Linux-Advocacy Digest #102, Volume #27 Thu, 15 Jun 00 17:13:03 EDT
Contents:
Re: Linux Tast Test (Cihl)
Re: Linux MUST be in TROUBLE (Cihl)
Re: An Example of the Superiority of Windows vs Linux (Cihl)
Re: Number of Linux Users (Michael Born)
Re: An Example of how not to benchmark (Pete Goodwin)
Re: Would a M$ Voluntary Split Save It? (Leslie Mikesell)
Re: An Example of how not to benchmark (Cihl)
Re: Number of Linux Users ("Christopher Smith")
Re: ApplixWare? More Build It As You Go Along Linux....
Re: Number of Linux Users (Cihl)
Re: OT Aboriginal Lifespans was Re: Canada invites Microsoft north (Bob Lyday)
Re: Microsoft migrates Hotmail to W2K (abraxas)
Re: OT Aboriginal Lifespans was Re: Canada invites Microsoft north (Eric Remy)
----------------------------------------------------------------------------
From: Cihl <[EMAIL PROTECTED]>
Subject: Re: Linux Tast Test
Date: Thu, 15 Jun 2000 20:13:06 GMT
David Steinberg wrote:
>
> aflinsch ([EMAIL PROTECTED]) wrote:
> : > Does this go for new versions, too? Or only older versions, like
> : > RH5.2, or something.
>
> : It goes for whatever they have in stock, in fact some of the older
> : distros might be yours for the cost of shipping.
>
> And they pretty much always have the current release of every distribution
> in stock.
>
> If you thought you actually have to pay more than $5 to get the most
> up-to-date version of a Free operating system, I guess this news is a
> pleasant surprise! :)
>
> (On the other hand, there are still obvious benefits to buying a distro
> with a manual and support.)
>
> --
> David Steinberg -o)
> Computer Engineering Undergrad, UBC / \
> [EMAIL PROTECTED] _\_v
Nah.. I was just curious. I have a cable modem so i just download and
try them all. My ISP (bART) is very happy with Linux-users, so they
don't care how many gigs of it i download. They are even willing to
set up BSMTP for you for a small fee.
------------------------------
From: Cihl <[EMAIL PROTECTED]>
Subject: Re: Linux MUST be in TROUBLE
Date: Thu, 15 Jun 2000 20:16:13 GMT
Pete Goodwin wrote:
>
> [EMAIL PROTECTED] (Cihl) wrote in <[EMAIL PROTECTED]>:
>
> >Yeah, and we've chosen to abduct all the people who post in this
> >newsgroup who like Windows. You're not wanted on your planet anyway!
>
> Yes, PLEASE, abduct me! Anything has got to be better than Earth!
>
> 8)
>
> Pete
We don't really know yet what OS you humans are running, so we're
going to have to open up your skull and pull out your brain for
examination.
------------------------------
From: Cihl <[EMAIL PROTECTED]>
Subject: Re: An Example of the Superiority of Windows vs Linux
Date: Thu, 15 Jun 2000 20:17:37 GMT
Pete Goodwin wrote:
>
> [EMAIL PROTECTED] (Charlie Ebert) wrote in
> <[EMAIL PROTECTED]>:
>
> >Pete,
> >
> >There are some 1,000 odd computer magazines world wide.
> >Go find me 10 of them which indicate by a performance graph
> >that Linux is slower than Windows.
> >
> >Just 10 is all I ask.
> >
> >You know, these magazines couldn't sell magazines if everybody thought
> >they were full of shit.
> >
> >Further, they'd get sued.
> >
> >So there you go Pete.
> >
> >Walnuts to A-bombs, your wrong.
>
> 1000000 magazines wouldn't make a difference. From my perspective, Linux is
> slower than Windows. My own tests tell me that.
>
> Pete
Linux is slower for you, because it adapts automatically to your
comprehension speed. It's a separate thread of the WindowsUser-daemon.
------------------------------
From: Michael Born <[EMAIL PROTECTED]>
Crossposted-To: comp.os.ms-windows.nt.advocacy
Subject: Re: Number of Linux Users
Date: Thu, 15 Jun 2000 20:25:55 GMT
If a product has increasing market share each year (which Linux has
achieved in the server os market), they are taking over.
Drestin Black wrote:
> "Michael Born" <[EMAIL PROTECTED]> wrote in message
> news:[EMAIL PROTECTED]...
> > Where Linux is superior now (as a server), it is in fact taking over.
>
> really? how does being in the minority indate "in fact taking over."?
------------------------------
Subject: Re: An Example of how not to benchmark
From: [EMAIL PROTECTED] (Pete Goodwin)
Date: Thu, 15 Jun 2000 20:27:36 GMT
[EMAIL PROTECTED] wrote in <8ib50k$oic$[EMAIL PROTECTED]>:
>So I had a close look at the configuration information the renderer
>outputs at the start of a run, and found a discrepancy in the "Bounding
>Threshold" --- under Windows, it was 3, under linux 25. In both cases,
>this was obviously the compile-time default; The original Povray source
>has 25, so it seems that whoever does the Windows port decided to change
>it.
[snip]
>Next, I reran the tests, using the "res640.ini" file on both platforms,
>and not enabling antialiasing on either (because at quality 9, the
>default, it gets disabled anyway). As expected, the Windows version
>still needed 21 minutes (20:58 to be precise). But the linux version
>needed only 19:29 minutes, i.e. a minute and a half less. Some of the
>statistics were still off in the final digits, but it was quite clear
>that, apart for minor variations caused by different rounding, the same
>thing was calculated on both platforms.
>So, in the end, getting the source and compiling it yourself with gcc
>for linux will give you a 7.5% speed advantage over getting the
>Windows version (not to mention that the download is considerably
>smaller).
8)
I built a version of POVray on Windows with VC++ and selected Pentium
class. I took a close look at the defaults used by this version, and its
the same as Linux, Bounding Threshold = 25
Now I ran this version and it does run slower, it generates the image in 26
minutes and 36 seconds.
Um, except, that's still faster than the 686 version on Linux which runs
for 32 minutes and 36 seconds.
Now, I've no idea why we should differing results, but it still looks like
Windows is a bit faster than Linux!
Pete
------------------------------
From: [EMAIL PROTECTED] (Leslie Mikesell)
Crossposted-To:
comp.os.os2.advocacy,comp.os.ms-windows.nt.advocacy,comp.sys.mac.advocacy
Subject: Re: Would a M$ Voluntary Split Save It?
Date: 15 Jun 2000 15:29:25 -0500
In article <%uU15.1665$[EMAIL PROTECTED]>,
Daniel Johnson <[EMAIL PROTECTED]> wrote:
>> If the window is resized (MS-windows is windows-aware, right?) a
>> telnet program should notify the other end of the new size
>> via the telnet protocol.
>
>I can't find anything about this in RFC 764, which seems to be
>the most refect RFC on the telnet protocol, and purports
>to define it. What am I missing?
You are missing the fact that standards evolve to meet users
needs and that large companies should pay attention instead
of going their own way. An overview of telnet-related
standards is at http://www.freesoft.org/CIE/Topics/110.htm
and specifically 'Negotiate About Window Size' is in RFC 1073.
>> >> It has nothing to do with terminals or terminal emulation.
>> >
>> >Wasn't your original complaint that the vt100 emulation was
>> >broken?
>>
>> That too, but I haven't really investigated the breakage so
>> I'm willing to blame DEC for giving us a dozen different
>> things call vt100's for this part,
>
>No, no, there's only one VT100. But there's also the
>more advanced VT102 and, in fact, many others.
History is much stranger than it appears historically. There
were in fact, options to VT100 terminals before there were
named variations. I think some were locally configurable,
some were extra cost add-ons. The character-insert and
character-delete functions were among them.
>> although these days
>> everyone should just match up with what xterm does.
>
>Why? Just because it is popular on Unix?
Because it does what everyone finally concluded a canonical
vt100 would have done if it could... That is, the reason
everyone uses xterms instead of real terminals or consoles
is that it presents a predictable and well-known feature set
(unlike most vendors of real hardware).
>I could see that, actually, but that would also arguing
>for adhering to extentions popular on Windows also,
>surely.
Of course we should have extensions. As soon as everyone
agrees on them and the standard is updated.
>
>> Fixing
>> the terminal emulation can be done from the host side
>> by using a matching termcap/terminfo description for the
>> broken operations.
>
>Well, except for things that aren't implemented at all- like
>variable screen size.
But it does vary the screen size. It just doesn't follow
the standard to let the program displaying on it know.
>> >> Unix has an elegant simplicity.
>> >
>> >That is not the same thing as being easy for you to use.
>>
>> Yes it is, because there is an underlying consistancy
>> that is mostly possible to understand.
>
>I'm afraid in practice it isn't. Most people find it very
>difficult to deal with the level of abstraction Unix
>asks for.
For most of unix's history it was used mostly in multi-user
configurations where an experienced administrator would
be expected to do the initial configuration and perhaps build
custom shell scripts for a user's frequent tasks. Now that
everyone is fending for themselves, it is playing catchup
in the realm of presenting a limited set of canned operations
in GUI menus. However it is easier to dumb down a general
interface than to add generality to a limited set of choices,
so if you need both unix is already the better choice.
>> Windows seems
>> to take the approach that the system is impossible to
>> understand and must be hidden from the operator.
>
>Oh, I don't know. I don't find it impossible to understand.
How do you do an operation involving several programs
repeatedly, varying only one step?
>> This
>> is fine when everything works and they have anticipated
>> what you want to do. It is not so fine when you try to
>> combine things in an unanticipated way or you have to
>> try to troubleshoot anything.
>
>I think that it is still preferable to the Unix 'elegant simplicitly'
>approach for most users, because most users couldn't trouble
>shoot Unix at all. They can at least do a little troubleshooting
>in Windows with those troubleshooter thingies, such as they are.
>
>Combining things is a great use of the 'elegant simplicity' idea, but
>it's beyond most of the users out there. Or at least the Unix way
>of doing it is.
Maybe, but you can hide the underlying simple tools with a
wrapper - GUI or shell script, or both. If you don't need to vary the
options or combinations you never need to know what is under the
covers. If you do, you still have complete access to the
individual tools.
>> >Those may be better telnet programs, but so far what I've
>> >heard from you is that MS's telnet's problem is that it
>> >can't do anything better than a lousy vt100.
>>
>> You could work around that, but there is no excuse for a
>> windows program not doing the correct thing when resized.
>
>Perhaps it's just being *strictly* standards compliant, with no
>extensions. That *is* what you want, right?
I want the extensions as soon as possible after they become
standards themselves.
>[snip]
>> >Well, what you call "annoying flaws" other people call "useful
>> >features", for the most part. Things like frontpage extensions are
>> >not comparable to the limitations in MS's telnet.
>>
>> Yes they are. Anything that intentionally locks you into
>> running only MS products or makes you put up with problems
>> when something else is on the other end is all the same.
>
>Is not. :P
Well, if you gave up long ago and no longer make your
own choices...
>[snip]
>> The 'problems' they claim are solved by their non-standard
>> extensions don't really exist until you bring in the
>> MS program that creates them.
>
>Oh, I don't know. I certainly won't take *your* word for it.
Fine, but look at how several other systems do things and
you'll see that the best is rarely Microsoft, although
they may come in cheaper compared to systems that don't
use off-the-shelf generic hardware.
>> The standards bodies aren't
>> dead - if the problems were real they would have been
>> addressed years ago.
>
>You have great faith in standards bodies; I think they are
>very slow to do anything, and are usually way behind the
>curve.
They got telnet right. Microsoft didn't. Telnet is pretty
simple. This is a pattern that repeats over and over...
>> OK then, how about annoying your customers unless they give
>> up using any other vendors products?
>
>I think the people who are annoyed with MS are going to be
>annoyed no matter what- it is MS's success that irritates
>them primarily.
You keep confusing cause and effect. People who are annoyed
with MS are annoyed because MS products do annoying things.
The problematic business practices are secondary.
>[snip]
>> >I assume so. The idea of using POSIX on NT was very silly
>> >to start with; intended, I think, to check a checkbox on somebody
>> >list and no more.
>>
>> It wasn't silly at all, it was a brilliant marketing ploy. Making
>> it unusable was equally brilliant.
>
>Oh, I don't it was in *any* way effective as a marketing tool-
>no-one was fooled into thinking NT was even a passable
>Unix.
I disagree. MS is very good at fooling people - and they did.
>> And evil.
>
>Certainly not evil; you're giving far to much weight to these things.
>It's harmless to provide a POSIX personality, even though it is
>of little practical value. I don't think harmless things should be
>seen as Evil (tm).
I suspect it gave them hundreds of millions of dollars of business
they would not have had otherwise - on a promise of something
that wan't usable. Is that harmless?
>> >MS has adhered to the letter of the (basic) POSIX API, well, you
>> >know what came out.
>> >
>> >POSIX is just the minimum that the Unix vendors could agree
>> >to, and as such it is almost completely useless by itself.
>>
>> That's not true at all. Many programs could be done using
>> only ansi-C standard functions and posix specified routines.
>
>Well, more than a bakers dozen. But POSIX by itself does not
>even give you X-Windows.
X is just a protocol. If you have TCP (in theory any connection
stream) you can run X.
>> The ones that can't can generally have the system/device-specific
>> operations encapsulated in a library so it can be ported to
>> anything with equivalent functionality without touching the
>> bulk of the code.
>
>Sure; but you should realize that POSIX is a small part of the
>"system specific operations" that you need to encapsulate. There's
>typicaly more.
Yes, posix.1 didn't cover everything - much like windows 1.
That is why both have later versions.
>
>> >To make it useful what you need to a Unix clone, pretty much all the
>> >way- and not just POSIX.
>>
>> No, the usability of the posix subsystem is broken in a very
>> specific way. I have enough faith in the MS engineers to
>> think it could have been usable if they had wanted it to be.
>
>What specific way is that?
The main problem is that fork() exists and apparently meets the
letter of the specification, but does not allow the open
file descriptors that are network sockets to be inherited
by the child process. This alone breaks any existing code
that expects to be started from inetd, inheriting the network
connection as its standard files, or any standalone servers
that accept connections in the main process and fork a process
to handle each one. I don't have any experience
with this - I expected an empty promise of interoperability as
usual and avoided even trying. But that is what the people
who tried say.
>Bear in mind, it's *fundamentally* unusable because it's
>just POSIX, and isn't enough for most real tasks done with
>NT. Even if it were a sterling implementation that would still
>be true. POSIX just ain't Unix.
Posix.1 just wasn't enough unix. It has been updated. NT
hasn't.
>> >Fortunately, MS has declined to go that route.
>>
>> There is nothing fortunate about it. They are just manipulating
>> their customers.
>
>Sure; offering them better technology in a brazen attempt to
>separate them from their money.
Not better - just different, non-standard, and non-inoperable.
>Of course, I happen to see this as a good thing.. :D
Until you find some other thing that you would like to
use but can't because it is too much trouble to integrate
with what you have already been sold.
Les Mikesell
[EMAIL PROTECTED]
------------------------------
From: Cihl <[EMAIL PROTECTED]>
Subject: Re: An Example of how not to benchmark
Date: Thu, 15 Jun 2000 20:32:19 GMT
Just for information:
When running one single task at a time, Windows is actually a bit
faster than Linux. Linux takes somewhat more overhead from proper
multitasking, memory management and I/O distribution. Windows can't
quite do that. Instead it can effectively lock up your desktop while
you scan something, or connect to your ISP, or install a new package,
to name just a few examples.
The speed difference really shouldn't be as large as you say, though.
It's still mainly the speed of your CPU/MEM/IO that determines that,
not the OS.
Actually, multimedia/3D performance isn't that hot on Windows, either.
It needs a rather large set of extra tools for bypassing most of the
OS-layers Windows has, just to enable the user to play games on it.
(I'm talking about DirectX, obviously)
------------------------------
From: "Christopher Smith" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.ms-windows.nt.advocacy
Subject: Re: Number of Linux Users
Date: Fri, 16 Jun 2000 06:36:38 +1000
"Michael Born" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> If a product has increasing market share each year (which Linux has
> achieved in the server os market), they are taking over.
Only if the market has a fixed size.
------------------------------
From: [EMAIL PROTECTED] ()
Subject: Re: ApplixWare? More Build It As You Go Along Linux....
Date: 15 Jun 2000 16:41:18 -0400
On Thu, 15 Jun 2000 03:53:32 GMT, billy ball <[EMAIL PROTECTED]> wrote:
>On Wed, 14 Jun 2000 20:30:05 GMT, [EMAIL PROTECTED]
><[EMAIL PROTECTED]> wrote:
>the suite [ApplixWare] is robust, recovers gracefully from any
>problems, and works in low-memory situations (read 32MB systems)...
>
>what's the problem?
No singing and dancing paper clip. <G>
------------------------------
From: Cihl <[EMAIL PROTECTED]>
Crossposted-To: comp.os.ms-windows.nt.advocacy
Subject: Re: Number of Linux Users
Date: Thu, 15 Jun 2000 20:45:45 GMT
Christopher Smith wrote:
>
> "Michael Born" <[EMAIL PROTECTED]> wrote in message
> news:[EMAIL PROTECTED]...
> > If a product has increasing market share each year (which Linux has
> > achieved in the server os market), they are taking over.
>
> Only if the market has a fixed size.
Eh?
Market share == Part of the market
Market share of Linux = Number of Users / Total Users
------------------------------
Date: Thu, 15 Jun 2000 13:47:53 -0700
From: Bob Lyday <[EMAIL PROTECTED]>
Crossposted-To:
comp.os.os2.advocacy,comp.os.ms-windows.nt.advocacy,comp.sys.mac.advocacy
Subject: Re: OT Aboriginal Lifespans was Re: Canada invites Microsoft north
billy ball wrote:
>
> ring-knockers are a bastard plague in the services, and their usefulness
> (to ensure a faithful core of officers in case of military mutiny) is long
> past...
Ring-knockers = ????????????
--
Bob
Never argue with an idiot. They bring you down to their level and then
beat you with experience.
------------------------------
From: [EMAIL PROTECTED] (abraxas)
Crossposted-To: comp.os.ms-windows.nt.advocacy
Subject: Re: Microsoft migrates Hotmail to W2K
Date: 15 Jun 2000 20:53:26 GMT
In comp.os.linux.advocacy Drestin Black <[EMAIL PROTECTED]> wrote:
> again - actually prove something.
> GO ahead, we are all waiting.
> You talk and talk and pose and waste space but have you proven anything?
> Lets see a deja url, how about that microphone? how about that laptop?
Wow, you really should wait until you read all of the posts regarding
a specific topic before you reply, huh?
> have you actually done ANYTHING other than selfmutilation and
> selfhumiliation?
I invented a radon detector which was 61% more accurate in 50% of the
time by utilizing a cylindrical gamma vector detector and a box of
doughnuts as inspiration. I also came up with a way to accellerate
an estherification reaction between n-butyl acetate and and ethyl
alchohol without exothermic results. I know how to make Saurbroten,
I understand The Dubliners, and I know how to tie my own shoes.
And I can change the toner cartridge on any printer.
How bout you?
=====yttrx
------------------------------
From: Eric Remy <[EMAIL PROTECTED]>
Crossposted-To:
comp.os.os2.advocacy,comp.os.ms-windows.nt.advocacy,comp.sys.mac.advocacy
Subject: Re: OT Aboriginal Lifespans was Re: Canada invites Microsoft north
Date: Thu, 15 Jun 2000 17:08:14 -0400
In article <[EMAIL PROTECTED]>, Bob Lyday
<[EMAIL PROTECTED]> wrote:
>billy ball wrote:
>>
>> ring-knockers are a bastard plague in the services, and their usefulness
>> (to ensure a faithful core of officers in case of military mutiny) is
>> long
>> past...
>
>Ring-knockers = ????????????
Military academy graduates. I've always heard it in reference to West
Point grads, but then again I was in the Army. I assume USNA and USAFA
grads are the same in spirit. (Although I've had mixed experiences:
some are normal guys, some are pricks.)
--
Eric Remy. Chemistry Learning Center Director, Virginia Tech
"I don't like (quantum mechanics), | How many errors can
and I'm sorry I ever had anything | you find in my X-Face?
to do with it."- Erwin Schrodinger |
------------------------------
** 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
******************************
