlw" <[EMAIL PROTECTED]> wrote in message
> news:[EMAIL PROTECTED]...
> > JS/PL wrote:
> > >
> > > "Andy Newman" <[EMAIL PROTECTED]> wrote in message
> > > news:[EMAIL PROTECTED]...
> > > > In article <8tk12f$614$[EMAIL PROTECTED]>, Christopher Smith
> wrote:
> > > > >
> > > > >Which "fixes" are present in Win98 you can't download for free ?
> > > > >
> > > > >Similarly with NT, perhaps you've heard of service packs ?
> > > > >
> > > >
> > > > You didn't get it did you. All versions of Win are fixes
> > > > to the previous one. It's only recently with Win2K that
> > > > they're getting the complete set of functions together in
> > > > a package that's half-well implemented. It's taken long
> > > > enough.
> > >
> > > Win2k isn't just a fix of NT4.
> >
> > True, this time they broke more than they fixed ;-)
>
> What are you talking about? What's broken? I've been using it since the
> final betas and havent had a single system crash. What is broken in
> Windows2000?
My bank account after getting it?
------------------------------
From: "Ayende Rahien" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.ms-windows.nt.advocacy,comp.os.ms-windows.advocacy
Subject: Re: 2.4 Kernel Delays.
Date: Thu, 2 Nov 2000 10:32:53 +0200
"Aaron R. Kulkis" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> Ayende Rahien wrote:
> >
> > "Colin R. Day" <[EMAIL PROTECTED]> wrote in message
> > news:[EMAIL PROTECTED]...
> > > Erik Funkenbusch wrote:
> > >
> > >
> > > >
> > > > Which will run a data center for at most, a few hours. What happens
if
> > an
> > > > earthquake hits, or a plane crashes into the data center, or any
number
> > of
> > > > other natural catastrophies that might occur to a single site.
> > > >
> > >
> > > Great. And what happens if a nuclear missile hits Redmond?
> > > Linux, on the other hand, you'd have to nuke half the planet.
> >
> > I don't think so, all you need to do is to kill one man, Linus, and then
the
> > linux compunity is going to be:
> > A> In shock
> > B> Un-unified.
> >
> > Very soon there will be no official kernel, no one with the autority to
> > release it, Linux will split up to various groups which will be totally
> > incompatible. Reasonable people will move to BSD.
>
> A new leader will be appointed, and unification will resume.
By who? You are going to have an election?
Shorten your sig.
------------------------------
From: [EMAIL PROTECTED] (Terry Porter)
Subject: Re: to all who responded
Reply-To: No-Spam
Date: 02 Nov 2000 08:31:26 GMT
On Thu, 02 Nov 2000 02:24:54 GMT, C. Nolan McKinney
<[EMAIL PROTECTED]> wrote:
>Thanks, everyone, for the responses. The more I read here and on web pages,
>the more I realise that I need to read more! I will in the next few weeks
>read up on Linux installation, as I am already sold on its benefits.
>
>Again, thanks,
>Nolan
>
>--
>
>
>
Wise idea Nonan, in 1997 when I installed Linux, I had already purchased 2
books, "Unix for the impatient" and "running Linux" by O'riely. I read the
"Running Linux" book 3 times before I even attempted to install Linux, and
I didnt have any problems at all.
I'm still reading "Linux for the Impatient" as its designed to be direct
with no hand holding, and there is so MUCH to learn about Unix, if one is
interested.
Of course the net will provide a huge wealth of How-to's and FAQ's about every
kind of Linux area you could possibly want.
Good luck and take your time, put Linux on a secondhand old pc, and "play" with
it, and allow a few years to become knowledgeable.
--
Kind Regards
Terry
--
**** ****
My Desktop is powered by GNU/Linux, and has been
up 2 weeks 4 days 4 hours 22 minutes
** Registration Number: 103931, http://counter.li.org **
------------------------------
From: "Ayende Rahien" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.ms-windows.advocacy,comp.os.ms-windows.nt.advocacy
Subject: Re: IBM to BUY MICROSOFT!!!!
Date: Thu, 2 Nov 2000 10:38:51 +0200
"Aaron R. Kulkis" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> Bruce Schuck wrote:
> > broken root and is running the box nice and smooth so it will be up and
>
> As opposed to NT, where all one has to do is get on the machine,
> and there is not need to 'break root'...because you already have it!
Oh? If I get to a NT box, I'm suddenly administrator? Do tell me how.
And shorten your sig while you are at it.
------------------------------
From: [EMAIL PROTECTED] (Terry Porter)
Subject: Re: Linux
Reply-To: No-Spam
Date: 02 Nov 2000 09:22:45 GMT
On 02 Nov 2000 04:46:14 GMT, JoeX1029 <[EMAIL PROTECTED]> wrote:
>>On 01 Nov 2000 01:36:46 GMT, JoeX1029 <[EMAIL PROTECTED]> wrote:
>>>>On 27 Oct 2000 23:27:37 GMT, JoeX1029 <[EMAIL PROTECTED]> wrote:
>>>>>i tend to agree. GNU/Linux was not and is not for the desktop.
>>>>
>>>>I tend to disagree, GNU/Linux is most certainly for the desktop :)
>>>>
>>>>I should know, its been on my desktop exclusively since August97.
>>>>
>>>>
>>>>Kind Regards
>>>>Terry
>>>>--
>>>
>>>What i meant was Linux is not ready for the mass desktop (ie average windoze
>>>user).
>>Sorry but I have to disagree, the average Windows user doesnt know a lot
>>about anything, its all point n click.
>I know thats why i said GNU/Linux is NOT for the desktop.
Oh but it is, GNU/Linus IS point and click too :-)
>>
>>> Moreover, I dont think it should be. Im decently certain theres no way
>>>to make it completely idiot friendly (refer to mass desktop) and still
>>reatin
>>>the security/scalibilty/roubustness it currently enjoys.
>>This just doesnt make sense to me ...
>>Windows has NO security and no remote admin, Linux has these.
>Thats what im saying! GNU/Linux cant be made like windoze (for the average
>idiot) and still be a excellent, stable etc platform.
???
I believe it already is, and I have proof :)
there is fellow, whose pc I installed Linux on about 2 years ago, hes comp
illiterate, totally. Ive telnetted into his pc via the net a couple of times,
to update some software for him (the last about 12 months ago) and hes still
on the net, still IRCing, and doesnt even have a single BOOK about Linux!
His Linux system is still running, no probs, and he wouldnt know what the CLI
is!
>>A properly installed Linux box, would reduce the worldwide pain of having a
>>desktop by a huge margin!
>>
>>No IloveYOU virii, no overwritten dlls, no lost security etc.
>i fscking hate those damn dll's....
>>
>>Kind Regards
>>Terry
>>--
>
>
--
Kind Regards
Terry
--
**** ****
My Desktop is powered by GNU/Linux, and has been
up 2 weeks 4 days 4 hours 22 minutes
** Registration Number: 103931, http://counter.li.org **
------------------------------
Crossposted-To:
comp.os.ms-windows.advocacy,comp.os.ms-windows.nt.advocacy,comp.os.os2.advocacy,comp.sys.mac.advocacy,comp.unix.advocacy
From: "Lennart Gahm" <[EMAIL PROTECTED]>
Reply-To: "Lennart Gahm" <[EMAIL PROTECTED]>
Subject: Re: A Microsoft exodus!
Date: Thu, 02 Nov 2000 09:38:34 GMT
On Thu, 2 Nov 2000 09:18:46 +1000, Christopher Smith wrote:
>
>"lyttlec" <[EMAIL PROTECTED]> wrote in message
>news:[EMAIL PROTECTED]...
>> Christopher Smith wrote:
>> >
>> > "Weevil" <[EMAIL PROTECTED]> wrote in message
>> > news:VyHL5.1544$[EMAIL PROTECTED]...
>> > >
>> > > > How hard would it be to write something like this for unix/linux?
>> > > > If I understand the way those viruses worked, you open the attach
>file,
>> > it
>> > > > reads the adress book and sent it to the first 50 people there,
>right?
>> > > > Can't you do the same in unix? (I'm asking, not insulting. I want to
>> > know
>> > > > the answer.)
>> > >
>> > > It used to be a truism that you can't get a virus merely by reading
>your
>> > > email. Microsoft changed all that (to the shock of most
>> > security-conscious
>> > > observers) when they added Visual Basic macro capability to their
>email
>> > > program. Any of their apps which use such macros (e.g. Word) are
>> > obviously
>> > > vulnerable.
>> >
>> > This is false. You have to *deliberately* double click on the
>attachment
>> > and answer yes to the resultant dialog box (which defaults to no) to
>> > open/run it. It's a long way from getting a virus "just by reading
>email".
>> >
>> I understood you could get the virus via the "preview" feature without
>> even opening the e-mail. Is this true or not?
>
>Not as far as I know.
>
>There was, I believe, a buffer overrun that caused some problem along those
>lines a while back, but that's a long way from it being a design choice and
>deliberate feature.
Take a look hove microsoft has done on w98. There they have added something
they calls "windowsupdate". Clicking on that takes you to a website there you
can click on "productupgrade". This will start some sort of scriptfunction
that looks through your system and lists upgrades you can install, and
upgrades you already have installed (many is for security holes in IE/OL).
In microsofts case, you have to answer some dialogboxes and click on the
things you want.
But what will prevent any hacker from putting this sort of script embedded
in html code without dialogs? Script that can do anything with your system.
IE and outlook uses the same "security" settings and has this sort of script
active by default. What happens if someone sends you an html-formatted email,
with scripts in it (without dialogs)? Does they get executed?
One can also note that if you turn scripts off, the windowsupdate feature
will
no longer works. Good luck all you lookout users.
/Lennart
------------------------------
** 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
******************************
