Linux-Advocacy Digest #149, Volume #29 Sat, 16 Sep 00 23:13:04 EDT
Contents:
Re: Another "feature" in IE discovered. (D. Spider)
Re: Because programmers hate users (Re: Why are Linux UIs so crappy?) ("Tristan
Wibberley")
Re: Windows+Linux=True (D. Spider)
Re: ZDNet reviews W2K server; I think you'll be surprised.... (T. Max Devlin)
Re: ZDNet reviews W2K server; I think you'll be surprised.... (T. Max Devlin)
Re: Unix more secure, huh? ("Otto")
Re: ZDNet reviews W2K server; I think you'll be surprised.... (T. Max Devlin)
Never tell me again that Windows is easy to install!!! It's a lie! (OSguy)
Re: Because programmers hate users (Re: Why are Linux UIs so crappy?) (Richard)
Re: Unix rules in Redmond ("Otto")
----------------------------------------------------------------------------
From: [EMAIL PROTECTED] (D. Spider)
Subject: Re: Another "feature" in IE discovered.
Date: Sun, 17 Sep 2000 02:11:58 GMT
It appears that on Sat, 16 Sep 2000 08:31:10 -0400, in
comp.os.linux.advocacy "MH" <[EMAIL PROTECTED]> wrote:
>I'd gladly trade my movements being tracked for a browser that can:
>
>Deal with text entry forms properly. Some work, many don't. Basic
>functionality. Don't have it? Your browser is worst than second rate.
What browser doesn't have it?
>
>Allow copy operations from any part of the page, and paste into any of my
>text editors
>I can't copy text off a page in NN and paste it into any of KDE's editors.
>This blows.
*ROFL* that's funny, I can do that fine.
>
>Fully functioning hot keys for all important operations.
>I'm not a mouser. NN makes me have to be one for may operations.
Hrmm I'm a big hotkey user myself. One of many reasons Opera is the
best browser around IMOP. NN does as good a job at this as IE, neither
is really good at it.
>I.E., for all of the faults it does have, handles all of the above without
>flaw.
No better than NN, and much worse than Opera.
>If Linux needs anything, it's a decent web browser. I'd be logged into Linux
>*MUCH* longer if I wasn't stuck with second and third rate web browsing. I
>really hope Mozilla is the solution, but I'm losing hope with that fast.
Mozilla+Galleon looks very promising, and I hear the new Konquerer is
very nice as well. Opera for linux can't come out a moment too soon
though.
>Besides, there are easy ways around the 'tracking' issue, as with most MS
>induced 'features'.
Perhaps, but I don't appreciate a company that tries to slip crap like
that by me, and I don't want to waste my time constantly trying to
figure out what new trojan-horse "feature" they've decided to saddle
me with.
#####################################################
My email address is posted for purposes of private
correspondence only. Consent is expressly NOT given
to receive advertisements, or bulk mailings of any
kind.
#####################################################
------------------------------
From: "Tristan Wibberley" <[EMAIL PROTECTED]>
Subject: Re: Because programmers hate users (Re: Why are Linux UIs so crappy?)
Date: Sun, 17 Sep 2000 03:29:31 +0100
Richard <[EMAIL PROTECTED]> wrote in message
<[EMAIL PROTECTED]>...
>[EMAIL PROTECTED] wrote:
>> On Fri, 15 Sep 2000 21:58:23 GMT, Richard <[EMAIL PROTECTED]> wrote:
>> >Incorrect. Unix types files into arbitrary "executable" and
"non-executable"
>> >categories. This is much worse than Windows associating only a single
process
>> >with each type; Unix only ever associates one process for *any* file.
>>
>> At the level of the kernel, that's all that's really required.
>
>And we all know that "the level of the kernel" is all that matters,
>because that's how all users and applications interact with the system,
>right? Get a clue man, who gives a shit about the kernel?
You were describing what the filesystem provides. And it provides
information for the kernel in the data found in the inodes (the executable
bit that you described), and information for user-space in the data refered
to by the inodes. Any scheme can be implemented on top of this. It's silly
to say that the existence of programs which do not work how you want them to
implies that you should not use computers upon which they can run - the
shareware available for MS Windows (for example) is mostly crud, so by the
rule you gave, you should not use MS Windows. A better assertion would be:
forall x such that x is a task for which I need a computer, forall y where y
is an operating system for a computer which is available to me : "The set of
combinations of software which can be usefully applied to x does not include
any combination of software for y" -> "I should not use y for x"
>> Furthermore, this is an extended attribute rather than just an
>> element of the filename.
>
>Which makes it much, much worse. Implementing useful functionality in
>conventions (like object names) is a kickass strategy whenever you can
>get away with it. Smalltalk does this again and again because it's simply
>the right way to go.
Yeah, when you can get away with it. I have a file *named*
'slutdoesitbackwards.jpeg' but it is a text file containing my email
address. The correct approach to use is the mathematical approach of
narrowing down the set of software which is not known to be unable to help
the user to use the file in the way that they would like. There are many
algorithms, and most discard methods of choosing the program right from the
start; AFAIK, the *nix command "file" throws away the ability to have a
mime-type stored in the filesystem in some way (but that's because UNIX
doesn't define a prefered way to do that, nor does it reserve filesystem
data or interface for doing this ATM), but keeps all other possibilities. I
think this will change for the better within the next year or two, at least
for Linux.
>> So what Unix has is actually a superset
>> of what Windows can provide.
>
>And yet *somehow*, it has less functionality (in the very narrow context
>we are discussing). It has a superset of /restrictions/, not functionality!
>(Having to use an arbitrary function to use functionality that does not
>logically require it is a constraint.)
But what does this sentence mean?
>> Why should a user process resemble a server process?
>
>Because there is no conceptual difference between the two??
>Because it DOES??
Don't try to get conceptual in an advocacy newsgroup like this one,
conceptual is for academia - it does not apply well to engineering problems
(which choosing an OS is).
>> What is
>> really the point of keeping some file handler forever resident.
>
>resident where? In memory??? That's only the broken Unix paradigm!
Computers only have memory for keeping information. Even if it is in a code
segment - you can keep it in logic space if you want, but I don't think that
helps anyone.
>> >Take a text editor for example. Conceptually, it's a type of windowed
>> >server that takes filenames as input and generates file changes as
output.
>> >There is *no* reason why a different "text editor" server process needs
>> >to be started for every single user, let alone for every single file.
The
>> >whole Unix paradigm is fucked up from top to bottom.
>>
>> No, it's not unecessarily complicated.
>
>For whom, the users or the coders? I do believe you're talking about the
>latter group. Well, boo hoo hoo, here's a free clue: I don't give a shit
>about how hard it is for the coders.
Great argument. Includes expletives and all.
>If programmers were held to the standards of *any* engineering discipline,
>mass executions would swiftly follow. We've all heard this umteenth times
>but enlightenment is obviously a different thing from mere comprehension.
They are held to those standards *mostly*, that is why software doesn't work
all the time - programmers get told to stop by the users because their
software is "good enough". Programmers are not generally held to the
standards of any *mathematical* discipline except where it is required to
meet the engineering constraints (like for life-support systems where
machine code for software should be proven correct).
>> You've mutated a potentially
>> quite rediculously simple process into something that now has to
>> deal with and be aware of security and concurrency issues.
>
>It only has to worry about security in a system that has none, like Unix.
?! UNIX's security is excellent - you can get better, but I can't afford any
machines which can run such an operating system, so I'll have to make do
with UNIX.
>And you have to worry about concurrency anyways, at least this way it's
>the programmer who's worrying about it and not the helpless user being
>shafted by some lazy SOB.
True, but it's trivial - if you're editing a file, apply a mandatory lock -
unless the user doesn't want it locked. Most of the time, the user doesn't
give a crap, so don't bother unless they ask for it.
>> The benefit of complicating a simple text editor in this fashion
>> is far from obvious.
>
>On the contrary, making them "one-off" is what complicates processes.
>You just don't see this because Unix makes it impossible to have
>persistent processes, or secure processes, or multi-user processes,
>or pretty much anything useful.
When you want to abstract software into it's logical structure you should
stop talking about processes, they are an implementation detail. A
"persistent process" as you call it can be implemented as one process that
doesn't exit, and you start one-off processes which finds the persistent
processes and delegate the handling of messages to that process - This is
very common in UNIX. My Linux machine has a web server running - it is a
persistent process to which I can send messages by running a one-off
process. I suggest to read up on OOP and design patterns and you will see
these things being done behind the scenes for the implementation of exactly
what you describe. UNIX provides the primitives which can be used to
implement these things; it's not perfect, but it's much better than most
systems which is why it's so much more widely used by people who need to
implement them.
>> Emacs is convoluted enough on it's own (compared to apps like vi)
>> without finding the need to turn an editor into a little Oracle
>> wannabe.
>
>Emacs is proof that when given the choice, people won't lower themselves
>to coding in a nightmarish low-level language
Of course not, that would be stupid (try writing an OS kernel in prolog and
getting people to trust it :-).
>, nor interface with the
>nightmarish system that C is associated with.
This is a bit of a pathetic statement really.
--
Tristan Wibberley
------------------------------
From: [EMAIL PROTECTED] (D. Spider)
Subject: Re: Windows+Linux=True
Date: Sun, 17 Sep 2000 02:15:46 GMT
It appears that on Fri, 15 Sep 2000 04:14:18 GMT, in
comp.os.linux.advocacy Tim Hanson <[EMAIL PROTECTED]> wrote:
>Ingemar Lundin wrote:
>>
>> aaa...c'mon! command line interfaces is much harder to learn than a GUI
>> based one...
>> (altough i have to admit...i really like the Bash shell in Linux, specially
>> the tab-completion)
>>
>> /IL
>>
>> > Which brings up the eternal question: who determines of a given interface
>> > is easy to use or not? Why do people keep insisting that a graphical user
>> > interface in the style of Windows or MacOS is easy to use and the command
>> > line is not user friendly? There are so many things that are easy from
>> the
>> > command line but so difficult from the graphical user interface.
>> >
>> >
>
>To this day my fastest writing, hour after hour, day-in day-out, was
>using the Borland Sidekick editor on a DOS '286. I could steadily do
>110 wpm because my fingers never had to leave the keyboard. After the
>edit I would paste it into Wordperfect 5.1 for formatting.
This is one of the reasons why I use emacs regardless of OS (it's been
ported to just about anything, including DOS and Windows.) There's
never any need to take your hands off the keyboard to touch that
stupid mouse, and in fact you don't even need to leave the program to
add formating - just change major mode.
#####################################################
My email address is posted for purposes of private
correspondence only. Consent is expressly NOT given
to receive advertisements, or bulk mailings of any
kind.
#####################################################
------------------------------
From: T. Max Devlin <[EMAIL PROTECTED]>
Crossposted-To: alt.destroy.microsoft
Subject: Re: ZDNet reviews W2K server; I think you'll be surprised....
Date: Sat, 16 Sep 2000 22:27:07 -0400
Reply-To: [EMAIL PROTECTED]
Said <[EMAIL PROTECTED]> in alt.destroy.microsoft;
>In alt.destroy.microsoft T. Max Devlin <[EMAIL PROTECTED]> wrote:
>: Said Stuart Fox in alt.destroy.microsoft;
>:>
>:>IIRC, Microsoft have been around for 25 of those 30 years.
>:
>: Perhaps you missed the point, then, which is that monopoly crapware is
>: not real technology, and it doesn't matter how long its been around.
>:
>
>The 30 years T. Max refers to is probably meant to be the Internet,
>which is in fact 40 years old. Novell dominated Microsoft's DOS/Win 3.11
>early days at least upto 1993. So Microsoft in networks is more like 7
>years.
I'd have to disagree, on both counts. First, the Internet has not
existed since 1960. A precise date is debatable, of course, based on
what you peg as the start of "the Internet", but 1970 is a better
average date, if that. The modern Internet certainly didn't exist until
even later than that, but I simply used a round figure.
Second, in referencing Novell, you are confusing 'networking' with
'internetworking', i.e. the Internet. Microsoft has been screwing that
up for only 5 years, since they built TCP/IP (and IE) into Win95.
>Throughtout this whole discussion Fox has stubbornly refused to see
>any benefit of interoperability (and thus the damage of
>non-interoperability does to a network). Pretty sad for a Network
>Administrator.
I wish it were a unique situation, to be sure.
--
T. Max Devlin
-- Such is my recollection of my reconstruction
of events at the time, as I recall. Consider it.
Research assistance gladly accepted. --
====== Posted via Newsfeeds.Com, Uncensored Usenet News ======
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
======= Over 80,000 Newsgroups = 16 Different Servers! ======
------------------------------
From: T. Max Devlin <[EMAIL PROTECTED]>
Crossposted-To: alt.destroy.microsoft
Subject: Re: ZDNet reviews W2K server; I think you'll be surprised....
Date: Sat, 16 Sep 2000 22:31:08 -0400
Reply-To: [EMAIL PROTECTED]
Said <[EMAIL PROTECTED]> in alt.destroy.microsoft;
>In alt.destroy.microsoft [EMAIL PROTECTED] wrote:
>: In alt.destroy.microsoft T. Max Devlin <[EMAIL PROTECTED]> wrote:
>: : Said Stuart Fox in alt.destroy.microsoft;
>: :>
>: :>IIRC, Microsoft have been around for 25 of those 30 years.
>: :
>: : Perhaps you missed the point, then, which is that monopoly crapware is
>: : not real technology, and it doesn't matter how long its been around.
>: :
>:
>: The 30 years T. Max refers to is probably meant to be the Internet,
>: which is in fact 40 years old.
>
>Correction! T. Max was right.
Thanks for that. Was it just a math error, or have you changed what you
consider to be 'the start of the Internet'? Either way, it was an
off-handed comment. If I were to discuss the modern Internet, I'd
actually put it closer to 20 years. DNS wasn't developed until the 80s,
AFAIK, nor was Usenet (though the earlier BBS replication systems, split
between uucp and PC-based systems [fidonet?] were begun in the 70s, I
think). And the World Wide Web wasn't real technology until almost
1990, and later (Mosaic wasn't written until 1991, IIRC, though gopher
was a strong precursor). The decade of the 80s is when everything
really came together.
Thanks for your time. Hope it helps.
--
T. Max Devlin
-- Such is my recollection of my reconstruction
of events at the time, as I recall. Consider it.
Research assistance gladly accepted. --
====== Posted via Newsfeeds.Com, Uncensored Usenet News ======
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
======= Over 80,000 Newsgroups = 16 Different Servers! ======
------------------------------
From: "Otto" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.ms-windows.nt.advocacy
Subject: Re: Unix more secure, huh?
Date: Sun, 17 Sep 2000 02:32:15 GMT
"A transfinite number of monkeys" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
: On Sat, 16 Sep 2000 13:33:58 GMT, Otto <[EMAIL PROTECTED]> wrote:
: : And that suppose to diminish the validity of the actual news how? Maybe
you
: : should look at the following link, CERT released the warning about Linux
and
: : DDoS on Friday:
: :
: : http://www.cert.org/incident_notes/IN-2000-10.html
:
: Oh wow. It cites two vulnerabilities that have had patches available for
: quite some time, all within 24-48 hours after being found.
Oh wow, hundreds of systems are compromised on the daily basis with old
exploits. Availability means nothing, applying the patch might. It doesn't
really matter how quickly the patch is available, if people don't use them
anyway. All CERT did is, issued a warning about widespread use of old
exploits.
: How about all of the Windoze users out there that have (and continue) to
: fall prey to Netbus|BO|SubSeven|remote access trojan du jour? My firewall
: and IDS logs here at home can attest to the widespread use of those. My
: machines get scanned ALL THE TIME.
That's kind of interesting how a not so good news about Linux can turn out
to be really bad news for Windows (correct spelling). Let's forget the
rpc.statd problem and focus on Windows exploits. Linux computers will still
continue to fall pray to old exploits, but that's ok, as long as you can
show that Windows OSs are more vulnerable. What a great argument....
Every PC on the web gets scanned all the time, regardless of the OS. In case
you didn't know it's done by scripts and the results are recorded for later
use. The IDS is great to indentify the source, but that's about it.
: Here's a nickel. Go get a new arguement.
Here's a dime, go call yourself....
Otto
------------------------------
From: T. Max Devlin <[EMAIL PROTECTED]>
Crossposted-To: alt.destroy.microsoft
Subject: Re: ZDNet reviews W2K server; I think you'll be surprised....
Date: Sat, 16 Sep 2000 22:38:02 -0400
Reply-To: [EMAIL PROTECTED]
Said Stuart Fox in alt.destroy.microsoft;
[...]
>Minor point - complicate your DNS and split it in two. Your network can do
>what it likes
LOL!
--
T. Max Devlin
-- Such is my recollection of my reconstruction
of events at the time, as I recall. Consider it.
Research assistance gladly accepted. --
====== Posted via Newsfeeds.Com, Uncensored Usenet News ======
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
======= Over 80,000 Newsgroups = 16 Different Servers! ======
------------------------------
From: OSguy <[EMAIL PROTECTED]>
Crossposted-To: comp.ms.windows.advocacy,comp.ms.windows-nt.advocacy
Subject: Never tell me again that Windows is easy to install!!! It's a lie!
Date: Sat, 16 Sep 2000 21:38:51 -0500
I have wasted a whole day putting on Windows ME from scratch on my
daughter's computer (Celeron-350Mhz, 32M Ram). Her machine under
Windows 98 just got too bogged down, and slowly but surely, all of her
games and educational programs were getting too much for Win98 to
handle. So this morning I completely wiped and reformatted her hard
disk, then started the Windows ME installation. Two Hours later, it was
over. The Windows Install program, of course, told me it would be 30-60
minutes, but funny how it was 10 minutes to go for installation for
about 25 minutes while it was installing.
Linux has never taken me over 30 minutes to install on a machine.
Here where the fun began. Once Windows ME fully started up, it couldn't
find a driver for my jukebox-CD unit....in fact it couldn't find the
drive, and I couldn't get WinME to understand that there is a CD Drive
there. The insane thing is that the floppy disk used to boot and then
install WinME from the CD Drive not only recognized my jukebox CD
drive....it recognized ALL of the CDs in the drive. What type of morons
have an install program with a driver much more advanced then the
system? I sat there wondering why that driver on the floppy wasn't
integrated at all in the Windows ME System. I spent a lot of time
opening up the case of the machine and wondering what the h*LL is wrong
with this ME garbage, then it hit me. The jukebox CD was on a separate
IDE card creating the 3rd IDE port on the machine. Wondering if the
problem was that WinME couldn't see the 3rd IDE port, I switched the CD
to the second IDE port and took out the external IDE card. Finally, I
had a CD. So, the Install floppy, and Windows 98 could recognize my
external IDE, but WinME proper can't. What's wrong with this picture.
Usually one something works under linux, I have no trouble keeping it
working from upgrade to upgrade...the worst I have to do is to recompile
the 'driver' again under the kernel version I'm using, and it works just
fine. Also, If the Linux install can recognize the device, the fully
installed linux distro can understand it too. And the kernel-2.4.0-test
series understand additional IDE ports just fine.
Of course, my 16-bit sound card which worked under Win95 & Win98 didn't
work under WinME....why aren't I surprised? So I got to buy a new
Soundblaster Card just so WinME would understand it.
Somehow, I was under the impression WinME was supposed to be an updated
Win98...and the box says it maintains compatibility with Win95 and
Win98. Well, guess what? I found out that anything that installed its
driver in the Config.sys section of DOS and expected its driver to be
there before WinME comes up is now broken. WinME won't allow any DOS
'legacy' drivers in the system.
Well, 12 hours later, I finally have a system working as good as it was
before I wiped off the Win98 system, except now it goes to the disk a
lot......Well, I have 32M Ram, the Windows recommended, but what's this
note that says performance will improve with more memory? Has it ever
occurred to Microshaft to tell anybody what the optimum amount of memory
is for their system? Why must I try to flush that info out from all of
the MS Blurbmeisters? So later this week, I'll be buying more memory.
After this little episode, I really appreciate my linux machine which
never gives me trouble, works with existing hardware where I only
upgrade when I want to, and, while it make take a while to get all the
drivers for the devices I need, at least I can rest easier knowing that
once the driver is out, it will work for as long as I keep it compiled
to my system. And, I've never seen a Linux distro regress on
capabilities from update to update (No, I've never used Corel Linux).
So, as long as you Windows advocates keep coming into the
comp.os.linux.advocacy groups telling us how much easier Windows is to
install, I'm just going to laugh at you because I know you're either a
moron, ignorant, or just plain lying.
Oh, and to the trolls, this is a first-hand account true story, unlike
your works of fiction that you keep spewing.
Microshaft Sucks! I hope to NEVER buy another product that Gates had a
hand in selling!
------------------------------
From: Richard <[EMAIL PROTECTED]>
Subject: Re: Because programmers hate users (Re: Why are Linux UIs so crappy?)
Date: Sun, 17 Sep 2000 02:48:43 GMT
Donovan Rebbechi wrote:
> On Sat, 16 Sep 2000 23:43:06 GMT, Richard wrote:
> >One of the first things I noticed when I first got Linux was that
> >there were no extensions, and that seemed stupid. When I asked about
> >it, someone waved me at magic and file, and that shut me up. It wasn't
>
> The point is that there's some typing functionality, and that your
> shell is actually something that someone could write in LInux. It's
> not a design limitation.
That's alright, there are plenty of other limitations (many
of them because of design) in Unix.
> It'd be an interesting project. Like I said, I don't think that it didn't
> get implemented because of "nasty programmers". I think that it's not
> the kind of thing that anyone thought of doing.
There is an attitude common among programmers that users are
supposed to make the effort to learn the system. So if a user
overwrites /etc by accident because they wanted to restore
/home from backup then that is Their Fault, and not the fault
of the protection scheme that let them have write access to
/etc even when they never asked for it. Programmers accept the
limitations of software, forgetting that any such limitations
exist only because they put them there. And when users fail to
make the effort to learn all of these arbitrary limitations
(and why should they?), many programmers start bashing them
as lazy or stupid.
> It's worth mentioning that a lot of these obnoxious types are not
> the same people who are contributing actual code to any project.
Point well taken. Though, I find it hard to believe that the
pervasive mythos of the Hacker versus the Luser isn't widely
present among contributors.
> > If users don't seem to want something useful, it's
>
> The users certainly want a bunch of useful things.
>
> >through intimidation of the kind you are using. If you actually cared
>
> "intimidation" ? Sorry, I was having a bad day, and then you came
> on the usenet having the nerve to bash people who give their time to
> writing free software, and I got pissed off.
>
> Try being less obnoxious and you'll find the "evil programmers"
> will be more receptive. No one is going to be nice to you when
> you hurl abuse at them.
Usually it starts because I'm seriously pissed at Linux for
doing something completely stupid. So I'll plead a bad day
as well. :-)
> >to listen to user demands you would find their list is endless. And
>
> Well, the KDE project allow users to make feature requests, though there's no
> gaurantee that a feature will get implemented.
>
> Personally, I work closely with users to develop software which solves their
> problem. The software I write is written to the users specifications.
> If the user wants something I can't do, I say "sorry, I don't know how
> to do that".
This assumes they will admit their needs to you. Two of the biggest goals
any user has are:
1) to not feel stupid
2) to not make unrecoverable mistakes
When a user can wipe out everything with rm -rf, this is a failure of #2.
When they can't delete anything without a stupid confirmation box, this
is a failure of #1. The only acceptable solution is to have a logging FS
that allows users to easily undo any operation (move, delete, overwrite,
anything). And you're not going to find this in any spec, because creating
such a specification would require more self-knowledge and self-worth than
most people possess.
You can read more about user's real goals at:
http://www.cooper.com/articles/drdobbs_goal_directed.html
and for a concrete example,
http://www.cooper.com/articles/vbpj_ban_the_bomb.html
> Sometimes, a user will have conflicting demands. For example,
> software should be compatible and well designed. Or from a
> developers point of view, a toolkit should be high performance,m
> and "safe", and "object oriented".
>
> The demands are reasonable things to wish for, but implementing a
> solution is near impossible, and the end result is usually an attempt
> at a "best compromise".
Not to be trite, but I believe there is always a way to resolve such
conflicts if you think <swallowing distaste> outside of the box. In
the case of Smalltalk, I know there are ways to speed it up so you
get the best of both worlds: advanced compilers, native compilers,
(probably automatable) techniques to systematically fuse objects and
destroy boundaries, et cetera.
And in the case of rm -rf, Linuxers are remarkably quick to assume
that making a user feel stupid, and allowing unrecoverable errors
are an either-or proposition. They don't seem to work very hard at
thinking up ways to avoid both of these intolerable situations, and
that's probably the biggest reasons why I believe that programmers
despise users.
> Well here we go again with compatibility vs good design. UNIX is a
> shining example of the former. The things you mention are not, though they
> are a fairly good example of the latter.
>
> Compatibility seems to be winning for better or worse. And the reason that it's
> winning is because the users demand it.
I don't know that compatibility has had any serious competition yet.
Plan 9 doesn't give the end user any compelling reason for adoption,
nor does VSTa, and Grasshopper is a research OS so I don't think it
gives even programmers a compelling reason for adoption. And despite
what Alan Kay claimed, Smalltalk is /not/ an OS (unless something as
insecure as DOS qualifies).
Besides, there can be peaceful coexistance between OSes. I use Linux
for VisualWorks Smalltalk; Windows for games, media, and burning CDs;
and I want to roll my own for everything else.
------------------------------
From: "Otto" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.ms-windows.nt.advocacy
Subject: Re: Unix rules in Redmond
Date: Sun, 17 Sep 2000 02:53:12 GMT
"Chad Myers" <[EMAIL PROTECTED]> wrote in message
news:APSw5.5109$[EMAIL PROTECTED]...
: The Penguinistas have their head burried so far in the sand they can't
: even realize their own beloved OS' weakness.
The more appropriate substance where the Penguinistas burry their head would
be snow. However, that might induce some people to post tasteless
association about the snow and Linux. One might even call Linux a snow
job....
Otto
------------------------------
** 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
******************************
