Linux-Advocacy Digest #346, Volume #29 Thu, 28 Sep 00 13:13:03 EDT
Contents:
IBM and Linux ([EMAIL PROTECTED])
Re: Linux? ("Ingemar Lundin")
Re: programming languages and design (Donal K. Fellows)
Re: Is Linux some kind of a joke? ([EMAIL PROTECTED])
Re: Space Shuttle uses Windows software almost exclusively ("Stuart Fox")
Re: IBM and Linux (Roberto Alsina)
Re: Is Linux some kind of a joke? (No Name)
Re: Linux? (Flint Slacker)
Re: Space Shuttle uses Windows software almost exclusively (dc)
Re: Space Shuttle uses Windows software almost exclusively (Jason McNorton)
Re: programming languages and design (The Ghost In The Machine)
Re: Is Linux some kind of a joke? ("sandrews")
----------------------------------------------------------------------------
From: [EMAIL PROTECTED]
Subject: IBM and Linux
Reply-To: [EMAIL PROTECTED]
Date: Thu, 28 Sep 2000 14:42:05 GMT
I had to send my ThinkPad back to IBM for repair the other day. They
include a paper you have to fill out listing the options you are
sending and a description of the problem etc.
It was interesting to see Linux listed as a choice in the operating
systems section, meaning the person sending back the unit has Linux
installed. They wanted access to /var/logs and the root password as
well for problem determination I suppose.
To me, this is a clear sign that Linux is fast becoming a major player
in the operating systems wars. This is a good thing for all of us.
claire
------------------------------
From: "Ingemar Lundin" <[EMAIL PROTECTED]>
Subject: Re: Linux?
Date: Thu, 28 Sep 2000 14:45:14 GMT
try SuSe Linux...you will find that it is next to fucking addictive ;)
you will never want to try another Linux distro after that ;)
/IL
[writing this in windows 98, on the work, what should you do?]
"Coconut Ming" <[EMAIL PROTECTED]> skrev i meddelandet
news:[EMAIL PROTECTED]...
> Which is Linux version or clone that have many user is using now?
> Redhat? Corel? Or original linux?
> Any thing on that?
> Can give some comment on that?
> i need some advice.
> thanks
>
> From
> kokming
>
------------------------------
From: [EMAIL PROTECTED] (Donal K. Fellows)
Subject: Re: programming languages and design
Date: 28 Sep 2000 14:39:19 GMT
In article <8qu5rc$o4o$[EMAIL PROTECTED]>,
Steve Mading <[EMAIL PROTECTED]> wrote:
> The Ghost In The Machine <[EMAIL PROTECTED]> wrote:
>: [5] Multiple vtbl pointers, because of multiple inheritance.
>: Java is slightly cleaner in that respect, and languages such
>: as Smalltalk probably use a single pointer to a table in their
>: message dispatcher, at the head of each class -- their equivalent
>: of a C++ vtbl. (I don't know Smalltalk beyond a very general usage
>: model, though.)
>
> Java is only "cleaner" because it can't do it at all. If you program
> in C++ and don't use multiple inheritence, then you don't get the
> multiple vtbl pointers. When I do multiple-inheritence in C++, I
> generally do it like Java does - all the parent classes except for
> one are merely lists of un-implemented methods, taken on as a promise
> to the compiler that I will implement them in this class.
The serious problems come whenever you have multiple inheritence of
implementations; whereas MI of specifications is fairly easily
understood. This is one of the areas that Java gets right (along with
forcing the object hierarchy to be single rooted and for every object
to be able to report its class, with every class being an object
itself.)
>: [9] Stdarg/varargs invite trouble, but can also be very useful.
>
> I agree - and I wonder - do any other languages, other than
> weakly types scripting languages such as perl, provide for
> varying numbers of arguments?
Only finite variation through mechanisms like overloading AFAIK. This
is an area where scripting languages have a big advantage.
>: [15] Very little introspection is available in C++; one can probably get
>: the type name of a pointer to a class and do such things as
>: dynamic casting, but that's about it.
>
> Agreed. This is a problem. When I needed a set of classes with some
> sort of "what type am I"? check, I had to manually make a virtual
> method "myType()", and then make it part of the base class all
> these classes derived from, then override it in each class derived
> from the base. That way I could take a generic pointer of the base
> type and ask, "what is your type". Really a lot of work. The Java
> version of this same program is much simpler in that regard.
This is really what I noted in respect of [5] above. It is one of the
things that the Java people got *right*. And it has saved my a** a
number of times...
>: [17] C++ doesn't have garbage collection. Granted, this is a
>: minor issue in most programs, but daemons that run 24/7
>: have to be very careful, lest they "leak". (Side note:
>: AmigaOS, at least as of 1.3 or so, did not have the
>: ability to clean up memory allocated with AllocMem(),
>: should a process die precipitiously. The C library worked
>: around this by maintaining a chain of allocated memory
>: blocks internally, and freeing it should a process exit.
>: It worked most of the time....)
>
> Hmm. Yes, that's a problem. In general, when working with a
> language that doesn't have garbage collection, I try to make daemons
> on the fork-off model. The only part of the daemon that runs 24/7
> is the loop that watches for when to "do something" The actual work
> is carried out by a forked process, like what happens in most TCP/IP
> server programs. That limits the area of code that I have to be
> meticulous with to a very small loop.
This is all very well if the forked off processes do not need to
interact. With the interesting cases (i.e. where you have got
interaction between clients via the server) things get awkward.
Frankly, this sort of thing is not easy to code well in any language,
even if the memory handling is looked after for you, due to the number
of different things that can go wrong and the number of places where
the problem can occur...
Donal.
--
Donal K. Fellows http://www.cs.man.ac.uk/~fellowsd/ [EMAIL PROTECTED]
-- This may scare your cat into premature baldness, but Sun are not the only
sellers of Unix. -- Anthony Ord <[EMAIL PROTECTED]>
------------------------------
From: [EMAIL PROTECTED]
Subject: Re: Is Linux some kind of a joke?
Date: Thu, 28 Sep 2000 14:51:19 GMT
In article <[EMAIL PROTECTED]>,
Jacques Guy <[EMAIL PROTECTED]> wrote:
[snippity]
> Tim! Tim! You're back old sport! ***smooooooch***
>
> But, but... what's happened to your wonderfull
> speallign czecher? Oh... pleeze bryngue itt baq,
> oled koqq, prittye pliese, wee thy scream on top?
>
Looks like maybe the court-ordered medication is starting to help.
Sent via Deja.com http://www.deja.com/
Before you buy.
------------------------------
From: "Stuart Fox" <[EMAIL PROTECTED]>
Crossposted-To:
comp.os.ms-windows.nt.advocacy,comp.sys.mac.advocacy,comp.os.os2.advocacy
Subject: Re: Space Shuttle uses Windows software almost exclusively
Date: Thu, 28 Sep 2000 16:09:56 +0100
"Christopher Smith" <[EMAIL PROTECTED]> wrote in message
news:39d3441d$0$26528$[EMAIL PROTECTED]...
>
> > Win95? The default protocol installed at the time the NIC is
> > isntalled is probably NetBEUI....that's not quite the same as the
> > "default protocol" which would vary by box, but anyway...
>
> NetBEUI is installed by default. If you then install TCPIP, which is what
> most people will do, what is the default protocol afterwards ? TCPIP or
> NetBEUI ?
>
> IOW, when you install a new protocol, does it make itself the default ?
No. You have to manually tell it to be the default.
------------------------------
From: Roberto Alsina <[EMAIL PROTECTED]>
Subject: Re: IBM and Linux
Date: Thu, 28 Sep 2000 12:18:55 -0300
El jue, 28 sep 2000, [EMAIL PROTECTED] escribió:
>I had to send my ThinkPad back to IBM for repair the other day. They
>include a paper you have to fill out listing the options you are
>sending and a description of the problem etc.
>
>It was interesting to see Linux listed as a choice in the operating
>systems section, meaning the person sending back the unit has Linux
>installed. They wanted access to /var/logs and the root password as
>well for problem determination I suppose.
>
>To me, this is a clear sign that Linux is fast becoming a major player
>in the operating systems wars. This is a good thing for all of us.
Since they are selling Thinkpads with Linux preinstalled, they better let you
tell them not to wipe it ;-)
--
Roberto Alsina
------------------------------
From: [EMAIL PROTECTED] (No Name)
Subject: Re: Is Linux some kind of a joke?
Date: 28 Sep 2000 15:06:31 GMT
Reply-To: [EMAIL PROTECTED]
On Thu, 28 Sep 2000 05:50:19 GMT, [EMAIL PROTECTED] said:
>Is Linux some kind of a joke or something?
No.
> I mean I instaled Redhat and it looks like shit.
If you like something that looks nice go and buy yourself a nice painting,
not too avant gard, otherwise your brain (or what is left of it) will
be fried.
> No games,
Every distro of Linux comes with games, which are better than what you get
with Windows out of the box. Unless you are a Minesweeper addict, that is.
> no support
>for my video card.
How do you know? Who did you ask?
> No support for my soundcard or any of my USB
>devices...
Their is support for practically any sound card, you have to buy a very
cheap sound driver.
USB you are right, but not for long.
>
>This has to be a joke?
It is not, an OS can do far more many things than run games and host USB
devices, but you would know that if you were computer literate.
>
>Why should I return to the 1980's just to run Linux?
You are just showing your ignorance, even the most anti Linux person, who is
computer literate, (which you are not) knows that to use Linux is not go
back to the 80s.
>
>Linux is a piece of shit....
That could be in your opinion, then don't use it and tutti contenti.
>
>
>The Whore...
>
>
???
------------------------------
From: [EMAIL PROTECTED] (Flint Slacker)
Subject: Re: Linux?
Reply-To: [EMAIL PROTECTED]
Date: Thu, 28 Sep 2000 15:43:13 GMT
Hmmmm.... Suse is good, but Slackware, well it makes my knees weak!
Flint
On Thu, 28 Sep 2000 14:45:14 GMT, "Ingemar Lundin"
<[EMAIL PROTECTED]> wrote:
>try SuSe Linux...you will find that it is next to fucking addictive ;)
>
>you will never want to try another Linux distro after that ;)
>
>/IL
>
>[writing this in windows 98, on the work, what should you do?]
>
>"Coconut Ming" <[EMAIL PROTECTED]> skrev i meddelandet
>news:[EMAIL PROTECTED]...
>> Which is Linux version or clone that have many user is using now?
>> Redhat? Corel? Or original linux?
>> Any thing on that?
>> Can give some comment on that?
>> i need some advice.
>> thanks
>>
>> From
>> kokming
>>
>
------------------------------
From: dc <[EMAIL PROTECTED]>
Crossposted-To:
comp.os.ms-windows.nt.advocacy,comp.sys.mac.advocacy,comp.os.os2.advocacy
Subject: Re: Space Shuttle uses Windows software almost exclusively
Date: Thu, 28 Sep 2000 10:46:48 -0500
On Thu, 28 Sep 2000 21:04:06 +1000, Chris Sherlock
<[EMAIL PROTECTED]> wrote:
>To regional areas which have traditionally had very very poor access
>(mainly through old phone exchanges). I think that they are offering
>some country users the use of satellite links.
>
>ISDN links cost too much. How much do people in the US pay for an ISDN
>link, BTW?
It's a fortune - I think a friend lives "out in the country" where
there's no DSL or cable modems, and he's (well, his employer) is
paying something like $80 a month for it....for all of 12-15K/s speed.
I'm paying $50 per month (and cheaper deals exist) for 150K/s DSL.
ISDN is very, very expensive.
DC
------------------------------
From: Jason McNorton <[EMAIL PROTECTED]>
Crossposted-To:
comp.os.ms-windows.nt.advocacy,comp.sys.mac.advocacy,comp.os.os2.advocacy
Subject: Re: Space Shuttle uses Windows software almost exclusively
Date: Thu, 28 Sep 2000 10:52:54 -0500
In article
<[EMAIL PROTECTED]>,
[EMAIL PROTECTED] says...
> On Thu, 28 Sep 2000 21:04:06 +1000, Chris Sherlock
> <[EMAIL PROTECTED]> wrote:
>
> >To regional areas which have traditionally had very very poor access
> >(mainly through old phone exchanges). I think that they are offering
> >some country users the use of satellite links.
> >
> >ISDN links cost too much. How much do people in the US pay for an ISDN
> >link, BTW?
>
> It's a fortune - I think a friend lives "out in the country" where
> there's no DSL or cable modems, and he's (well, his employer) is
> paying something like $80 a month for it....for all of 12-15K/s speed.
>
> I'm paying $50 per month (and cheaper deals exist) for 150K/s DSL.
>
> ISDN is very, very expensive.
Tell me about it, I pay $80 plus $40 for the ISP.
It's insane, but it's all I can get. Either that or 26.4k dialup, which
is near useless for me.
And I live right down the road from several company headquarters, and in
a densely populated, pretty affluent area. I just don't get it.
------------------------------
From: [EMAIL PROTECTED] (The Ghost In The Machine)
Subject: Re: programming languages and design
Date: Thu, 28 Sep 2000 16:17:28 GMT
In comp.os.linux.advocacy, Steve Mading
<[EMAIL PROTECTED]>
wrote
on 28 Sep 2000 01:11:08 GMT
<8qu5rc$o4o$[EMAIL PROTECTED]>:
>The Ghost In The Machine <[EMAIL PROTECTED]> wrote:
>
>: Flaws in C++, IMO:
>
>: [1] Almost every C program compiles in C++. This means that
>: Obfuscated C contests can still be held, even with C++
>: compilers. (Not clear if the C preprocessor is a feature,
>: a giant bug, or merely a coil of rope just waiting to tangle
>: up near someone's foot :-) ).
>
>Backward compatiability is a flaw? I'm confused. If you are trying
>to imply that this is bad because C *allows* confusing code, I'd
>disagree. It's not the language's job to make code feel pleasant and
>in good form - that's the programmer's responsibility. C *allows* bad
>form, but it also *allows* good form. That's far more desirable to me
>than a system that ONLY allows "good form", since what constitutes
>"good form" is a matter of opinion, and my opinion rarely agrees
>with that of the language designers.
This is a good point, admittedly -- the flaw might merely be that
C++ still allows this sort of abuse. And that might not be a
flaw at all, as you point out.
>
>: [2] Pointers -- more rope to trip over. (Rope can also be used
>: to tie recalcitrant programs together, though, so it's a
>: flaw *and* a feature.)
>
>Actually, what I don't like is the fact that references have no
>syntactic indicator down in the body of the function - they look
>like normal value variables in their syntax.
That they do -- another double-edged sword.
>
>This is not to say that references are bad - they added better
>type checking, but one thing I always liked about vanilla C was
>the visual obviousness of pass-by-value vs pass-by-pointer, which
>forced the programmer to think, "This isn't really the variable
>I'm modifying, I can tell because I can see the de-reference in the
>expression, I'm actually modifying what this thing points at."
>That visual clue goes away with C++ references.
>
>: [3] The C library, with such goodies as printf ("%s", 1234)
>: or printf ("%d", "The rain in Spain falls here").
>: (I think Gnu C++ will warn on both, though.)
>: Also, the ever-dangerous gets() is still around (the GNU linker
>: does issue a warning, though).
>
>I actually like this. The ability to look at a piece of data
>as if it was several different things is handy. No need for
>ord() or chr(), which are silly do-nothing routines anyway,
>just print the char as if it was an int if you want to see its
>ord(). After all, inside the computer, it *is* a little number.
>It only turns into a character when you render it on the screen
>for the user.
I would at least want it to be marked with a reinterpret_cast<...>.
(That's a do-nothing routine too, admittedly.)
>
>: [4] No built-in array bounds checking. None. Nada. Zip. If you want it,
>: you can use STL vectors, with additional overhead -- and even
>: then, some implementations may not bother. Or you can roll
>: your own. (Libraries such as Electric Fence and tools such
>: as Rational Software's Purify may help in finding trouble spots,
>: but they aren't quite as easy to use as Java's built-in checking.)
>
>I have mixed feelings on this one. I like array-bounds checking,
>but I don't like it being forced down my throat with no way to
>circumvent it. Ideally, it should be a compiler-time flag, so
>I can have it when debugging, but turn it off for speed when not
>debugging.
Agreed, that would be useful. Of course, how does one do it
elegantly when the bounds aren't passed?
int ary[5];
someFunc(ary)
...
void someFunc(int * ary)
{
...
}
someFunc has *no* idea how big ary is.
>
>: [5] Multiple vtbl pointers, because of multiple inheritance.
>: Java is slightly cleaner in that respect, and languages such
>: as Smalltalk probably use a single pointer to a table in their
>: message dispatcher, at the head of each class -- their equivalent
>: of a C++ vtbl. (I don't know Smalltalk beyond a very general usage
>: model, though.)
>
>Java is only "cleaner" because it can't do it at all. If you program
>in C++ and don't use multiple inheritence, then you don't get the
>multiple vtbl pointers. When I do multiple-inheritence in C++, I
>generally do it like Java does - all the parent classes except for
>one are merely lists of un-implemented methods, taken on as a promise
>to the compiler that I will implement them in this class.
>
>: [6] cin >> char *, which is just as bad as gets(), and for the
>: exact same reason! (I'd personally recommend using cin >> string&,
>: if one's using STL.)
>
>Agreed. This is the advantage of the fact that C++ didn't throw
>C away. I can still drop back to using the stdio.h functions, thereby
>avoiding the iostream mess of C++, which sucks big time for trying
>to make formatted output. I'd take printf( "%12d %10.2lf\n", i, x );
>any day over the iostream equivilent. (push the module that changes
>the format width - 12, then push i, then push a space, then push
>the module that changes the width - 10, then push the module that
>changes the precision - 2, then push x.)
I'll admit, the '<<' and '>>' constructs are a bit mucky to work with.
I'm not sure I like them either, but they do allow for typechecking.
>:
>: [7] There is apparently some confusion in implementations of STL;
>: some require 'std::' to be prefixed to each and every STL
>: class (e.g., std::vector, std::map). Some do not.
>
>This, agreed, is a problem. C++ took absurdly long to
>standardize on things, which encouraged everyone to go off
>and say "screw the standards committee, I'd like to write a
>compiler and library some time this decade if possible." This
>inevitably led to incompatabilities.
>
>: [8] Exception handling is spotty, especially with signals
>: (SIGSEGV, SIGILL, SIGINT) thrown in. I can see this confusing
>: a number of people.
>
>: [9] Stdarg/varargs invite trouble, but can also be very useful.
>
>I agree - and I wonder - do any other languages, other than
>weakly types scripting languages such as perl, provide for
>varying numbers of arguments?
>
>: [10] The routine __pure_virtual_called simply exits, with a return code.
>: If I'm lucky, the return code will return failure...
>: (This routine is called if a class is incompletely constructed
>: (or has been destroyed!), and a subsequent caller attempts to call
>: a virtual function which has no definition, e.g.,
>: 'virtual void doIt() = 0'.)
>
>: Java throws an exception, and my understanding is that Smalltalk
>: will hand the message to some sort of 'unknown message' handler,
>: even if the message is destined for some other class hierarchy!
>: Mind you, one can go overboard on this...but it does make
>: code evolution a bit easier.
>
>: [11] Subtle issues regarding malloc/new, threading, and signals that
>: occasionally bite program developers. For example, malloc() cannot
>: be used within a signal handler (a holdover from C), and
>: C++ doesn't thread well, athough with work it's not too broken.
>: It's like walking on a tightrope without a net ... be very
>: very careful, especially if one thread needs another's structures.
>
>: [12] Multiple inheritance is problematic. Consider a class C which
>: inherits from A and B; if a header or source file fails to declare C
>: properly, one might get a missoffset pointer! (Fortunately,
>: this isn't a common problem. At least, I don't think it is...)
>: Also, the "diamond diagram" problem, which might surface during
>: code evolution if A and B inherit from a common baseclass D.
>: C now has two D's, vtbl pointers and all. Yuck.
>
>: [13] It is unclear how well C++ supports wide characters (such as
>: Unicode) in literals. (One of the strengths of Windows
>: is its Unicode handling.)
>
>: [14] Constructors MUST declare initializer lists (the bit after the ':')
>: if they want their fields initialized! Otherwise, one gets
>: garbage, even from those classes which have a default constructor.
>: (At least, in some implementations; I haven't checked g++ lately.)
>
>?? You can initialize things there, or down in the body of the
>constructor - I don't understand the complaint. Are you referring
>to initializing the parent classes? In that case, I can't see where
>else it would have made sense to put it other than up in the header
>of the constructor.
A point of pedantry.
class A
{
int b;
int c;
B d;
B e;
A()
: b(0), d()
{
c = 0;
e = B("foobar");
}
};
In the case above, b and d are initialized, while c and e are
assigned. e in particular is never initialized -- a problem if e
is supposed to be an intelligent pointer which frees whatever it's
pointing to.
>
>: [15] Very little introspection is available in C++; one can probably get
>: the type name of a pointer to a class and do such things as
>: dynamic casting, but that's about it.
>
>Agreed. This is a problem. When I needed a set of classes with some
>sort of "what type am I"? check, I had to manually make a virtual
>method "myType()", and then make it part of the base class all
>these classes derived from, then override it in each class derived
>from the base. That way I could take a generic pointer of the base
>type and ask, "what is your type". Really a lot of work. The Java
>version of this same program is much simpler in that regard.
>
>: [16] If one wants to add a non-virtual method to a class, one must have
>: write access to the declaration of that class (i.e., the bit with
>: the 'class Foo { ... }' in it.) Java also has this problem.
>: Smalltalk, as I understand it, does not.
>
>Why is this bad?
Mostly because some people like to be able to add capabilities to
library files. Admittedly, one can work around it by very careful
editing of the file headers (can't change the size of, or relative
positions of, the data members or the vtbl methods), coupled with a
"patch object", which would be supplied by the implementer and
linked into the program.
It's still not that pretty, and it's quite error-prone, from what
I can see; it's probably easier to just declare another class which
encapsulates the first one - if one can.
>
>: Note that adding virtual methods, or data members, to a class,
>: is a different thing entirely. I don't know if Smalltalk
>: supports the notion of adding members to a class from
>: outside its equivalent of a class declaration; in theory,
>: it could if it understands the notion of a default value,
>: so that already-created objects do the right thing in new code
>: referencing the new members. (I simply don't know; I suspect
>: that it does.) Java basically is all virtuals (although
>: it might do some optimizations, and methods declared final
>: might incur some checking and the wrath of either the compiler
>: or the runtime environment, under certain conditions).
>
>: [17] C++ doesn't have garbage collection. Granted, this is a
>: minor issue in most programs, but daemons that run 24/7
>: have to be very careful, lest they "leak". (Side note:
>: AmigaOS, at least as of 1.3 or so, did not have the
>: ability to clean up memory allocated with AllocMem(),
>: should a process die precipitiously. The C library worked
>: around this by maintaining a chain of allocated memory
>: blocks internally, and freeing it should a process exit.
>: It worked most of the time....)
>
>Hmm. Yes, that's a problem. In general, when working with a
>language that doesn't have garbage collection, I try to make
>daemons on the fork-off model. The only part of the daemon
>that runs 24/7 is the loop that watches for when to "do
>something" The actual work is carried out by a forked process,
>like what happens in most TCP/IP server programs. That limits
>the area of code that I have to be meticulous with to a very
>small loop.
Indeed; that's probably the best model for this sort of thing.
>
>: Granted, some of these flaws may simply be the old performance/safety
>: tradeoff; in particular, introspection would probably require a lot
>: of extra overhead for proper support, and exception handling/signal
>: handling may be because of C and Unix compatibility issues.
>: Array limits would be very troublesome to implement in a clean way,
>: and dynamic adding of methods to a class implies that all methods
>: go through some sort of dynamic dispatcher
>
>: But C++ sure isn't perfect. (It's pretty good if one knows what one
>: is doing, but then, so is a blowtorch around a car's gas tank. :-) ) [*]
>
>:>
>:>However, I still believe C++ is a wonderful language for software
>:>development.
>
>: It works reasonably well. Java might be faster to prototype, though,
>: because one doesn't have to worry quite as much about the sillinesses
>: in C++ that get in the way of getting the idea working. (Of course,
>: in Java, one might spend more time looking up something in the API
>: than simply coding the routine, too -- it's very rich; there's also
>: the issue that Java loves to consume memory.)
>
>My problem in Java is often finding ways to do something that
>wasn't in the API - the system doesn't let you go low-level
>enough to implement anything fast at the library-level. For
>example, when I found out how absurdly slow StringBuffers are
>in Java, I wanted to re-implement my own similar class using
>arrays of chars. That was a pain to deal with when I wanted
>to "re-alloc" when the char array got too large. It was hard
>to just to a fast copy to populate the new larger array, because
>of all the checks and trigger-locks that were built-in to Java.
>
--
[EMAIL PROTECTED] -- insert random misquote here
------------------------------
From: "sandrews" <[EMAIL PROTECTED]>
Subject: Re: Is Linux some kind of a joke?
Date: Thu, 28 Sep 2000 12:13:11 -0500
In article <[EMAIL PROTECTED]>,
[EMAIL PROTECTED] wrote:
> Is Linux some kind of a joke or something?
> I mean I instaled Redhat and it looks like shit. No games, no support
> for my video card. No support for my soundcard or any of my USB
> devices...
>
> This has to be a joke?
>
> Why should I return to the 1980's just to run Linux?
>
> Linux is a piece of shit....
>
>
> The Whore...
>
>
[[EMAIL PROTECTED]]$ grep troll /var/log/messages
Sep 28 08:51:22 localhost trolld[668]:high troll-o-meter reading detected
Sep 28 08:51:22 localhost trolld[668]:troll factor: 95.5 %
Sep 28 08:51:22 localhost trolld[668]:whine factor: 97.5 %
Sep 28 08:51:22 localhost trolld[668]:clue factor: 0.0 %
Sep 28 08:51:22 localhost trolld[668]:suspected troll post details follow:
Sep 28 08:51:22 localhost trolld[668]:From: [EMAIL PROTECTED]
Sep 28 08:51:22 localhost trolld[668]:Subject: Is Linux some kind of a joke?
Sep 28 08:51:22 localhost trolld[668]:Date: Thu, 28 Sep 2000 05:50:19 GMT
Sep 28 08:51:22 localhost trolld[668]:Message-ID:
<[EMAIL PROTECTED]>
Sep 28 08:51:22 localhost trolld[668]:running /usr/local/bin/troll-be-gone
Sep 28 08:51:22 localhost troll-be-gone[1396]:redirected 1 message to /dev/null
Sep 28 08:51:22 localhost troll-be-gone[1396]:updating /home/sandrews/News/Score
Sep 28 08:51:22 localhost troll-be-gone[1396]:[*]
Sep 28 08:51:22 localhost troll-be-gone[1396]:Score -9999
Sep 28 08:51:22 localhost troll-be-gone[1396]:From: [EMAIL PROTECTED]
[[EMAIL PROTECTED]]$
------------------------------
** 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
******************************
