Ordinary Computer User's Manifesto
Anonymous
your forums and mailing lists -- we are everywhere. We have been raped
by The Redmond Giant and we are desperate. We NEED our computers, and
we aren't going to go away. You have good computing, and we want it. We
turn to you for help. We buy Linux, and BSD, and whatever else catches
our eyes. We've heard of your OS, and we believe your stories about how
good it is.
You are the technicians, enthusiasts, hobbyists, and the gurus. Many of
you got into computing as a refuge from the teeming mass of idiot
humanity. Computers are more predictable, have rules, can be told what
to do. So you toil away at understanding them and making them do what
they do ever better. We clueless users are immensely impressed. We want
some. You won't escape us easily.
We are NOT willing to wade through your technical manuals, not at
first. Far too many of us want it handed over, finished, on a silver
platter. Wait long enough and that bunch will go away. But there are
plenty of us left over who will hang around until we get that one
little snippet we need to get the next technical problem solved. Even
if we read your manpages, manuals and handbooks, we could not possibly
understand them. There are portions that we grasp, but most of it is
not written for us, and we know it. So we may try, and then once we've
really gotten confused, we will invade your sacred sanctuaries and
pester you.
We are writers, artists, businessmen, musicians, and soccer moms. Never
mind our idiot teenage kids; some of them have already joined you, and
the rest have long been out of our hands. They are a separate problem.
We are the ones who have the money to go to computer stores and
bookstores and buy a copy of _XXXX for Dummies_ with our computers and
varied hardware. And in the cover of that book we find probably the
worst version of your favorite OS. Sometimes we get lucky and it's good
one. We won't know the difference. Some of us have no big bucks; we get
our machines second hand, or even out of the dumpster. And it works,
and we need to use it.
We'll pop that CD or DVD into the drive and install. Maybe. We are so
certain we don't want MicroStuff that we keep trying, even for hardware
that has no drivers in your OS. Because we usually do manage to stumble
through the process and get a workable system, we keep messing with it.
And then we come to your sanctuaries -- forums, mailing lists, Usenet
groups -- and we plead cluelessly, and sometimes not so cluelessly.
Since it seems most of the time we are rudely dismissed, we go and buy
more books, or borrow them, puzzle hopelessly through technical
documentation, and occasionly we find an answer that we understand.
But we keep using your OS. And we pollute the purity of the user pool,
because we don't bother to learn all the ins and outs that you know.
Well, we don't have time. We will learn what we need to know for our
own uses, and then we have work to do. The next time we mess something
up, we will then come ask again, or read some more. But computers are
not our lives, just our best tools. Worth some time and effort, but
nothing like what you invest in yours.
We will not be gurus; we don't have time. We are the ones who, for
better or worse, keep this crazy world going. You do your part; we do
ours. If you won't help us, we'll just keep blundering along. We don't
think like you, can't work like you, so stop demanding it. We aren't
you. If you as an individual don't want to help, we'll learn to leave
you alone. It's about freedom, and we treasure it for you as much as for
ourselves.
A few rare members of your elite cadre understand. They seem to have
kept somehow an understanding of the human race. They are the ones who
write the "...for Stupids" books and get some of our money. We gladly
give them what little or much we may have to get a glimpse inside your
mysterious world, but written at our level.
No we aren't going away. If we do, the Internet, as well as the
hardware manufacturers will be so overrun with That Monopoly OS that
all the fun is taken out of computing. Not to mention the freedom of
making choices. And all your carefully studied efforts at setting
standards are flushed down the toilet. Then the monster corporations
and governments and criminals (who can tell the difference?) will lock
the world out of the Internet we once knew. Free information comes to a
screeching halt. Nightmare visions follow.
We know we need you. Can you imagine you need us?
Oliver Douglas
> We will not be gurus; we don't have time. We are the ones who, for
> better or worse, keep this crazy world going. You do your part; we do
> ours. If you won't help us, we'll just keep blundering along. We don't
> think like you, can't work like you, so stop demanding it. We aren't
> you. If you as an individual don't want to help, we'll learn to leave
> you alone. It's about freedom, and we treasure it for you as much as for
> ourselves.
Are you one of the dumshits
that shops at Wal@Mart?
Cause they have some cheap
Linux desktops for you ( starting
at $200 ).
They are easy to use, and meet
all your criteria.
You'll save so much money
not paying the m$oft tax, you
can buy a nice LCD monitor as well.
There -- is that 'ordinary'
enough for you?
jpd
You only think you do. You can still write letters on typewriters, and
patience can be played on an ordinary desk-top as well as on that virtual
one. What else do you need? Email? C'mon, you still have faxes and
telephones and even *gasp* a fine postal service. Most of you at least
still barely remember how to write a proper letter. Not so with email,
even though it isn't all _that_ different.
> we aren't going to go away. You have good computing, and we want it. We
> turn to you for help. We buy Linux, and BSD, and whatever else catches
> our eyes. We've heard of your OS, and we believe your stories about how
> good it is.
No technical person was ever unwilling to offer help to those who were
genuinely interested. Shouting `YOU MUST HELP ME NOW FOR I AM STUPID'
sounds hopeless and not worth the effort. Worse, there's a real danger
for /my/ carpet to get permanently soaked in your drool if I'm going
to help you this once. I'm not particularly fond of carpets, but I
prefer them dry.
> You are the technicians, enthusiasts, hobbyists, and the gurus. Many of
> you got into computing as a refuge from the teeming mass of idiot
> humanity. Computers are more predictable, have rules, can be told what
> to do. So you toil away at understanding them and making them do what
> they do ever better. We clueless users are immensely impressed. We want
> some. You won't escape us easily.
There is a price. The price is opening your mind and trying to understand
our logic behind the tools. Heroic Faillure is no problem, unwillingness
to even try, is. Play with the haves, the knows, and you'll become one.
Scorn them, and get ignored, if you're lucky.
> We are NOT willing to wade through your technical manuals, not at
We are patient. You will learn.
[snip]
> But we keep using your OS. And we pollute the purity of the user pool,
> because we don't bother to learn all the ins and outs that you know.
> Well, we don't have time. We will learn what we need to know for our
> own uses, and then we have work to do. The next time we mess something
> up, we will then come ask again, or read some more. But computers are
> not our lives, just our best tools. Worth some time and effort, but
> nothing like what you invest in yours.
Not /taking/ the time to learn your /best tool/ is folly. Once you've
/taken/ the time to learn the basics, you'll need to learn to keep an
eye open, to use your /best tool/ with your /full attention/.
And why not? It is /your best tool/, after all.
> We will not be gurus; we don't have time. We are the ones who, for
> better or worse, keep this crazy world going.
Ha, I say. Ha!
You may keep it _warm_, but without us, you won't keep it going.
[snip: drivel]
>
> We know we need you. Can you imagine you need us?
Yes, we need you. To finally stop whining and to start learning what you
need to know. You cannot expect every tech to spoonfeed you the basics
over and over again, during your entire life. There simply are not
enough of us, and we've got better things to do. We don't _want_ to take
the time for a hopeless task, it'd be better, more efficient and a hell
of a lot less annoying to just kick you out and replace you with a small
shell script. You'd be amazing how much more a small shell script can do.
`we' are not interested in people taking pride in their faillures. It
doesn't matter if you don't know everything, just be sure that what
you know, you know well. And never be affraid to learn something new, or
at least try to learn.
That way, when one of us needs help in your field, you can return the
favour. That is all we ask, really.
--
j p d (at) d s b (dot) t u d e l f t (dot) n l .
Peter da Silva
for help.
Well, damnit, I agree with you. Now, I'm a pretty fair UNIX geek
myself, and I rarely run into problems with a plain vanilla UNIX
system. But lately I've been having the same experience over and
over:
Ingredients:
1 Middling fair UNIX geek.
1 Recent user-friendly UNIX distribution (Linux, Mac OS X, BSD, ...).
1 Canned script that's trying to make the thing user-friendly.
1 Requirement that the canned script doesn't quite handle.
Method:
Spend half an hour to half a day depending on processor speed,
interruptions, and impatience trying to use the canned scripts
the way they were supposed to be used.
Spend the rest of the day poring through 6000 lines of perl or
bash or python or tcl that's trying to emulate the Microsoft tool
the user's supposedly familiar with, that doesn't actually fit
the way UNIX works.
Go back to first principles, hardcode the configuration and make
it work, go home.
Leave in computer center 12 hours to ripen.
Next morning, get paged at 6:30, "The server's down!" "what server?"
"The NETWORK server!" (pause, request a moment to wake up, grab a
coffee cup, shrug, and chew a mouthful of Folgers) "Can you tell me
what you're trying to do? OK, OK, I think I know what the problem
is, I gotta hang up so I can dial in. Bye..."
Dial in, discover that the friendly script has trashed the config file
when it woke up and did its user-friendly check-if-the-user-messed-up
and-fix-it-for-them cron job.
Spend the morning removing all traces of the user-friendly script
and installing from source.
Conclusion.
If you are going to try and make a user-friendly configuration
script, base it on what the program you're configuring actually
does. Don't make it any longer than necessary. Build scripts that
have hooks for manual configuration. If you have the choice of
two programs, pick the one that's harder to mess up and easier
to fix, not the one with most features unless that's also the
easiest one for the user to deal with.
I'll start off with kudos, just for the heck of it.
FreeBSD's sysinstall seems pretty foolproof. It adds stuff to rc.conf
instead of changing what's there, so you can back it out. The defaults
files are pretty well-commented. You can look at them and see how you're
supposed to do things, and when you do things the way they're already being
done it works. That's Jordan's work, yes? Bravo, if they were all that
good everyone would be a lot happier. At least I would.
Now my gripes:
FreeBSD's mail configuration is the new damnable split sendmail, the one
where you're never quite sure which configuration is the one you're using.
Sendmail is probably still the most featureful mail transport, but it's a
real pain to configure. Try Postfix instead?
Red Hat seems to have a completely different config tool for every major
release, and I still find that I have to go in and edit the config scripts
themselves to change stuff.
Gnome's login manager is really pretty, but it's extremely unobvious how you're
supposed to add a new window manager choice. Still, it's no CDE.
Gnome again... how are you supposed to add a program to the default launcher?
Not "what's a way of doing it", I've found several, but what's the easy "drag
an icon here or type a program name there" option that end-users are supposed
to use?
SUSE: Is it my imagination, or does YAST really reinstall all your config
files from its database every time you log in?
Mac OS X: SystemStarter is nifty, I guess, but it does have a bit of a steep
learning curve.
--
I've seen things you people can't imagine. Chimneysweeps on fire over the roofs
of London. I've watched kite-strings glitter in the sun at Hyde Park Gate. All
these things will be lost in time, like chalk-paintings in the rain. `-_-'
Time for your nap. | Peter da Silva | Har du kramat din varg, idag? 'U`
Andy Baxter
This is one of the reasons I run debian - it has a few things like this,
but far less of it than mandrake for example, which does piss me off when
I have to do something like what you're talking about.
andy.
--
remove 'n-u-l-l' to email me. html mail or attachments will go in the spam
bin unless notified with [html] or [attachment] in the subject line.
Peter da Silva
Andy Baxter <ne...@earthsong.null.free-online.co.uk> wrote:
> This is one of the reasons I run debian - it has a few things like this,
> but far less of it than mandrake for example, which does piss me off when
> I have to do something like what you're talking about.
If I were running Linux I'd probably run debian, though I've found
myself in a twisty little maze of package-management tools when I've
run into it in the past. It certainly seems to have avoided the
demons of stupidity better than most.
But if I were recommending a version of "linux" for a novice user I'd
recommend "FreeBSD, and install the Postfix package", because it has
pretty few of these kinds of gotchas, AND it is pretty painless
building up to the kind of box the original poster is probably
expecting.
I'd probably suggest Windowmaker as the window manager / desktop, but I
don't know what to really suggest for a file manager. The Gnome one
just doesn't seem really happy without the Gnome world, and the others
are kind of crummy. I'd honestly like some suggestions here, actually.
Ori
> This is one of the reasons I run debian - it has a few things like this,
> but far less of it than mandrake for example, which does piss me off when
> I have to do something like what you're talking about.
>
> andy.
>
Slackware here, but same idea (plus the fact that it seems fast, and I'm
learning. I'm still a newbie, but I'm learning Linux faster than I would
normally be)
Jim Richardson
Hash: SHA1
On Thu, 16 Oct 2003 20:25:46 +0000 (UTC),
Peter da Silva <pe...@abbnm.com> wrote:
> In article <pan.2003.10.16...@earthsong.null.free-online.co.uk>,
> Andy Baxter <ne...@earthsong.null.free-online.co.uk> wrote:
>> This is one of the reasons I run debian - it has a few things like this,
>> but far less of it than mandrake for example, which does piss me off when
>> I have to do something like what you're talking about.
>
> If I were running Linux I'd probably run debian, though I've found
> myself in a twisty little maze of package-management tools when I've
> run into it in the past. It certainly seems to have avoided the
> demons of stupidity better than most.
>
> But if I were recommending a version of "linux" for a novice user I'd
> recommend "FreeBSD, and install the Postfix package", because it has
> pretty few of these kinds of gotchas, AND it is pretty painless
> building up to the kind of box the original poster is probably
> expecting.
>
> I'd probably suggest Windowmaker as the window manager / desktop, but I
> don't know what to really suggest for a file manager. The Gnome one
> just doesn't seem really happy without the Gnome world, and the others
> are kind of crummy. I'd honestly like some suggestions here, actually.
>
DFM works well with Windowmaker, as do the usual suspects in the cli
world (mc, emacs-dired, etc) Gentoo (filemanager, not distro) is a two
pane system that works very well with pentabs. It's similar to Opus from
the Amiga.
The one I would recommend you look at is XFE. XFE has a sort of
Explorer/Konqueror feel to it. It is simple to configure, fast, and
stable and very small. It has a couple of quirks, I *hate* horizontal
scrolls for directory listings, they suck. It uses the Fox toolkit. This
is the only app I use that uses it. On the plus side, it seems like a
nice toolkit, and there are python bindings. So maybe I'll tinker with
it sometime.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (GNU/Linux)
iD8DBQE/jw1wd90bcYOAWPYRAjhNAJ9DCp/TYwgzJxKnF8AIkAaTKb5o8gCdGEE+
aqrxjKZ65S+rBhvZeIiS/aE=
=t/sP
-----END PGP SIGNATURE-----
--
Jim Richardson http://www.eskimo.com/~warlock
Nine out of ten of the voices in my head say "Don't shoot!"
Andrew Reilly
> I'd probably suggest Windowmaker as the window manager / desktop, but I
> don't know what to really suggest for a file manager. The Gnome one
> just doesn't seem really happy without the Gnome world, and the others
> are kind of crummy. I'd honestly like some suggestions here, actually.
I've been running GNOME myself, but I hate the default file manager, so I
turn it off. I don't like root-window icons.
Go with ROX. You (Peter) have probably encountered RiscOS on an
Archimedes, well ROX started as a clone of the RiscOS file manager, but
now it's better. It really understands drag-n-drop, you can enter
directory paths from the keyboard, with filename completion, you can run a
one-line shell command with CWD set to the directory in question, it's
just great. Pretty quick to start, too.
If I was going to have a whinge about it at all, I'd complain that it
should sort file names in columns vertically, like ls, rather than
horizontally. At least it doesn't do the horizontal-scrolling thing.
--
Andrew
M. C. Widerkrantz
> I'd probably suggest Windowmaker as the window manager / desktop,
> but I don't know what to really suggest for a file manager. The
> Gnome one just doesn't seem really happy without the Gnome world,
> and the others are kind of crummy. I'd honestly like some
> suggestions here, actually.
When I set up an environment for my wife, a physician, I used the
fvwm2 window manager and the xftree file manager from the XFCE
project. I find that xftree looks and works a bit like the Macintosh
Finder in tree mode. It's small and doesn't require much resources or
colours and works quite OK on my wife's Sun Ultra 1.
Me, I've been stuck in ctwm for ages and I generally use Emacs dired
as my file manager.
--
MC, http://hack.org/mc/
jpd
> Well, that's three for three responses slamming you for crying out
> for help.
>
> Well, damnit, I agree with you. Now, I'm a pretty fair UNIX geek
> myself, and I rarely run into problems with a plain vanilla UNIX
> system. But lately I've been having the same experience over and
> over:
Maybe I'm silly, or maybe I'm just dead arrogant, possibly both. But,
I've found that it usually helps to not demand the world works like you
want it to, but to grab the manual, however arcane, and spend some time
with it. One might pick up a clue or two. Too bad they don't sell those
canned --like they do with london fog.
This goes for everything, including, but not limited to, the world of
computing. Nevermind that such manuals are often more than a millenium
old, written in drooling gibberish, or plain absent. It simply is easier
and faster to play by and with the rules than to go against them. Even
for purposes of ``resistance within''.
Maybe the first thing is to stop assuming you cannot change anything.
You might not be able to change _much_, on the grand scale of things,
but you'd be surprised how much one _can_ achieve. Even if it only
changes yourself.
[snip: cooking with user friendly scripts]
> Spend the morning removing all traces of the user-friendly script
> and installing from source.
>
> Conclusion.
>
> If you are going to try and make a user-friendly configuration
> script, base it on what the program you're configuring actually
> does. Don't make it any longer than necessary. Build scripts that
> have hooks for manual configuration. If you have the choice of
> two programs, pick the one that's harder to mess up and easier
> to fix, not the one with most features unless that's also the
> easiest one for the user to deal with.
I'd say this is caused by too many lusers crying out for help, the wrong
way. The result, in a way, has a lesson that still hasn't been learned,
or that people insist on not learning over and over again: all the world
is not a vax. And then you get stuck by not having the choice of not
using the oh-so-helpful userfriendly script, because, obviously, not
having the choice is much more userfriendly that having it.
It is also said that having choice brings responsibility. So taking that,
as a first step, gets you back the choice, and hence the power to choose.
Enough people do that, and ``the other side'' starts to get a grip.
Enough people sue someone else for their own stupidity _and win_,
and nobody gets to choose anymore. Instead of getting the legal system
and the litigous abusers whacked.
[snip]
> Now my gripes:
>
> FreeBSD's mail configuration is the new damnable split sendmail, the one
> where you're never quite sure which configuration is the one you're using.
> Sendmail is probably still the most featureful mail transport, but it's a
> real pain to configure. Try Postfix instead?
You can, or use exim or well, what options do we have. Lots of stuff in
the ports collection. Just not directly supported by the sysinstall tui.
> Red Hat seems to have a completely different config tool for every major
> release, and I still find that I have to go in and edit the config scripts
> themselves to change stuff.
So there's no point in having it, really, beyond installing the basic stuff
on your machine. So, maybe, it should not pretend to be more?
[snip: gnome]
I don't use a DE. I won't comment on that.
> SUSE: Is it my imagination, or does YAST really reinstall all your config
> files from its database every time you log in?
Every time you run it, at least, and it's pretty undoable to try and reconfig
it without yast. *sigh* My DHCP server doesn't have that problem. *grin*
Byron A Jeff
Anonymous <dam...@hotmail.com> wrote:
Phenominal piece! I'll throw in some comments and some basic do's and don'ts.
-We are here. On your doorstep, in your lab, in your inbox, invading
-your forums and mailing lists -- we are everywhere. We have been raped
-by The Redmond Giant and we are desperate. We NEED our computers, and
-we aren't going to go away. You have good computing, and we want it. We
-turn to you for help. We buy Linux, and BSD, and whatever else catches
-our eyes. We've heard of your OS, and we believe your stories about how
-good it is.
I think you're overhyping the reality of the situation. Many folks have only
heard of alternatives. Many others cannot comprehend the fact that there
are alternatives. Finally the toughest group are the ones who have a bunch
of misconceptions about the alternatives.
-You are the technicians, enthusiasts, hobbyists, and the gurus. Many of
-you got into computing as a refuge from the teeming mass of idiot
-humanity. Computers are more predictable, have rules, can be told what
-to do. So you toil away at understanding them and making them do what
-they do ever better. We clueless users are immensely impressed. We want
-some. You won't escape us easily.
Interesting point. I'll continue to run away screaming though! ;-)
-We are NOT willing to wade through your technical manuals, not at
-first. Far too many of us want it handed over, finished, on a silver
-platter. Wait long enough and that bunch will go away.
Third adopters needs to be retrained.
- But there are
-plenty of us left over who will hang around until we get that one
-little snippet we need to get the next technical problem solved. Even
-if we read your manpages, manuals and handbooks, we could not possibly
-understand them. There are portions that we grasp, but most of it is
-not written for us, and we know it. So we may try, and then once we've
-really gotten confused, we will invade your sacred sanctuaries and
-pester you.
Second adopters are always welcomed because they are always eager to learn.
-
-We are writers, artists, businessmen, musicians, and soccer moms. Never
-mind our idiot teenage kids; some of them have already joined you, and
-the rest have long been out of our hands. They are a separate problem.
-We are the ones who have the money to go to computer stores and
-bookstores and buy a copy of _XXXX for Dummies_ with our computers and
-varied hardware. And in the cover of that book we find probably the
-worst version of your favorite OS. Sometimes we get lucky and it's good
-one. We won't know the difference. Some of us have no big bucks; we get
-our machines second hand, or even out of the dumpster. And it works,
-and we need to use it.
OK. Now here are some rules:
1) You don't have to know everything, however you do need to hookup with
someone who does. The reason is that one learns the landscape by experience.
Events that flat stymie the newbie are been there done that for more advanced
users. Having the learn everything for the first time is slow and frustrating.
2) One partial retraining is to get 2nd adopters to clue into the fact that
they should start out as users and leave the admin task for later. It's one
of the reasons I often give new users copies of Knoppix. It autoconfigures
almost everything, so there is only a small initial learning curve to get
rolling.
3) When it comes to hardware, we live in a Windows world. PC hardware is
designed almost exclusively with Windows in mind. So you have to do your
research before your buy. Often a simple "device" +Linux in google will
give you a sense of what the support is, even if you don't understand the
details.
4) The mantra: It's not Windows, it's not Windows, it's not Windows. Nothing
irrates more than users insisting that the new tools behave precisely like
the tools they are used to. The differences exists for a reason. Take the
time to figure how why things are done differently.
-We'll pop that CD or DVD into the drive and install. Maybe. We are so
-certain we don't want MicroStuff that we keep trying, even for hardware
-that has no drivers in your OS. Because we usually do manage to stumble
-through the process and get a workable system, we keep messing with it.
-And then we come to your sanctuaries -- forums, mailing lists, Usenet
-groups -- and we plead cluelessly, and sometimes not so cluelessly.
-Since it seems most of the time we are rudely dismissed, we go and buy
-more books, or borrow them, puzzle hopelessly through technical
-documentation, and occasionly we find an answer that we understand.
Another issue: Asking the right questions in the right manner. Eric Raymond
wrote an excellent article on the subject. Find it here:
http://www.catb.org/~esr/faqs/smart-questions.html
it's easier to attract flies with honey than vinegar.
-
-
-But we keep using your OS. And we pollute the purity of the user pool,
-because we don't bother to learn all the ins and outs that you know.
It's not pollution until cross the line from being a clueless user to
being an obnoxious, pushy, clueless user that assumes that we are here
simply to service their needs.
-Well, we don't have time. We will learn what we need to know for our
-own uses, and then we have work to do. The next time we mess something
-up, we will then come ask again, or read some more. But computers are
-not our lives, just our best tools. Worth some time and effort, but
-nothing like what you invest in yours.
That's a tough one for much of the developer community to comprehend. The
art of teaching where the students are is complicated and requires
enlightened teachers.
-
-
-We will not be gurus; we don't have time. We are the ones who, for
-better or worse, keep this crazy world going. You do your part; we do
-ours. If you won't help us, we'll just keep blundering along. We don't
-think like you, can't work like you, so stop demanding it. We aren't
-you. If you as an individual don't want to help, we'll learn to leave
-you alone. It's about freedom, and we treasure it for you as much as for
-ourselves.
Excellent point. But the responsibility goes both ways. There's a difference
between starting out clueless, and choosing to remain there. First choose
to invest some time learning the new tool. Second invest just a little bit
of time searching for the answer yourself before pulling others into the
loop.
There's a huge chasm between "Think like us." and "Think for yourself."
-A few rare members of your elite cadre understand. They seem to have
-kept somehow an understanding of the human race. They are the ones who
-write the "...for Stupids" books and get some of our money. We gladly
-give them what little or much we may have to get a glimpse inside your
-mysterious world, but written at our level.
Actually it's probably just as well to keep the abstraction and pay someone
else to work under the hood. The perfect analogy is the shadetree mechanic.
Not everyone has the skill, understanding, tools, and time to work on their
own vehicles. But somehow everyone thinks they have the skillset to know
everything about their own computers.
-No we aren't going away. If we do, the Internet, as well as the
-hardware manufacturers will be so overrun with That Monopoly OS that
-all the fun is taken out of computing. Not to mention the freedom of
-making choices. And all your carefully studied efforts at setting
-standards are flushed down the toilet. Then the monster corporations
-and governments and criminals (who can tell the difference?) will lock
-the world out of the Internet we once knew. Free information comes to a
-screeching halt. Nightmare visions follow.
Agreed. So a symbiotic relationship needs to be forged.
-
-
-We know we need you. Can you imagine you need us?
We do. I hope everyone figures this out.
BAJ
Peter da Silva
Jim Richardson <war...@eskimo.com> wrote:
> DFM works well with Windowmaker, as do the usual suspects in the cli
> world (mc, emacs-dired, etc) Gentoo (filemanager, not distro) is a two
> pane system that works very well with pentabs. It's similar to Opus from
> the Amiga.
I'll have a look at DFM, though I'm a little concerned with its use of the
root window, since Windowmaker needs the root window as well. I'll check out
XFE as well. Thanks.
I'll let the two-pane babies slide. Not what I'm looking for, at all.
Wish I had as much free time as I did in the '80s, I'd port my Amiga
Browser to UNIX.
Peter da Silva
Andrew Reilly <and...@gurney.reilly.home> wrote:
> Go with ROX.
Um, not only no but hell no. It's what I'm using right now, and the layers
of fiddly stuff you have to do to add actions, combined with the totally
counterintuitive and uncustomisable menus... no thanks. It works, but it's
not what I'd recommend to someone who wants a program instead of a lesson
in MIME and XML.
It's got a lot of good features, but it's got way too many of the kinds
of things I was complaining about in my first message to consider.
Peter da Silva
jpd <read_t...@do.not.spam.it> wrote:
> Maybe I'm silly, or maybe I'm just dead arrogant, possibly both. But,
> I've found that it usually helps to not demand the world works like you
> want it to, but to grab the manual, however arcane, and spend some time
> with it.
Oh, if I was king of the world I'd force people to write their own
firewall before they were allowed on the Internet, but I'm not, and
your not, and the problems the original poster and myself are talking
about *are* problems that people who write software, want it to be used
by people who don't think pretty much exactly like themselves, and
aren't (for good or bad) king of the world bloody well should be
thinking about.
> Maybe the first thing is to stop assuming you cannot change anything.
No, the first thing is to listen.
I write a lot of software, I give away a lot of software, a lot of the
software I give away turns out to be configuration tools for other people's
software, most recently for a keyboard remapper under Mac OS X. I didn't
like the shortcomings in Amiga's Workbench in AmigaDOS 1.3, and wrote a
replacement... and lo and behold most of its features showed up in the
Workbench in AmigaDOS 2.0...
Maybe I'm tired of just leading by example?
Oh, I can change lots of things. It's funny how the Workbench in AmigaDOS 2.0
picked up just about every feature I'd put in Browser including almost exactly
the same file requestor.
[more BOFHishness deleted]
Quit telling me what I can do, and start thinking about WHY I'm pointing this
stuff out.
Mark
> In article <h4f361-...@grendel.myth>,
> Jim Richardson <war...@eskimo.com> wrote:
>
>>DFM works well with Windowmaker, as do the usual suspects in the cli
>>world (mc, emacs-dired, etc) Gentoo (filemanager, not distro) is a two
>>pane system that works very well with pentabs. It's similar to Opus from
>>the Amiga.
>
>
> I'll have a look at DFM, though I'm a little concerned with its use of the
> root window, since Windowmaker needs the root window as well. I'll check out
> XFE as well. Thanks.
>
> I'll let the two-pane babies slide. Not what I'm looking for, at all.
>
> Wish I had as much free time as I did in the '80s, I'd port my Amiga
> Browser to UNIX.
>
How about xffm that is with Xfce4?
jpd
> In article <bmom9s$79c$1...@news1.transmedia.de>,
> jpd <read_t...@do.not.spam.it> wrote:
> Quit telling me what I can do, and start thinking about WHY I'm pointing this
> stuff out.
My you wasn't ment as a specific you, mind you. I'm impressed by what you
(specific this time) did and achieved.
Anyway, yes, I can see why you point it out. The main reason why I'm
pointing the other direction is that, IMAO, there's far too much telling
``hey, this is (well, ought to be) easy''. With the effect that everyone
starts to believe everything should be easy and intuitive (whatever that
may be), even when it shouldn't, really.
That arrogance, --it's arrogance, even if totally different from the
arrogance encountered at the tech side-- is just as bad as its
counterpart and shouldn't be promoted.
Of course this is crying for a better world, just as the OP did, but from
a different perspective. Now for the golden middleroad...
Peter da Silva
It's important to remember who it is that is trying to attract whom.
There's a lot of peole who *are* trying to sell Linux to Windows users,
names like Red Hat, Gnome, KDE, ...
They're doing it wrong. One of the big gripes I have with Windows is that
Microsoft thinks a pretty face excuses a clumsy design. These people give
us a pretty face, too, but they don't even do as good a job of being easy
for people to personalise as Windows.
Peter da Silva
> > release, and I still find that I have to go in and edit the config scripts
> > themselves to change stuff.
> So there's no point in having it, really, beyond installing the basic stuff
> on your machine. So, maybe, it should not pretend to be more?
It's not even good for that.
Look, I'm capable of reading the python source to the scripts to find
the name of the rc file that you have to edit to fix it, but I doubt if
one in a hundred Red Hat users could. And I really DID have to, because
changing the settings in an "rc.local" resulted in the interface
getting reconfigured too late, or too early.
And, again, there's the FreeBSD sysinstall as an example of a way to do
it that works for both kinds of users and doesn't stick you with a bad
configuration.
You run sysinstall, and it adds the lines to /etc/rc.conf:
# -- sysinstall generated deltas -- # Sun May 3 10:15:30 2000
nfs_server_enable="YES"
hostname="fileserver1.example.com"
# -- sysinstall generated deltas -- # Tue Jun 25 20:02:39 2002
nfs_server_enable="NO"
sendmail_enable="NONE"
hostname="socksproxy.example.com"
ifconfig_fxp1="inet 192.168.0.2 media autoselect netmask 255.255.255.0"
So, the system was configured as a fileserver in May 2000, then it was
re-used as a SOCKS proxy in June 2002. Changes always go on the end, and
only include stuff that's changed from the default. If you change your
mind, you can back up. If you want to change it manually, you add your
own changes to the end, it doesn't mess anything up.
Liam Slider
> In article <bmp3sm$5...@cleon.cc.gatech.edu>,
> Byron A Jeff <by...@cc.gatech.edu> wrote:
>
>>it's easier to attract flies with honey than vinegar.
>
>
> It's important to remember who it is that is trying to attract whom.
>
> There's a lot of peole who *are* trying to sell Linux to Windows users,
> names like Red Hat, Gnome, KDE, ...
>
> They're doing it wrong. One of the big gripes I have with Windows is that
> Microsoft thinks a pretty face excuses a clumsy design. These people give
> us a pretty face, too, but they don't even do as good a job of being easy
> for people to personalise as Windows.
>
That's simply not true.
--
"Soon Bill Gates will have something called Total Information
Technology, TIT!!! and when your sucking on the TIT, I'VE GOT YOU BY THE
MOTHERBOARD!!!" --Robin Williams
Peter da Silva
jpd <read_t...@do.not.spam.it> wrote:
> In article <bmpfmv$13e9$4...@jeeves.eng.abbnm.com>, Peter da Silva wrote:
> > In article <bmom9s$79c$1...@news1.transmedia.de>,
> > jpd <read_t...@do.not.spam.it> wrote:
> > Quit telling me what I can do, and start thinking about WHY I'm pointing this
> > stuff out.
>
> My you wasn't ment as a specific you, mind you. I'm impressed by what you
> (specific this time) did and achieved.
>
> Anyway, yes, I can see why you point it out. The main reason why I'm
> pointing the other direction is that, IMAO, there's far too much telling
> ``hey, this is (well, ought to be) easy''.
I believe that if there isn't a good reason why it's hard (and, yes, some
of it does have a good reason) then it really ought to be easy.
One of the reasons I like working on Open Source software is that when
someone says "hey, it'd be nice if it did this" and I can see a good
way to make it DO that, easily, I can do it. For my day job just adding
a link to a webpage can require a week's worth of meetings.
And that's why it really peeves me when there's a piece of open source
software that's hard where it doesn't need to be. And it *really* peeves
me when I write the code to fix it, and send it in, and the author comes
back with "yeh, I can see that's a useful feature, but I don't use it that
way so I won't put it in".
And I want to start cracking skulls when someone justifies leaving some
clumsy workaround in place because it's more user-friendly than not needing
the workaround, because they're confusing "user friendly" with "has a nice
step-by-step interface to configure things the way the author thinks they
should be configured".
So, the problem is not that people say "it should be easy" too often, it's
that people don't try and figure out what "it should be easy" means.
It doesn't mean "you shouldn't have to know how it works".
It doesn't mean "it should be just like this other tool I like", even
if it's phrased that way.
It doesn't mean "I'm lazy".
It means that there should be an obviously right way to use the tool for the
most common cases, and that obviously right way should be the way the user
is going to find when they start using the tool, and that right way should
work well with the other ways that more advanced users are going to use
the tool.
So, let's look at some examples.
Let's say you want to add an "Open With" menu item in a file browser. How do
you think you should have to do it?
A. Select "Open with other...", locate an application in the file
system, and click "Add"?
B. Shift-click on the menu to open a window, and drag a file into
that window?
C. Select 'prefs' or 'configure', open 'file handlers'. type in
the file extension, and type in the name of the program?
D. Open prefs again, open 'file types', find the file extension,
select 'add', and type in the name of the program.
E. Open a hidden directory in your home directory, copy an obscure
system file there, edit the XML to add a new MIME type for the
file, select "reload", create a subdirectory with a name that
matches that MIME type, and drag the application there with a
key held down, and select "make link".
F. Select the application, copy it to your home directory, select
"Show package contents", use an XML editor to add the file
extension, purge the launch services cache, and then always
run *that* copy of the application.
I think most users would be happy with A through D. I wish I could say E and
F were parodies. And yet both are supposed to be user-friendly.
Peter da Silva
Liam Slider <li...@NOSPAM.liamslider.com> wrote:
> Peter da Silva wrote:
> > They're doing it wrong. One of the big gripes I have with Windows is that
> > Microsoft thinks a pretty face excuses a clumsy design. These people give
> > us a pretty face, too, but they don't even do as good a job of being easy
> > for people to personalise as Windows.
> That's simply not true.
I dearly wish you were right.
Liam Slider
> In article <bmpm1n$q0vj8$2...@ID-169482.news.uni-berlin.de>,
> Liam Slider <li...@NOSPAM.liamslider.com> wrote:
>
>>Peter da Silva wrote:
>>
>>>They're doing it wrong. One of the big gripes I have with Windows is that
>>>Microsoft thinks a pretty face excuses a clumsy design. These people give
>>>us a pretty face, too, but they don't even do as good a job of being easy
>>>for people to personalise as Windows.
>
>
>>That's simply not true.
>
>
> I dearly wish you were right.
It's just as easy, or easier, using one of the modern desktop
envornments, to personalise Linux as it is to personalise Windows. Often
you have more options. I can change my wm on a whim, I certainly can't
do that in Windows. I can easily change backgrounds, change "wallpaper",
change themes, I can have certain sound files play at certain times
using an easy to set up cron file. I can set up a "wallpaper" that
rotates between a number of different images if I want. I can have
multiple virtual desktops, each with a different "wallpaper" image. Can
you personalise that on Windows? Oh wait, Windows doesn't even have
multiple virtual desktops.
In fact, Windows doesn't have many of these options. Usually, you can
theme, change some sounds, and change your "wallpaper." On some versions
(I don't know if you can on XP) you can hack the boot screen. Yeah, lots
of personalisation there.
Kingbarry2000
"Liam Slider" <li...@NOSPAM.liamslider.com> wrote in message
news:bmpq2b$opvqp$1...@ID-169482.news.uni-berlin.de...
> Peter da Silva wrote:
> > In article <bmpm1n$q0vj8$2...@ID-169482.news.uni-berlin.de>,
> > Liam Slider <li...@NOSPAM.liamslider.com> wrote:
> >
> >>Peter da Silva wrote:
> >>
> >>>They're doing it wrong. One of the big gripes I have with Windows is
that
> >>>Microsoft thinks a pretty face excuses a clumsy design. These people
give
> >>>us a pretty face, too, but they don't even do as good a job of being
easy
> >>>for people to personalise as Windows.
> >
> >
> >>That's simply not true.
> >
> >
> > I dearly wish you were right.
>
>
> It's just as easy, or easier, using one of the modern desktop
> envornments, to personalise Linux as it is to personalise Windows. Often
> you have more options. I can change my wm on a whim, I certainly can't
> do that in Windows. I can easily change backgrounds, change "wallpaper",
> change themes, I can have certain sound files play at certain times
> using an easy to set up cron file. I can set up a "wallpaper" that
> rotates between a number of different images if I want. I can have
> multiple virtual desktops, each with a different "wallpaper" image. Can
> you personalise that on Windows? Oh wait, Windows doesn't even have
> multiple virtual desktops.
You can do all of that on windows. Freeware and shareware and commercial
apps do
that and more.
>
> In fact, Windows doesn't have many of these options. Usually, you can
> theme, change some sounds, and change your "wallpaper." On some versions
> (I don't know if you can on XP) you can hack the boot screen. Yeah, lots
> of personalisation there.
You can make windows look and feel like any OS / windowing system.
3rd party software can do it all.
>
>
> --
> "Soon Bill Gates will have something called Total Information
> Technology, TIT!!! and when your sucking on the TIT, I'VE GOT YOU BY THE
> MOTHERBOARD!!!" --Robin Williams
>
--
I work for SUTI, an offshoot of SETI.
( Search for Usenet Terrestrial Intelligence )
No luck so far.
All outgoing mail is certified content free.
Its cookies, all the way down.
Andrew Reilly
> In article <pan.2003.10.16....@gurney.reilly.home>,
> Andrew Reilly <and...@gurney.reilly.home> wrote:
>> Go with ROX.
>
> Um, not only no but hell no. It's what I'm using right now, and the layers
> of fiddly stuff you have to do to add actions, combined with the totally
> counterintuitive and uncustomisable menus... no thanks. It works, but it's
> not what I'd recommend to someone who wants a program instead of a lesson
> in MIME and XML.
Bollocks. I've never had to tweak MIME or XML to use ROX.
Rox doesn't have an "open with" menu item because that's not the way it
works. If you want some non-default application to read a particular
file, then you open the application and then drag the file into it. If
you want to *set* a default run action, then you either drag the
application icon into the spot in the set-run-action dialog, or you type
in the command you want to use. It's just not that difficult.
> It's got a lot of good features, but it's got way too many of the kinds
> of things I was complaining about in my first message to consider.
I must have missed your earlier message. Perhaps you're asking it to do
things that it's not designed to do? While I use ROX every day, I hardly
ever have a ROX window open. I just don't need to move files around that
often. Web browsing doesn't need it. E-mail reading and writing doesn't
need it, although it can be used to specify attachments and destinations
if you like.
--
Andrew
Jim Richardson
Hash: SHA1
["Followup-To:" header set to comp.os.linux.advocacy.]
On Sat, 18 Oct 2003 03:46:55 GMT,
assuming you don't mind more crashes, and freezes. Not to mention, that
setting it all up, is a PITA compared to Linux or *BSD. Especially since
for some of it (the open source stuff mostly) you have to set up a gnu
developement toolchain, since the MICROS~1 toolchain won't work, (and
costs big $$)
>>
>> In fact, Windows doesn't have many of these options. Usually, you can
>> theme, change some sounds, and change your "wallpaper." On some versions
>> (I don't know if you can on XP) you can hack the boot screen. Yeah, lots
>> of personalisation there.
>
> You can make windows look and feel like any OS / windowing system.
> 3rd party software can do it all.
I now wait for you to show me a port of, or clone of, the ION
windowmanager for MICROS~1 windows.
<http://modeemi.cs.tut.fi/~tuomov/ion/>
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (GNU/Linux)
iD8DBQE/kMSZd90bcYOAWPYRAlyGAJ4qTXvPr4HlAy19a2+hoXAytvneiQCdE+Ks
5PJc/QVfH4rUr9XvPb3cKMA=
=1Djr
-----END PGP SIGNATURE-----
--
Jim Richardson http://www.eskimo.com/~warlock
Why is American beer served cold? So you can tell it from urine.
-- David Moulton
Ilari Liusvaara
<pan.2003.10.16....@gurney.reilly.home>. Dumping datagram.
>
> Go with ROX. You (Peter) have probably encountered RiscOS on an
> Archimedes, well ROX started as a clone of the RiscOS file manager, but
> now it's better. It really understands drag-n-drop, you can enter
> directory paths from the keyboard, with filename completion, you can run a
> one-line shell command with CWD set to the directory in question, it's
> just great. Pretty quick to start, too.
It also has application directories, which are like Mac virtual disks.
Very nice for application management.
-Ilari
--
Guys, Linux is not a research project. Never was, never will be. --
Linus Torvalds
Linux LK_Perkele_IV9 2.4.22-rc3 #2 Sun Aug 24 14:36:19 EEST 2003 i686 unknown
11:18pm up 33 days, 11:12, 13 users, load average: 0.10, 0.03, 0.01
Liam Slider
But not out of the box? Not built out of stable components of the OS?
Well fuck. Looks like Linux wins. I was talking about the stuff we can
do standard, you want to get fancy of course we can do more.
>
>
>>In fact, Windows doesn't have many of these options. Usually, you can
>>theme, change some sounds, and change your "wallpaper." On some versions
>>(I don't know if you can on XP) you can hack the boot screen. Yeah, lots
>>of personalisation there.
>
>
> You can make windows look and feel like any OS / windowing system.
> 3rd party software can do it all.
And you can do the same thing on Linux...even make it look like Wintendo
if you want. And again, you can't do any of that *out of the box*.
Windows is pathetic.
jpd
> In article <bmpipf$edm$1...@news1.transmedia.de>,
> jpd <read_t...@do.not.spam.it> wrote:
>
> I believe that if there isn't a good reason why it's hard (and, yes, some
> of it does have a good reason) then it really ought to be easy.
Then we get to the next problem: one's easy isn't necesairily the other's.
I don't want a gui, I'm _much_ quicker off writing a longish commandline
using the standard tools than trying where $company hid the such-and-so
magic button /this release/. If I need to do more of the same, I'll dump
the commands in a script. I do understand there's a lot of people that
balk at that, for a variety of reasons, some of them legitimate.
But that doesn't solve the issue. What Is Easy?
> One of the reasons I like working on Open Source software is that when
> someone says "hey, it'd be nice if it did this" and I can see a good
> way to make it DO that, easily, I can do it. For my day job just adding
> a link to a webpage can require a week's worth of meetings.
That sounds like the organisation is broken. I'm going to be totally
arrogant and point out that's a people problem, and that one shouldn't
try and solve that through technical means. Nevermind that you very
often simply _cannot_ do that in the first place.
> And that's why it really peeves me when there's a piece of open source
> software that's hard where it doesn't need to be. And it *really* peeves
> me when I write the code to fix it, and send it in, and the author comes
> back with "yeh, I can see that's a useful feature, but I don't use it that
> way so I won't put it in".
There's the other side of the author not wanting to put it in because
(s)he's deemed it unmaintainable. There is a legitimate reason in that,
even if the wording is unfortunate. Doesn't mean _I_ would react that
way, but that's besides the point.
Then again, this medium seems to invite the very worst of human non-
communication. Because it's supposed to be easy nobody pays attention.
That's not what I'd call a /desirable/ kind of easy.
> And I want to start cracking skulls when someone justifies leaving some
> clumsy workaround in place because it's more user-friendly than not needing
> the workaround, because they're confusing "user friendly" with "has a nice
> step-by-step interface to configure things the way the author thinks they
> should be configured".
Or the other variety of that: Let's invent /yet another tool/ to do the
same because I have these weird-ass preferences nobody else has.[1]
> So, the problem is not that people say "it should be easy" too often, it's
> that people don't try and figure out what "it should be easy" means.
>
> It doesn't mean "you shouldn't have to know how it works".
>
> It doesn't mean "it should be just like this other tool I like", even
> if it's phrased that way.
>
> It doesn't mean "I'm lazy".
>
> It means that there should be an obviously right way to use the tool for the
> most common cases, and that obviously right way should be the way the user
> is going to find when they start using the tool, and that right way should
> work well with the other ways that more advanced users are going to use
> the tool.
Well said.
Now, what about learning curve?
> So, let's look at some examples.
>
> Let's say you want to add an "Open With" menu item in a file browser. How do
> you think you should have to do it?
>
> A. Select "Open with other...", locate an application in the file
> system, and click "Add"?
>
> B. Shift-click on the menu to open a window, and drag a file into
> that window?
>
> C. Select 'prefs' or 'configure', open 'file handlers'. type in
> the file extension, and type in the name of the program?
>
> D. Open prefs again, open 'file types', find the file extension,
> select 'add', and type in the name of the program.
>
> E. Open a hidden directory in your home directory, copy an obscure
> system file there, edit the XML to add a new MIME type for the
> file, select "reload", create a subdirectory with a name that
> matches that MIME type, and drag the application there with a
> key held down, and select "make link".
>
> F. Select the application, copy it to your home directory, select
> "Show package contents", use an XML editor to add the file
> extension, purge the launch services cache, and then always
> run *that* copy of the application.
>
> I think most users would be happy with A through D. I wish I could say E and
> F were parodies. And yet both are supposed to be user-friendly.
C and D sound OK, don't know about A and B. E and F sound like the
application isn't finished yet. But then again some people or companies
have varied, interesting and ever changing notions of what user-friendly
and/or easy is or ought to be. OTOH, it all depends awfully much on file
extentions. Can't we do better than that?
[1] Cue the various ifconfig replacements for a certain kernel on cra^Wgn00.
Among others, of course.
Peter da Silva
Liam Slider <li...@NOSPAM.liamslider.com> wrote:
> Peter da Silva wrote:
> > In article <bmpm1n$q0vj8$2...@ID-169482.news.uni-berlin.de>,
> > Liam Slider <li...@NOSPAM.liamslider.com> wrote:
> >>Peter da Silva wrote:
> >>>They're doing it wrong. One of the big gripes I have with Windows is that
> >>>Microsoft thinks a pretty face excuses a clumsy design. These people give
> >>>us a pretty face, too, but they don't even do as good a job of being easy
> >>>for people to personalise as Windows.
> >>That's simply not true.
> > I dearly wish you were right.
> It's just as easy, or easier, using one of the modern desktop
> envornments, to personalise Linux as it is to personalise Windows.
Personalise Linux? Oh, absolutely, you have a dozen window managers, half
a dozen file managers (though none are particularly good), and you have lots
of smart people working to make it easy to take advantage of the far greater
control that Linux gives you.
But that's not what I'm talking about. That's 'hey, I wish I could...' not
'hey, I need to...'.
I wish I could change the ugly window decorations in Windows, but I don't need
to. I *do* need to switch everything over to a larger font because my eyes
aren't so good any more. I *do* need to have programs show up in the destktop
menu or the panel menu when I install them, or at least it has to be fairly
easy to do it myself.
In Windows, most of the time (not all of the time, but most of the time) your
applications and subsystems and components are pretty easy to configure and
customize. There's a lot more limitations in how much you can do, which seems
to be where you're coming from, but it's easy to do it.
That's not true, even in the "desktop environments" like Gnome, on Linux. Not
only does every application have its own configuration, which would be OK if
they did a good job of it, but all too many of them DON'T do a good job of it.
Liam Slider
> In article <bmpq2b$opvqp$1...@ID-169482.news.uni-berlin.de>,
> Liam Slider <li...@NOSPAM.liamslider.com> wrote:
>
>>Peter da Silva wrote:
>>
>>>In article <bmpm1n$q0vj8$2...@ID-169482.news.uni-berlin.de>,
>>>Liam Slider <li...@NOSPAM.liamslider.com> wrote:
>>>
>>>>Peter da Silva wrote:
>>>>
>>>>>They're doing it wrong. One of the big gripes I have with Windows is that
>>>>>Microsoft thinks a pretty face excuses a clumsy design. These people give
>>>>>us a pretty face, too, but they don't even do as good a job of being easy
>>>>>for people to personalise as Windows.
>
>
>>>>That's simply not true.
>
>
>>>I dearly wish you were right.
>
>
>>It's just as easy, or easier, using one of the modern desktop
>>envornments, to personalise Linux as it is to personalise Windows.
>
>
> Personalise Linux? Oh, absolutely, you have a dozen window managers, half
> a dozen file managers (though none are particularly good), and you have lots
> of smart people working to make it easy to take advantage of the far greater
> control that Linux gives you.
>
> But that's not what I'm talking about. That's 'hey, I wish I could...' not
> 'hey, I need to...'.
>
> I wish I could change the ugly window decorations in Windows, but I don't need
> to. I *do* need to switch everything over to a larger font because my eyes
> aren't so good any more.
That's pretty damn easy to do...
> I *do* need to have programs show up in the destktop
> menu or the panel menu when I install them, or at least it has to be fairly
> easy to do it myself.
That's also something that you can do in Linux too.
>
> In Windows, most of the time (not all of the time, but most of the time) your
> applications and subsystems and components are pretty easy to configure and
> customize. There's a lot more limitations in how much you can do, which seems
> to be where you're coming from, but it's easy to do it.
>
> That's not true, even in the "desktop environments" like Gnome, on Linux. Not
> only does every application have its own configuration, which would be OK if
> they did a good job of it, but all too many of them DON'T do a good job of it.
Bullshit.
Peter da Silva
Andrew Reilly <and...@gurney.reilly.home> wrote:
> Bollocks. I've never had to tweak MIME or XML to use ROX.
Then you haven't set to add a run action for a file extension that ROX
wasn't already expecting.
> If you want some non-default application to read a particular
> file, then you open the application and then drag the file into it.
Yes, I suppose you could. And for the applications (the majority of the
applications, in my experience) that don't support drag and drop you can
write a wrapper that accepts a file then launches it. But that's hardly
a huge step up from editing XML. Actually, editing XML is probably easier.
And, yes, the author of ROX considers that a bin full of little wrapper
scripts is a perfectly reasonable way of working with ROX.
So, since I'm up to doing that, I've got a bin full of wrapper scripts. But
I wouldn't recommend that for Uncle Arthur or Aunt Mabel. :)
> I must have missed your earlier message. Perhaps you're asking it to do
> things that it's not designed to do?
Ah yes.
See, that's the problem. There's a lot of applications designed to do
something that the author wants to do, the way the author wants to do it.
The problem comes when someone who doesn't work the way the author does and
doesn't think the way the author does starts using it. There's this big
cognitive gap, and the user gets frustrated, and you get the message that
started this thread. Especially after the guy has tried a bunch of tools
and hasn't found any of them that aren't written for people who think nothing
of dropping to a command line to do basic things.
I don't mind doing that all that much, but I don't believe it should be
necessary.
Peter da Silva
You could have a creator and type attached to the file, like Apple and Be
did. You could make the document a collection containing multiple files, like
NeXT bundles or Amiga filename/filename.info pairs (or potentially, NTFS
multi-fork files).
But so far we've always come back to extensions, because everything supports
them and nothing supports everyone else's schemes. Even Apple's given up the
good fight, except for legacy apps.
Peter da Silva
Absolutely. That's the same problem from the other side. Hence my suggestion
that people look at the way FreeBSD's sysinstall works... it's a good basis
for building user-friendly configuration tools in a way that doesn't break
the command line people.
> > One of the reasons I like working on Open Source software is that when
> > someone says "hey, it'd be nice if it did this" and I can see a good
> > way to make it DO that, easily, I can do it. For my day job just adding
> > a link to a webpage can require a week's worth of meetings.
> That sounds like the organisation is broken. I'm going to be totally
> arrogant and point out that's a people problem, and that one shouldn't
> try and solve that through technical means. Nevermind that you very
> often simply _cannot_ do that in the first place.
Great thoughts, friend, but you're missing the point. I'm not going to debate
the virtues of writing to contract versus rapid prototyping, because we
probably don't disagree, but this isn't the problem I'm talking about, this
is just brought up as a contrast to the open source world.
> > And that's why it really peeves me when there's a piece of open source
> > software that's hard where it doesn't need to be. And it *really* peeves
> > me when I write the code to fix it, and send it in, and the author comes
> > back with "yeh, I can see that's a useful feature, but I don't use it that
> > way so I won't put it in".
> There's the other side of the author not wanting to put it in because
> (s)he's deemed it unmaintainable.
It's possible in a few cases, but more often the response is a lot more
vehement than that. "That isn't the way you should be doing things" is
not that unusual. A lot of people write freeware because it's a bully
pulpit to promote their theories of how people and computers should interact
and solving problems or gaining mindshare for the project is less important
than gaining mindshare for the theory.
> > It means that there should be an obviously right way to use the tool for the
> > most common cases, and that obviously right way should be the way the user
> > is going to find when they start using the tool, and that right way should
> > work well with the other ways that more advanced users are going to use
> > the tool.
> Well said.
> Now, what about learning curve?
Depends on what you're learning... but there shouldn't be a high barrier to
doing basic things unless maybe that crossing that barrier provides a skill
that generalises beyond the application and makes other things easier later.
Learning to read, for example, is something that generalises beyond Dick and
Jane. If you needed to learn a new language every time you ran into a new
author, though, very few people would bother.
> C and D sound OK, don't know about A and B. E and F sound like the
> application isn't finished yet. But then again some people or companies
> have varied, interesting and ever changing notions of what user-friendly
> and/or easy is or ought to be.
Maybe it's not finished yet, but if it's been not-finished for long enough
and new features have come in without the basic stuff getting fixes, I think
we're dealing with "people shouldn't be doing that". Actually, in one case
I know we're dealing with "people shouldn't be doing that".
> OTOH, it all depends awfully much on file
> extentions. Can't we do better than that?
A bit of a digression, this. Because I've run into the same problem for
file types (MIME types). I'll start a new thread for that.
Peter da Silva
Liam Slider <li...@NOSPAM.liamslider.com> wrote:
> > I *do* need to switch everything over to a larger font because my eyes
> > aren't so good any more.
> That's pretty damn easy to do...
... and you do it here to cover apps using this toolkit, there for apps using
that toolkit, and Mozilla has a this or that theme, and these media players,
let's see, I think there's an OK theme for them here. Well, that one's kind
of ugly but it works, and that app's stuck but there's another one that does
pretty much the same thing...
> > I *do* need to have programs show up in the destktop
> > menu or the panel menu when I install them, or at least it has to be fairly
> > easy to do it myself.
> That's also something that you can do in Linux too.
There are apps that make it reasonably easy. Windowmaker for example. But I'm
not talking about "in Linux", I'm talking about "in Linux applications", and
it's not easy at all in a lot of them.
You can cherrypick the apps that have the configuration capabilities you want,
but you need to know what they are and how to do it, and learning that is also
not easy.
> Bullshit.
Ah, the crushing grip of sweet reason and clear self-evident logic.
jpd
> In article <bmrrad$smh$1...@news1.transmedia.de>,
> jpd <read_t...@do.not.spam.it> wrote:
>> OTOH, it all depends awfully much on file
>> extention. Can't we do better than that?
>
> You could have a creator and type attached to the file, like Apple and Be
> did. You could make the document a collection containing multiple files, like
> NeXT bundles or Amiga filename/filename.info pairs (or potentially, NTFS
> multi-fork files).
Or you could apply file(1) magic, if it were anywhere near reliable. Or
just for the parts where it's known reliable. Extentions aren't. And they
get abused as often as not (eg micros~1 grabbing .nfo for something silly).
Wonder what that'd do to performance.
>
> But so far we've always come back to extensions, because everything supports
> them and nothing supports everyone else's schemes. Even Apple's given up the
> good fight, except for legacy apps.
Somewhere I have a feeling this relates to the fallacious sticking to
proprietary software ``for compatabiliy'' I've so often been angry about.
Shmuel (Seymour J.) Metz
10/15/2003
at 09:04 PM, dam...@hotmail.com (Anonymous) said:
>We are here. On your doorstep, in your lab, in your inbox, invading
>your forums and mailing lists -- we are everywhere. We have been
>raped by The Redmond Giant and we are desperate. We NEED our
>computers, and we aren't going to go away.
Who asked you to? This is an advocacy group, and anyone exploring the
desirability of using Linux is welcome.
>You are the technicians, enthusiasts, hobbyists, and the gurus. Many
>of you got into computing as a refuge from the teeming mass of idiot
>humanity.
Alas, no. We are fully exposed to that teeming mass on the job. Never
forget that "Dilbert" is a nature study.
>You won't escape us easily.
Why would we want to, as long as you listen to the answers when given,
and don't believe that your time is more valuable than ours? You will
find lots of people willing to help you online, and more at meetings
of user groups.
>We are NOT willing to wade through your technical manuals, not at
>first. Far too many of us want it handed over, finished, on a silver
>platter.
That isn't going to happen. Not on Linux and not on windoze. But we
can help you to locate the proper manuals, and we can help clarify
things that are unclear or wrong.
>So we may try, and then once we've really gotten confused,
>we will invade your sacred sanctuaries and pester you.
And if you do it politely there will probably be someone willing and
able to help you.
>Since it seems most of the time we are rudely dismissed,
It happens, and sometimes for no good reason. But sometimes it's
because a poster was already given the answer and refused to believe
it, or because he refuses to do his homework. Do not be put off by a
response of RTMF as long as the poster tells you where TFM and what
section to look in.
>And we pollute the purity of the user pool,
Admittedly there are some posters with such an attitude, but I believe
that they are a small minority. Don't confuse those who don't welcome
neophytes with those who do not want harmful changes to the software.
>We will learn what we need to know for our
>own uses, and then we have work to do.
I'll let you in on a little secret; so do the rest of us. What we have
to know depends very much on what we are doing, but there is too much
software out there for everybody to know all of it.
>We don't think like you, can't work like you, so stop demanding it.
There are some things that come with the territory. If you have a
chauffer, you don't need to know the road signs. But if you're going
to drive, then you need to know quite a bit in order to not kill
yourself and to not be a hazard to others.
>If you as an individual don't want to help, we'll learn to leave you
>alone.
That's wise. I believe that most people will help if you ask nicely,
but will give you the backs of their hands if you demand help. But
you'd do well to do some research and to try to determine whose advice
is trustworthy and whose isn't.
>A few rare members of your elite cadre understand. They seem to have
>kept somehow an understanding of the human race.
You seem to have been taken in by the anti-intellectual prejudices of
the general culture. A lot of techies have much better people skills
than a lot of non-techies. I suggest that you ditch the idea, because
once you tell someone "I'm willing to tolerate your being a geek
because I need your assistance", you've provably alienated him for
good.
>They are the ones who write the "...for Stupids" books and get
>some of our money.
Maybe, and maybe not. I have my own suspicions.
>No we aren't going away.
Not should you.
>Can you imagine you need us?
Does it matter? If we help you, it won't be because we need you. It
will be because we're in a position to help you, you didn't ask in an
obnoxious way and you're willing to accept the help. The only payment
we ask is that when you finally learn enough, you help someone else a
little farther back on the same road you travelled.
--
Shmuel (Seymour J.) Metz, SysProg and JOAT
Unsolicited bulk E-mail will be subject to legal action. I reserve
the right to publicly post or ridicule any abusive E-mail.
Reply to domain Patriot dot net user shmuel+news to contact me. Do
not reply to spam...@library.lspace.org
Olathe
> In article <bmrrad$smh$1...@news1.transmedia.de>,
> jpd <read_t...@do.not.spam.it> wrote:
>
>>OTOH, it all depends awfully much on file
>>extention. Can't we do better than that?
>
>
> You could have a creator and type attached to the file, like Apple and Be
> did. You could make the document a collection containing multiple files, like
> NeXT bundles or Amiga filename/filename.info pairs (or potentially, NTFS
> multi-fork files).
BeOS used MIME-types (but not creators) fairly effectively. When I used
BeOS, I generally removed file extensions for aesthetic reasons (.mp3
doesn't look that nice). And, it also supported file extensions for
files that didn't have MIME-types stored.
File extensions could be tacked on when communicating with other types
of systems, as a sort of common language. That way, you'd get all the
benefit of MIME-types and you'd still be able to send and receive files.
> But so far we've always come back to extensions, because everything supports
> them and nothing supports everyone else's schemes. Even Apple's given up the
> good fight, except for legacy apps.
Web servers generally send the MIME-type of the file (and in several
BeOS web browsers, the browser would store the MIME-type of the file as
it stored the file), so I'd assume that MIME-types are a decent
direction for other OSs to move in as well, simply because it wouldn't
require rewriting communication protocols (HTTP with MIME-types is
fairly widespread).
How the OS stores the MIME-types can differ (forks, databases, index
files) without affecting intercommunication too much.
Peter da Silva
Olathe <Ola...@DALnet.irc> wrote:
> Web servers generally send the MIME-type of the file (and in several
> BeOS web browsers, the browser would store the MIME-type of the file as
> it stored the file), so I'd assume that MIME-types are a decent
> direction for other OSs to move in as well, simply because it wouldn't
> require rewriting communication protocols (HTTP with MIME-types is
> fairly widespread).
OTOH, you don't generally use HTTP for general passing files around, THAT
happens at the file system level over AFS or NFS or CIFS or Apple AFS.
In a heterogenous environment your options are to have the network file
server aware of all the operating systems and present the files in the
right way for each OS, which is just not going to happen, or abandon
file sharing and go back to things like mailing 200M powerpoints to the
guy in the next cubicle.
> How the OS stores the MIME-types can differ (forks, databases, index
> files) without affecting intercommunication too much.
The majority of the intercommunication where this matters, how the OS
stores the mime types is *precisely* how the OS in the next cubicle sees
them.
Olathe
> In article <vu82r6d...@corp.supernews.com>,
> Olathe <Ola...@DALnet.irc> wrote:
>
>>Web servers generally send the MIME-type of the file (and in several
>>BeOS web browsers, the browser would store the MIME-type of the file as
>>it stored the file), so I'd assume that MIME-types are a decent
>>direction for other OSs to move in as well, simply because it wouldn't
>>require rewriting communication protocols (HTTP with MIME-types is
>>fairly widespread).
>
>
> OTOH, you don't generally use HTTP for general passing files around, THAT
> happens at the file system level over AFS or NFS or CIFS or Apple AFS.
If those don't do what you want, perhaps something new is needed.
Nothing says we have to keep using old file transfer protocols and
nothing says we can't upgrade them.
> In a heterogenous environment your options are to have the network file
> server aware of all the operating systems and present the files in the
> right way for each OS, which is just not going to happen, or abandon
> file sharing and go back to things like mailing 200M powerpoints to the
> guy in the next cubicle.
Sorry, that's incorrect. All you need is for the network server to
include the MIME-type as metadata and have the *clients* do the
translation for the local OS. It's a bit like how the presentation
layer for, say, FTP text file transfer works.
>>How the OS stores the MIME-types can differ (forks, databases, index
>>files) without affecting intercommunication too much.
>
>
> The majority of the intercommunication where this matters, how the OS
> stores the mime types is *precisely* how the OS in the next cubicle sees
> them.
Then, whatever server you use will know what OS it's running on, be able
to get the type information, convert it to a standard format, and let
the other side convert it from the standard format to the local format.
It's all about metadata.
P.S. Learn from the Internet protocol or the FTP protocol or Java;
computers on token ring networks can't talk directly to computers on
Ethernet networks...text files are represented differently on different
OSs (line breaks being a big difference)...computer processors use
different machine languages. Yet, those examples bridge the gaps.
Peter da Silva
Olathe <Ola...@DALnet.irc> wrote:
> If those don't do what you want, perhaps something new is needed.
> Nothing says we have to keep using old file transfer protocols and
> nothing says we can't upgrade them.
I would love to have a widely implemented network file system protocol
that provided OS-independant representation and automatic marshalling
of arbitrary metadata and multiple data streams. Heck, I'd be ecstatic
to have a network file system protocol that even handled OS-specific
metadata between equivalent operating systems the way OpenNET did back
on the '80s. I mean, nothing says that you have to keep using the old file
transfer protocols, right? Nothing says we can't upgrade to a network
file system that only sucks sometimes, right?
Well, no, what says we have to keep using the old protocols is that we have
zillions of dollars sunk in infrastructure that uses them, and any new
operating system has to interoperate with a primarily metadata-free world
and do it *well*. Which means that even if you have great local support
for rich database files, every application and utility still has to support
files that don't have much more than a name, size, and a single
undifferentiated stream of bytes.
> > In a heterogenous environment your options are to have the network file
> > server aware of all the operating systems and present the files in the
> > right way for each OS, which is just not going to happen, or abandon
> > file sharing and go back to things like mailing 200M powerpoints to the
> > guy in the next cubicle.
> Sorry, that's incorrect. All you need is for the network server to
> include the MIME-type as metadata and have the *clients* do the
> translation for the local OS.
You talk to Netapp, I'll see what Quantum and EMC have to say, and Joe over
there can brace Microsoft.
> > The majority of the intercommunication where this matters, how the OS
> > stores the mime types is *precisely* how the OS in the next cubicle sees
> > them.
> Then, whatever server you use will know what OS it's running on, be able
> to get the type information, convert it to a standard format, and let
> the other side convert it from the standard format to the local format.
I already have "whatever server I use", and I've had to fight for 18 months
to get one that supports both NFS and CIFS adequately. I can't begin to
imagine what developing a business case for one to support an OS we're not
even currently using would require.
> It's all about metadata.
It's all about money. It's always all about money.
> P.S. Learn from the Internet protocol or the FTP protocol or Java;
> computers on token ring networks can't talk directly to computers on
> Ethernet networks...text files are represented differently on different
> OSs (line breaks being a big difference)...computer processors use
> different machine languages. Yet, those examples bridge the gaps.
I suggest you study FTP and Java a bit better.
FTP was developed back when there were a bunch of different EXISTING OSes
that had to be supported, and it still doesn't handle anything but stream
text files: it breaks for 36-bit binary data, or variant record style text
like RSX-11 and VMS use.
Java, on the other hand, is a really bad card for you to play. It encapsulates
the application in a pseudocode for a virtual machine. This means your Java
app interacts with humans well, but it doesn't interoperate with native apps
except in the roughest way... or if it uses native APIs then it loses all the
portability you're looking for.
As for the internet protocol, it provides a high level of interoperability
by remaining strictly agnostic about the semantics of what's being carried
over it. It's similar to Java, except at the other end of the stack.
Your hypothetical network file system client could behave like IP or Java
encapsulating the rich database file in a package that is treated as an
opaque object by the remote system, but then you'd still need to have every
applications know how to deal with foreign files or you couldn't share
your data with anyone using a forign OS.
Alternatively, it could act like FTP, and have dozens of clever heuristics
to try and convert the local data into something the network software could
deal with, and then reconstruct the discarded metadata and present that to
the application. It couldn't just save the metadata off and reapply it to
the modified file, because it doesn't know that the app that was doing the
mods updated it appropriately.
In practice, though, you'll end up with two sets of code in every app. One
to deal with local files well, and one to deal with remote files well. You
will end up with two communities of apps... one, like the classic MacOS
software, that takes full advantage of the rich file system and doesn't
play well with foreign files... and another set used in mixed environments
that mostly ignores the extra data except when it absolutely needs it.
And Apple, as I noted, is currently in transition from the first environment
to the second.
Thomas Mueller
> FTP was developed back when there were a bunch of different EXISTING OSes
> that had to be supported, and it still doesn't handle anything but stream
> text files: it breaks for 36-bit binary data, or variant record style text
> like RSX-11 and VMS use.
Are you sure FTP handles only stream text files? It seems to me FTP can
handle binary files too, 8-bit data, but you have to tell the server to go
into binary mode, otherwise the server will send a binary file in ASCII mode,
meaning garbled data. I have done a lot of binary FTP downloads.