Linux-Advocacy Digest #199, Volume #26 Thu, 20 Apr 00 19:13:05 EDT
Contents:
Re: Standard desktop... ("Davorin Mestric")
Re: Standard desktop... (Mike Marion)
Re: Sell Me On Linux (The Cat)
Re: Why Linux on the desktop? ("John W. Stevens")
Re: Become a Windows Registry Expert! (Mike Marion)
Re: Sell Me On Linux (Streamer)
Re: Become a Windows Registry Expert! (Charles Kooy)
Re: Become a Windows Registry Expert! (George Graves)
Re: Why Linux on the desktop? [OT] ("John W. Stevens")
Re: Unix is dead? ("John W. Stevens")
Re: Become a Windows Registry Expert! (Marty)
Re: Unix is dead? ("John W. Stevens")
Re: at the risk of ignorance...a little too late for that (Bart Oldeman)
Re: Guess How Many Windows Crashes.... (Gary Hallock)
Re: at the risk of ignorance...a little too late for that (David Steinberg)
----------------------------------------------------------------------------
From: "Davorin Mestric" <[EMAIL PROTECTED]>
Subject: Re: Standard desktop...
Date: Fri, 21 Apr 2000 01:01:08 +0200
yes, internet explorer is an user interface. you can write applications for
it. you can produce complete programs in it without touching any of the
native gdi controls. it has an api, it is very extensible and configurable.
it can also be a reporting tool and a shell (ie4). yes, internet explorer
is the next generation user interface for windows.
what qualities in an user interface are you looking for that are not
available in internet explorer? what is your definition of an user
interface?
davorin
<[EMAIL PROTECTED]> wrote in message news:8dnoem$972$[EMAIL PROTECTED]...
> Are you claiming that Explorer is a USER INTERFACE????? < Snicker >
> The only OS's that Explorer (AN APPLICATION) are ported to are OS's that
> MS ports them to. The MS Windows (THE USER INTERFACE) is NOT ported to
> any other platforms NOT EVEN TO OTHER MS OS'es. KDE (A USER INTERFACE)
> can be ported to ANY OS by ANYONE with enough skills. The point is, KDE
> (USER INTERFACE) can move faster and farther than can the MS desktop
> (USER INTERFACE). Explorer is NOT A DESKTOP (USER INTERFACE)!
>
------------------------------
From: Mike Marion <[EMAIL PROTECTED]>
Subject: Re: Standard desktop...
Date: Thu, 20 Apr 2000 22:12:49 GMT
mlw wrote:
> You can run IE on three platforms. Windows, Macintosh, and SUN because
And speaking from experience (mine and many others here) you often can't even
run IE5 under Solaris 2.6. On my box it either locks itself (or my entire wm)
about 3 seconds after I start it.. I have to login to the box from another to
kill it.
Under CDE on 2.6 it randomly locks up, or it crashes CDE for several others that
tried it.
--
Mike Marion - Unix SysAdmin/Engineer, Qualcomm Inc.
"There are no significant bugs in our released software that any
significant number of users want fixed." - Bill Gates in an interview with
Focus magazine, Oct 23, 1995.
------------------------------
From: The Cat <[EMAIL PROTECTED]>
Subject: Re: Sell Me On Linux
Date: Thu, 20 Apr 2000 22:15:24 GMT
For $1.99 at cheapbytes.com you can discover for yourself. I suggest
mandrake 7.0 as a good start.
thecat
Hint:rebooting every time you change a network setting is
optional.....
On Thu, 20 Apr 2000 19:36:15 GMT, Jason Portell <[EMAIL PROTECTED]>
wrote:
> I'm a corporate Network administrator and we are currently exploring
>different networking solutions in my company. I don't know very much
>about the OS, and I would like to know what is SOOOOO great about Linux.
>
>Thanks,
>
>Jason Portell
>NWE Communications
>St. Louis, MO
"Agent under Wine and powered by Mandrake 7.0"
------------------------------
From: "John W. Stevens" <[EMAIL PROTECTED]>
Subject: Re: Why Linux on the desktop?
Date: Thu, 20 Apr 2000 16:17:30 -0600
[EMAIL PROTECTED] wrote:
>
> In article <[EMAIL PROTECTED]>,
> "John W. Stevens" <[EMAIL PROTECTED]> wrote:
>
> > Even taking into account all of the following factors you describe,
> > a list is displayed as a list, a table as a table.
>
> What if the table has two columns, one of bullets and one of text data,
> with bordersize set to zero?
A list is still displayed as a list, a table is still displayed as a
table.
In this case, they are *VISUALLY* indistinguishable, but they are still
displayed different.
> > The simple fact is that a table is not a list, and the browser will
> > display them differently depending on the tag (The "instruction"
> > given).
>
> Not necessarily. See above
And you were focusing on "visually different". But my statement was
that the browser would display them differently, which *usually* means
that they will be visually different, but not always.
The point, however, is that the actions the *browser* takes to display
them are different.
> > > That is an extremely loose (not "abstract")
> > > definition: program use == programming
> >
> > Nope. 'Cause that isn't my description of programming.
> >
> > Programming is design, followed by communication, followed by
> > execution.
> >
> > Running a program is simply part of programming: it is the execution
> > of one step in a program.
> >
> > The program might be:
> >
> > Program: Find and Delete the Core file.
> >
> > 1) Change directory to binaries.
> >
> > 2) Run ls to determine contents of directory.
> >
> > 3) Search for a file named core.
> >
> > 4) IF found:
> > delete file core.
> >
> > Running ls is not in itself programming . . . but running a prograrm
> > is *part* of programming.
>
> So then, executing a program (e.g., deleting core) is programming . . .
Nope. You read all that, and got this out of it?
I really need to work on my writing skills . . .
> > Remember, the assertion wasn't about HTML . . . the assertion was that
> > all interaction with a computer was programming.
>
> Provided we have a loose, or abstract, enough of a definition of
> programming; "Programming is interacting with a computer, ergo, 'all
> interaction with a computer [is] programming.'"
Your response is a straw man argument.
That isn't my definition of programming.
Programming: The act of creating a set of instructions, to be executed
by a computing system, for the purpose of fulfilling a specific goal.
> If we say that booking a flight reservation (regardless of method or
> means) is "programming" then, obviously, booking said flight on
> Travelocity or Yahoo Travel qualifies as "programming."
If you created a set of instructions, that were then executed, to
fulfill that specific purpose, you did indeed engage in programming.
> > Then why respond? With no clear definitions, how can you categorize
> > anything?
>
> I respond precisely because I didn't consider the submitted definition
> to be a "clear definition."
Then please supply one. We've taken a hack at mine . . . what's yours?
>
> > > I simply take issue with your definitions as
> > > I consider them to be so overly broad as to be meaningless.
> >
> > No, you don't consider them to be overly broad . . . since you have no
> > definitions of your own to compare them to, you *cannot* consider them
> > to be to broad, to narrow, or anything else, as those terms require
> > you to be *COMPARING* my definition to some other definition . . . an
> > impossibility if you have no other definitions.
>
> Where did I claim to have no other definitions? No, I said I am trying
> to avoid ***MAKING*** definitions.
Having definitions, yet refusing to supply them, I am left with only one
valid conclusion.
Supply your defintions, and you've got a discussion. Simply saying that
you don't like mine, isn't very meaningful.
> I have an imprecise, gut-level defintion of programming,
IOW, you don't have a definition.
What you have is an emotional reaction.
> as it's what I
> do for a living, and based on that imprecise definition, I *CAN* and
> *DO* consider your definition to be overly broad.
Emotional reactions do not constitute definitions, due to the many to
one problem.
> > > Or, if an HTML *document* is a "program"
> >
> > Hint: all programs are documents.
>
> Again, you assume a two-way mapping that may not necessarily hold.
Not at all. I never said that all documents are programs.
I specifically stated the conditions under which the above would be
true, though.
> > One of the purposes of this discussion is to abstract enough to begin
> > the process of design.
>
> What are you trying to design?
A next-gen user interface.
> > No. Any instruction stream directed at any computational system is a
> > program.
>
> Even if there is no design behind the instruction stream?
Yes.
Nobody wrote the program, but it is still a program.
> Doesn't what you just said conflict with the following?
>
> JWS> Programming is design, followed by communication, followed by
> JWS> execution.
Nope. I never at any point required that all programs be created by
programming, did I?
> It depends on how we define "programming," doesn't it? ;) This, of
> course, sends us onto a slippery slope of circular logic.
No, it leads us back to the question: How do you define programming?
--
If I spoke for HP --- there probably wouldn't BE an HP!
John Stevens
[EMAIL PROTECTED]
------------------------------
From: Mike Marion <[EMAIL PROTECTED]>
Crossposted-To:
comp.sys.mac.advocacy,comp.os.os2.advocacy,comp.os.ms-windows.nt.advocacy
Subject: Re: Become a Windows Registry Expert!
Date: Thu, 20 Apr 2000 22:30:23 GMT
Bob Lyday wrote:
> What do you run on your PC's?
OS? Mostly NT.
Apps? Eudora (of course), and the usual productivity apps.
--
Mike Marion - Unix SysAdmin/Engineer, Qualcomm Inc.
gawk; talk; date; wine; grep; touch; unzip; touch; gasp; finger; gasp; mount;\
fsck; more; yes; gasp; umount; make clean; make mrproper; sleep
------------------------------
From: Streamer <[EMAIL PROTECTED]>
Subject: Re: Sell Me On Linux
Date: Thu, 20 Apr 2000 17:35:18 -0500
Jason Portell wrote:
> I'm a corporate Network administrator and we are currently exploring
> different networking solutions in my company. I don't know very much
> about the OS, and I would like to know what is SOOOOO great about Linux.
Had you asked politely, I might have referred you to some links. But the
"SOOOOO great" in your post already shows that you aren't about to give
Linux a chance. So please don't insult our intelligence by asking for
advice on something you have absolutely no intention of giving a chance.
Next time, use something else in your 'exploration' that you're just using
to justify the real way you want your 'exploration' to go.
Too bad. Linux might have provided a good solution for you....but I know
you already have your mind made up. Too bad your company will suffer as
the result.
------------------------------
From: [EMAIL PROTECTED] (Charles Kooy)
Crossposted-To:
comp.sys.mac.advocacy,comp.os.os2.advocacy,comp.os.ms-windows.nt.advocacy
Subject: Re: Become a Windows Registry Expert!
Date: Thu, 20 Apr 2000 23:42:47 +0100
Shock Boy <[EMAIL PROTECTED]> wrote:
> "Eric Bennett" <[EMAIL PROTECTED]> wrote in message
> news:[EMAIL PROTECTED]...
> > I should point out that multiple instances wouldn't help here on the
> > Mac, because most Mac printer drivers won't let you switch to *anything*
> > else, whether another app or another instance of the same app, while
> > they are writing the spool file.
>
> Yes, no matter how well an application is written, it can not overcome
> sloppy programming in the operating system.
>
> Hell, I remember that I'ld start the print job.. then go out for lunch..
> and hopefully when I get back, I could actually use my mac!
The only apps I've ever seen do this are Quark and Illustrator.
Presumably you're talking about those. Or are there other apps which do
it of which I'm not aware?
You did have background printing turned on?
ck
------------------------------
From: George Graves <[EMAIL PROTECTED]>
Crossposted-To:
comp.sys.mac.advocacy,comp.os.os2.advocacy,comp.os.ms-windows.nt.advocacy
Subject: Re: Become a Windows Registry Expert!
Date: Thu, 20 Apr 2000 22:41:23 GMT
In article <[EMAIL PROTECTED]>, Marty <[EMAIL PROTECTED]>
wrote:
>George Graves wrote:
>>
>> In article <[EMAIL PROTECTED]>, Marty <[EMAIL PROTECTED]>
>> wrote:
>>
>> >George Graves wrote:
>> >>
>> >> In article <[EMAIL PROTECTED]>, Marty
>> >> <[EMAIL PROTECTED]>
>> >> wrote:
>> >>
>> >> >George Graves wrote:
>> >> >>
>> >> >> Don't worry, I won't. I have learned that the only thing that
>> >> >> Apple
>> >> >> could ever do to please Wintrolls who post on CSMA is to roll
>> >> >> over,
>> >> >> belly-up and die. With Apple gone, they wouldn't have that little
>> >> >> nagging voice in their head that keeps saying "did I choose the
>> >> >> wrong
>> >> >> platform?" Because with no Apple, there would be only ONE platform
>> >> >> and
>> >> >> the Wintrolls could sleep secure in their beds with no nasty Apple
>> >> >> confusing them with that pesky Macintosh.
>> >> >
>> >> >A common misconception. PC owners are becoming increasingly aware
>> >> >that
>> >> >there are alternatives to MS based products, thus there are far for
>> >> >than
>> >> >"one" platform available.
>> >>
>> >> With what, pray tell, to run on them?
>> >
>> >It's called "software" I think.
>>
>> There just isn't enough of it for most people to get any work done.
>
>Let's leave that up to "most people" to decide for themselves.
They have. "most people" chose Windows, the rest chose Mac.
--
George Graves
------------------------------
From: "John W. Stevens" <[EMAIL PROTECTED]>
Subject: Re: Why Linux on the desktop? [OT]
Date: Thu, 20 Apr 2000 16:39:07 -0600
Matthias Warkus wrote:
>
> It was the Tue, 11 Apr 2000 13:28:00 -0600...
> ...and John W. Stevens <[EMAIL PROTECTED]> wrote:
> > IOW, a monitor is *PART* of a computing system, yes?
>
> Depends on your definitions. Certainly, a monitor is not involved with
> programs in any way.
Sure it is.
Here's where the monitor is involved in a program:
1) While not bright enough:
2) Increase brightness of monitor.
There you go.
> Algorithms execute whether you can see their
> output or not. Yes, even interactive ones do. For the purpose of this
> discussion, it's completely irrelevant whether the input to a program
> is a meaningful user reaction to the output or just something random.
Nope. For this discussion is based on the assertion I made, which
specifies *interaction*, not random input.
As such, if you specified that no planning or design were occurring *AT*
*ALL*, and that you still had *interaction*, then I'd wonder how you
define "interaction" . . . as usually an interaction is when two systems
are communicating, and reacting to the data that has been sent to them.
> After all, we are not talking about the inner workings of the user's
> brain, but about those of the computer's innards.
Nope. We are talking about the inner workings of the user, *AND* the
inner workings of the computer.
Where, in your opinion, does programming happen? If no user is
required, for programming to be taking place, then where *does* it take
place?
> > > > (a) Is using the slider on the GUI to the volume control program
> > > > programming?
> > >
> > > No.
> >
> > Why not?
>
> It does not involve implementing an algorithm in a programming
> language.
Sure it does. See above, though you'll have to replace a few nouns.
> Is turning up the volume on your stereo programming?
No. It is one step in a program.
Unless you randomly twitch your stereo volume knob in random directions
at random times, you did some programming while interacting with your
stereo.
Algorithm:
1) Loop until volume is loud enough to kill small animals:
2) Increase volume.
3) Subject a mouse to stereo output.
4) Did the mouse die?
5) Yes. Break out of loop.
;->
'Course, being a politically correct person who is a friend to all
animals, I know you test the stereo volume on yourself . . .
;_>
> > And, so, once again . . . why is the exact same operation (that of
> > adjusting the monitor) not-programming in one place, but in another, it
> > "may be" programmin?
>
> Programming means creating an implementation of an algorithm that can
> be executed by the computing machinery.
Bingo! Therefore, adjusting the monitor (interacting with the
computer), is one step taken in executing an algorithm.
If you define "interaction" as *ONLY* the knob twist, then you've shot a
hole in my assertion. But I assert that you wouldn't be twitching that
knob without having first done some programming, that in fact, you
cannot reasonably separate the act of twitching that knob and
programming.
> Spoon-feeding events into the machinery is not programming.
Nope. It is, however, part of *interacting* with the machine. It is
not the sum and total of interacting with the machine, but it is part of
interacting with the machine.
One could reasonably conclude that "programming" does not include
executing the program, I suppose, which would pretty well invalidate my
assertion, but programming must take place before you execute any
program, and the events you are feeding into the machine are, in short,
the program that you wrote.
> In fact,
> if you use a GUI, you usually execute an algorithm in your own head,
Umm. . . no. You write that algorithm in your head, and *PART* of that
algorithm is executed in your head, part is executed by the computer.
> and the output goes to the GUI in form of mouse and keyboard events.
> You hold the place of an interpreter, not a programmer, when using a
> GUI.
No, when using a GUI you are *PART* of the system. Think of a two CPU
computer, where one process
is running on each CPU, and they are communicating data back and forth .
. . but both are running *PART* of a program.
> > Really? How so? The computer is spitting out less light than it did
> > before . . . so therefore it is running differently, yes?
>
> Irrelevant. It does not change the state of the machine in any way.
Sure it does. The state of the machine has changed . . . the average
amount of light (the state) emitted by the monitor has changed.
> The state of a computer can be reconstructed from a number of values;
> they do not include the brightness setting of the monitor.
Yes, they do, or you have not reconstructed the computer, you've only
reconstructed part of it.
> > But my point is: how do the processes differ? A program uses variables,
> > into which you put specific values during different runs, but isn't that
> > precisely how you "use" a computer: you learn the pattern, then stick
> > different values into the pattern every time you "run" the pattern?
>
> Letting a computer interpret an algorithm is different from
> interpreting it yourself.
How so? Same algorithm, same inputs, no random factors . . . then you
should get the same output.
> If it weren't, there would be no point in
> using computers for any purpose at all.
Oh.
Then I see where you went astray: neither you, nor the computer, is a
whole. The complete system includes *BOTH* the user and the computer.
What is occurring here is a division of labor. Both you and the
computer are executing part of the algorithm . . . however, you usually
are the only part of the system to execute the "create algorithms" part
of the algorithm . . .
> I've got a definition. You've got one, too. Yours is childish and
> ridiculous, however.
Ohhhh!! Personal attacks!
These really support your point of view . . .
> "Anything you do to a computer is programming."
>
> Bother. "Anything you do to a woman is sex."
>
> Absurd.
You could at least have saved yourself the emberassment of *BOTH*
misstating my assertion, *AND* creating a false analogy.
--
If I spoke for HP --- there probably wouldn't BE an HP!
John Stevens
[EMAIL PROTECTED]
------------------------------
From: "John W. Stevens" <[EMAIL PROTECTED]>
Subject: Re: Unix is dead?
Date: Thu, 20 Apr 2000 16:42:44 -0600
mlw wrote:
>
> Chris Williams wrote:
> >
> > I keep hearing "Unix is dead or will die soon."
> > What can replace it?
> > Linux?
> > Linux is Unix.
> >
> > Chris Williams
>
> Linux is not UNIX. One could say there is no such thing as UNIX anymore.
Oh, maybe, maybe not. I'd say that Unix is now a class designator, so
the proper statement would be: Linux ISA Unix.
--
If I spoke for HP --- there probably wouldn't BE an HP!
John Stevens
[EMAIL PROTECTED]
------------------------------
From: Marty <[EMAIL PROTECTED]>
Crossposted-To:
comp.sys.mac.advocacy,comp.os.os2.advocacy,comp.os.ms-windows.nt.advocacy
Subject: Re: Become a Windows Registry Expert!
Date: Thu, 20 Apr 2000 22:50:54 GMT
George Graves wrote:
>
> In article <[EMAIL PROTECTED]>, Marty <[EMAIL PROTECTED]>
> wrote:
>
> >George Graves wrote:
> >>
> >> In article <[EMAIL PROTECTED]>, Marty <[EMAIL PROTECTED]>
> >> wrote:
> >>
> >> >George Graves wrote:
> >> >>
> >> >> In article <[EMAIL PROTECTED]>, Marty
> >> >> <[EMAIL PROTECTED]>
> >> >> wrote:
> >> >>
> >> >> >George Graves wrote:
> >> >> >>
> >> >> >> Don't worry, I won't. I have learned that the only thing that
> >> >> >> Apple
> >> >> >> could ever do to please Wintrolls who post on CSMA is to roll
> >> >> >> over,
> >> >> >> belly-up and die. With Apple gone, they wouldn't have that little
> >> >> >> nagging voice in their head that keeps saying "did I choose the
> >> >> >> wrong
> >> >> >> platform?" Because with no Apple, there would be only ONE platform
> >> >> >> and
> >> >> >> the Wintrolls could sleep secure in their beds with no nasty Apple
> >> >> >> confusing them with that pesky Macintosh.
> >> >> >
> >> >> >A common misconception. PC owners are becoming increasingly aware
> >> >> >that
> >> >> >there are alternatives to MS based products, thus there are far for
> >> >> >than
> >> >> >"one" platform available.
> >> >>
> >> >> With what, pray tell, to run on them?
> >> >
> >> >It's called "software" I think.
> >>
> >> There just isn't enough of it for most people to get any work done.
> >
> >Let's leave that up to "most people" to decide for themselves.
>
> They have. "most people" chose Windows, the rest chose Mac.
You've just called everyone using OS/2 and Linux and every other PC OS a
"nobody". If this is your intention then we will file your opinion
appropriately (that's what the "Shredder" on the OS/2 desktop is for, after
all).
------------------------------
From: "John W. Stevens" <[EMAIL PROTECTED]>
Subject: Re: Unix is dead?
Date: Thu, 20 Apr 2000 16:45:59 -0600
JoeX1029 wrote:
>
> actually i think i heard somewhere that the Linux kernel was based on the Minix
> kernel. Unsure if it really was thats just what i heard. Hopefully unix won't
> die:)
Nope. Linus was inspired by Minix, but did not "base" Linux on Minix
any more than Windows NT is based on Minix (IE, all modern OS'en share
some pretty general architechtural concepts . . .)
--
If I spoke for HP --- there probably wouldn't BE an HP!
John Stevens
[EMAIL PROTECTED]
------------------------------
Crossposted-To: rec.games.roguelike.nethack
From: Bart Oldeman <[EMAIL PROTECTED]>
Subject: Re: at the risk of ignorance...a little too late for that
Reply-To: [EMAIL PROTECTED]
Date: Thu, 20 Apr 2000 22:47:28 GMT
On Thu, 20 Apr 2000, Karl Knechtel wrote:
> <rant crossposted to comp.os.linux.advocacy, where it should hopefully be
> somewhat more relevant. Those of you in rec.games.roguelike.nethack are
> warned ;)>
>
> : (Typing 'nethack' while in the directory probably doesn't work because
> : '.' ("current directory") isn't in your path either, for good though
> : somewhat offtopic reason; invoking './nethack' would likely work,
> : but adding '/usr/games' to your path will save the directory-changing.)
>
> The whole idea of something being "in your path" is easily one of the three
> least intuitive things about *n?x I've run into (the others being the vi
Paths are there in Windows as well.
(e.g. type set in a command line box:
PATH=C:\WINDOWS\COMMAND;C:\WINDOWS;C:\DOS;C:\BAT;C:\UTIL)
You just don't start Word with
C:\>\MSOffice\Winword\winword.exe
do you? In *nix all binaries (or representations) are in a few
directories: you don't need a long path. To implement such a thing in
Windows you need a directory full of batch files or a very long path.
To stay on-topic: I just tried to install nethack on my Debian
GNU/Linux system:
apt-get install nethack
That's all, and an entry (together with Xnethack) shows up in the Debian
Application Menus, under Games -> Adventures. Nothing special.
> keys - which I avoid by using pico - and the idea of forcing something to
pico is fine for small editing jobs, but it lacks a lot of things (like
search & replace and syntax highlighting, although the former seems to
exist now if you start with "pico -b"); it's useless for serious editing
of programs and large texts.
Emacs (probably needs some customization) or Nedit or mcedit or ...
should be fine for you as well.
> run in the background with. I thought this class of OSes was supposed to
> do PMT; how come if I run netscape without an & from an xterm, the commands
> in the xterm window don't get executed until I quit Netscape, even if I
> minimize it and bring the xterm window into focus? Shouldn't a command
Just make the xterm active, type Ctrl-Z (stops netscape) and then bg
<enter> (puts it in the background) if you want to use that xterm for
commands again. It's not a big deal.
The reason why it goes this way, is that your shell (e.g. bash) doesn't
know whether netscape is going print some stuff in the xterm. And if you
don't like this, why don't you just make an alias?
alias netscape = 'netscape &'
or make an icon on your desktop, or start it from the menu?
I really don't see your point very much. You choose to use the shell. This
gives you a lot of flexibility, but a few minor 'inconveniences', which
are quite logical if you think about it.
Bart
------------------------------
Date: Thu, 20 Apr 2000 19:24:28 -0400
From: Gary Hallock <[EMAIL PROTECTED]>
Subject: Re: Guess How Many Windows Crashes....
"Dan J. Smeski" wrote:
> No, just common sense.
>
In other words, ALL crashes are caused by bad hardware. Windows is perfect.
If Windows crashes it must be the hardware so you would just go replace all of
the hardware hoping that the problem disappears. I'm glad you don't work for
my company.
Gary
------------------------------
From: [EMAIL PROTECTED] (David Steinberg)
Crossposted-To: rec.games.roguelike.nethack
Subject: Re: at the risk of ignorance...a little too late for that
Date: 20 Apr 2000 23:01:25 GMT
Karl Knechtel ([EMAIL PROTECTED]) wrote:
: The whole idea of something being "in your path" is easily one of the
: three least intuitive things about *n?x I've run into...
Funny, I didn't have a problem with it since it is so similar to the DOS
concept. The only thing to get used to, already mentioned, is that it's
not hardcoded anywhere to try the current directory before the path. If
you want to try the current directory as part of the path, you add a . to
the path. This increases your fexibility: you can chose not to try the
current directory, to try it before the rest of the path (PATH=".:$PATH"),
to try it after the rest of the path (PATH="$PATH:."), or to try it at any
point in between.
Note that not having the current directory as a special case is actually
MORE consistant (hence simpler) than the DOS way. A directory is either
in the path or it's not -- no special cases. You're likely just used to
the DOS way...it's not always easy to change our assumptions.
: Why on EARTH saying the equivalent of "go to the current directory and
: open this file" should under any circumstances allow you to do
: something that saying "open this file" wouldn't is beyond me. Off-topic
: (to RGRN), perhaps, but please do enlighten us (me). I presume there
: must be some obscure security reason for it.
There is a simple security reason for it.
Say some nasty person manages to put a shell script like this
#!/bin/sh
rm -rf *
in your home directory, calls it "ls," and makes it executible. The first
time you type "ls" in your home directory, you lose all your personal
data. Oops!
You lose your data, that is, if you have . in your path before
/bin. If not, /bin/ls is executed, and you get a listing of your
files. Of course, noticing that there's an executable named "ls", you
investigate further and save the day! :)
Note that the default, and ONLY, behaviour in DOS/Windows is to try the
current directory before the path, and to execute that nasty, destructive
trojan. In UNIX, you can protect yourself from this situation by making
sure that if you have "." in your path, it is the last directory listed
therein.
Remember: it's UNIX, it all makes sense... :)
--
David Steinberg -o) Boycott Amazon.com! Fight
Computer Engineering Undergrad, UBC / \ the "1-Click Order" patent:
[EMAIL PROTECTED] _\_v http://www.nowebpatents.org
------------------------------
** 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
******************************
