How to destroy a system
bobu...@yahoo.com
ways to destroy a system. The first way is to add feature after feature
after feature, and before you know it you have a 500 pages manual (how
many pages are there in MzScheme manual?). When I heard this I nodded
because I agree with this.
However when I heard the second way to destroy a system I couldn't
believe my ears. Because that other way was to add a mouse and windows
and stuff, in other words a GUI. Becuase Sussman said that all the
processor activity that the GUI required from the system would make it
slow, and if you happened to touch the mouse accidentally with your
elbow, then smoke would come out of your computer. Now I've used mouse
for many years and not once did any smoke come out of my computer. The
reason that many people like GUI is that their brain is not just good
for symbolics but also for images and kinetics. If GUI was such a bad
idea why has it taken such a big market. I think GUI is good becuase as
I see it the computers should adapt to people and not the other way
around.
Well, Graham Bell thought that the primary use of the telephone was to
hear opera from a long distance and Edison thought that the primary use
of the phonograph was to send audio messages with post instead of
writing letters. Had they exchanged their idea of use, they would have
been nearer to the truth. When Lumière Brothers, the creators of the
cinematographe, (a three-in-one motion picture camera, developer, and
projector) sold their invention to the stage magician George Melies
they warned him that there was no future for their invention. Luckily
Melies didn't take their advice and made many famous fils of which the
trip to the moon is best known. This proves that inventors not always
know what their invention is good for.
Bob
Pascal Bourguignon
> However when I heard the second way to destroy a system I couldn't
> believe my ears. Because that other way was to add a mouse and windows
> and stuff, in other words a GUI.
Indeed.
> Because Sussman said that all the
> processor activity that the GUI required from the system would make it
> slow, and if you happened to touch the mouse accidentally with your
> elbow, then smoke would come out of your computer.
It's half an old-timer joke. 20 years ago, moving the mouse could
easily activate the swap.
> Now I've used mouse
> for many years and not once did any smoke come out of my computer. The
> reason that many people like GUI is that their brain is not just good
> for symbolics but also for images and kinetics.
Monkeys and Neanderthals are good with images and kinetics too.
IMO, the more advanced, higher level functions of the brain are symbolic.
Anyways, with GUI you see people working three hours making
substitutions in "word processors" with copy-and-paste, or renaming
files in directories, or finding source and destination folder
windows to move files between them, all actions that can be done in 30
seconds of shell commands.
> If GUI was such a bad
> idea why has it taken such a big market. I think GUI is good becuase as
> I see it the computers should adapt to people and not the other way
> around.
What for is it good? GUI is good to sell computers to 90% of the
population. But it's no good to improve their productivity.
> Well, Graham Bell thought that the primary use of the telephone was to
> hear opera from a long distance and Edison thought that the primary use
> of the phonograph was to send audio messages with post instead of
> writing letters. Had they exchanged their idea of use, they would have
> been nearer to the truth. When Lumière Brothers, the creators of the
> cinematographe, (a three-in-one motion picture camera, developer, and
> projector) sold their invention to the stage magician George Melies
> they warned him that there was no future for their invention. Luckily
> Melies didn't take their advice and made many famous films of which the
> trip to the moon is best known. This proves that inventors not always
> know what their invention is good for.
Funny that you invoke Graham Bell. Nowadays, the primary use of the
telephone wire is to transmit MP3. All right, it's more RAP than
Opera, but Graham Bell was perfectly right. If you want to speak to
someone, you use GSM, not ATT.
--
__Pascal Bourguignon__ http://www.informatimago.com/
Kitty like plastic.
Confuses for litter box.
Don't leave tarp around.
bobu...@yahoo.com
> What for is it good? GUI is good to sell computers to 90% of the
> population. But it's no good to improve their productivity.
>
Thanks Pascal for opening up my eyes. Those poor suckers with
Photoshop, Powerpoint, Excel, Quicktime and so on. They don't know
what's good for them. But we the smart 10% raise ourselves above
Monkeys and Neanderthals. Because we use higher brain functions ;-)
Pascal Bourguignon
I intended to add this quotation:
Re graphics: A picture is worth 10K words -- but only those to
describe the picture. Hardly any sets of 10K words can be adequately
described with pictures.
As an exercice, take any medium sized program, and convert it into
diagrams and organigrams. Then give the graphics to a fellow
programmer and ask him to add a feature or remove a bug.
If your job is to draw a picture, or edit a photography, then
Photoshop can be nice. But if you mean to draw a diagram, MacDraw is
already better, and if you want to draw a graph, then dot is even
better. The input of dot is not gesticulations, it's a text file.
And even in the case of Photoshop, for some tasks gimp is superior
because you can program the graphical manipulations in scheme.
As for spreadsheets, there's a lot of material and serrious studies
explaining why they're bad; for example:
http://smallbusinessceo.blogspot.com/2005/12/spreadsheet-is-dead.html
the main problem being that they're in the hand of non-programmer, who
don't apply needed software engineering techniques.
As for PowerPoint, I'll once quote the 1,500,000 google hits for:
powerpoint disaster.
Quicktime, well of course, it's better to watch movies on graphic
screens. Nonetheless, try: telnet://towel.blinkenlights.nl ;-)
--
__Pascal Bourguignon__ http://www.informatimago.com/
CONSUMER NOTICE: Because of the "uncertainty principle," it is
impossible for the consumer to simultaneously know both the precise
location and velocity of this product.
bobu...@yahoo.com
since I don't know when. My wife came in and asked me if I was watching
some comedy on my computer. You have an answer for everything. It's
like those religious people coming to my door trying to convince me to
believe in God. Whatever I say they have an answer. We'll have to
accept that we have different world view and value different kind of
things. When a new paradigm comes a long the people living in the old
paradigm do not get convinced. The new paradigm (if it's any good) wins
because the old paradigmers die. That's why it usually take 25 to 50
years to a new paradigm to establish itself.
I notice that some subjects drive some people in defensive position and
nothing productive comes out of such discussion. So I will not discuss
GUI questions more.
Pascal Bourguignon
Ah, but don't misunderstand me. I was amongst the first to buy a
Macintosh. I got mine on June 25th 1984, one of the firsts Macintosh
sold in France. I loved the Mac GUI, and loved even more NeXTSTEP
GUI.
Nowdays I still use Linux with X11, and I've got more than twenty
windows in my screen. But they all contain text. The only windows
that don't contain text are the occasionnal mplayer window with the
SICP videos, erm... well, there's still a lot of text on Abelson &
Sussman's blackbox and edwin videoterminal...
--
__Pascal Bourguignon__ http://www.informatimago.com/
This is a signature virus. Add me to your signature and help me to live.
Anton van Straaten
> When a new paradigm comes a long the people living in the old
> paradigm do not get convinced. The new paradigm (if it's any good) wins
> because the old paradigmers die.
If GUIs are so great, why are you busy teaching yourself a very powerful
symbolic language with which to communicate with the computer?
A wise man named Bear once wrote, in this very forum:
"If you're stuck at the level of wanting to point at a picture and
click, you've got to understand that you're basically at the same
level as a preverbal child..."
It's not obscure enough to qualify as a koan, but it can be as
enlightening, if you contemplate it a little.
> That's why it usually take 25 to 50
> years to a new paradigm to establish itself.
If you go back 50 years, computers had barely been invented, and Lisp
was just about to be invented. Taking a historical perspective on the
history of computing, you realize we've hardly started. That's also
pretty clear when you examine the current state of GUIs, and software
written for GUIs. In many respects, they're mind-blowingly primitive
and limiting.
What works in favor of GUIs from an adoption perspective is that they're
flashy, to appeal to our chimp-derived hindbrains; and they can simulate
the preferred I/O mode of our preverbal inner child, allowing computers
to be used at a basic level by someone with minimal training, and
minimal interest in being trained.
There's a long way still to go there, so much so that I think it's
pretty safe to say that we haven't actually figured out the paradigm
yet, we're just experimenting and finding the limits of what we've done
so far. GUIs will no more replace symbolic command interfaces than
McDonald's hamburgers will replace steak.
Anton
Thant Tessman
> bobu...@yahoo.com wrote:
>
>> When a new paradigm comes a long the people living in the old
>> paradigm do not get convinced. The new paradigm (if it's any good) wins
>> because the old paradigmers die.
>
>
> If GUIs are so great, why are you busy teaching yourself a very powerful
This thread is so full of shit. Text interfaces are simply easier to get
working. That's all. Think about it. A mathematician doesn't write
formulae as a linear string of characters. Engineers (unless they're
programmers) don't type to solve problems, they draw diagrams. And even
programmers occasionally use notation that doesn't look all that purdy
once it's shoved through a qwerty keyboard. 3D modelers and animators
don't program in the typing on a keyboard sense. Or when they do, it's
because that's an aspect of their task that the designer of the modeling
application could slough off.
Once upon a time people did actually put some thought into how to
accomplish complex tasks (including programming) using visual
abstractions and metaphora. Unfortunately this was back when the
computers weren't quite up to the task. Consequently, the computer and
the text interface evolved together. Consequently, now that computers
*are* up to the task, visual interfaces usually just choose abstractions
to fit what's already in place, or provide the easy stuff like buttons
and sliders and text fields and folders which has the annoying side
effect of making every application look like a control panel.
> What works in favor of GUIs from an adoption perspective is that they're
> flashy, to appeal to our chimp-derived hindbrains; and they can simulate
> the preferred I/O mode of our preverbal inner child, allowing computers
> to be used at a basic level by someone with minimal training, and
> minimal interest in being trained.
This is the same "I've suffered for my craft--now it's your turn"
elitist bullshit I hear too often from C++ experts.
[...]
-thant
--
"If they can get you asking the wrong questions, they don't have to worry
about answers." -- Thomas Pynchon
Thomas Hafner
> bobu...@yahoo.com writes:
> > Well, Graham Bell thought that the primary use of the telephone was to
> > hear opera from a long distance and Edison thought that the primary use
> > of the phonograph was to send audio messages with post instead of
> > writing letters. [...]
>
> Funny that you invoke Graham Bell.
Maybe bobueland supposes that Graham was the inventor of the
telephone? (Antonio Meucci was it, and even Johann Philip Reis did
reinvent it before Graham Bell).
Thomas
Anton van Straaten
> Anton van Straaten wrote:
>
>> bobu...@yahoo.com wrote:
>>
>>> When a new paradigm comes a long the people living in the old
>>> paradigm do not get convinced. The new paradigm (if it's any good) wins
>>> because the old paradigmers die.
>>
>>
>>
>> If GUIs are so great, why are you busy teaching yourself a very
>> powerful symbolic language with which to communicate with the
>> computer? [...]
>
>
> This thread is so full of shit. Text interfaces are simply easier to get
> working. That's all. Think about it. A mathematician doesn't write
> formulae as a linear string of characters. Engineers (unless they're
> programmers) don't type to solve problems, they draw diagrams. And even
> programmers occasionally use notation that doesn't look all that purdy
> once it's shoved through a qwerty keyboard.
You've misunderstood me, but it's my fault for not having laid out a
more comprehensive position.
I'm saying that *current* GUIs and GUI applications leave a lot to be
desired. Although they've effectively replaced command line interfaces
for the average user, they haven't subsumed all the things that command
line interfaces and textual representations tend to do better, today,
which leaves a gap in what most GUI applications are capable of, or good at.
If GUIs do succeed in subsuming the best of command line interfaces, I
suspect that'll be at least partly by providing better support for
inputting arbitrary command combinations within the GUI. At that point
there'll be much less basis for arguing "which is better" (Houston or
Los Angeles?), except possibly on the basis of resource consumption
(which brings us back to the start of the thread).
> 3D modelers and animators
> don't program in the typing on a keyboard sense. Or when they do, it's
> because that's an aspect of their task that the designer of the modeling
> application could slough off.
For applications where the end product is very visual and non-symbolic,
that makes sense. However, you only have to look at the state of GUI
spreadsheets and word processors to see examples where a great deal of
usability has been sacrificed in the name of making it easy to do the
easy things with just a point and a click, while making it sometimes
very hard to do only slightly more advanced things.
> Once upon a time people did actually put some thought into how to
> accomplish complex tasks (including programming) using visual
> abstractions and metaphora. Unfortunately this was back when the
> computers weren't quite up to the task. Consequently, the computer and
> the text interface evolved together. Consequently, now that computers
> *are* up to the task, visual interfaces usually just choose abstractions
> to fit what's already in place, or provide the easy stuff like buttons
> and sliders and text fields and folders which has the annoying side
> effect of making every application look like a control panel.
You're supporting my point - current GUIs have a lot of room for
improvement.
>> What works in favor of GUIs from an adoption perspective is that
>> they're flashy, to appeal to our chimp-derived hindbrains; and they
>> can simulate the preferred I/O mode of our preverbal inner child,
>> allowing computers to be used at a basic level by someone with minimal
>> training, and minimal interest in being trained.
>
>
> This is the same "I've suffered for my craft--now it's your turn"
> elitist bullshit I hear too often from C++ experts.
Not at all. In fact, it's the opposite of elitist: my position is that
the companies that are responsible for the current GUIs and GUI apps
have done a great disservice to a large number of users by not providing
them with richer user interfaces, that scale beyond just pointing and
clicking.
The comment about minimal training or interest in being trained is not
elitist, it's realistic. The market forces which result from that
situation have led Windows to become almost kiosk-like in its interface,
which seems like a wonderful thing for TCO in large organizations.
However, that cost calculation doesn't consider the productivity lost
when people fight with applications that won't let them do anything
outside the confines of a series of dialog boxes, nor does it consider
the cost of consultants needed to engineer solutions to work around
those limitations, or the effect on the collective intelligence and
spirit of the human race. If you give any credence to even the weak
form of the Sapir-Whorf theory, ask yourself what current GUIs are doing
to people's thinking with respect to computers.
Anton
Anton van Straaten
> The comment about minimal training or interest in being trained is not
> elitist, it's realistic. The market forces which result from that
> situation have led Windows to become almost kiosk-like in its interface,
> which seems like a wonderful thing for TCO in large organizations.
TCO being Total Cost of Ownership, not Tail Call Optimization. :)
Ray Dillinger
> bobu...@yahoo.com wrote:
>
>> When a new paradigm comes a long the people living in the old
>> paradigm do not get convinced. The new paradigm (if it's any good) wins
>> because the old paradigmers die.
>
>
> If GUIs are so great, why are you busy teaching yourself a very powerful
> symbolic language with which to communicate with the computer?
>
> A wise man named Bear once wrote, in this very forum:
>
> "If you're stuck at the level of wanting to point at a picture and
> click, you've got to understand that you're basically at the same
> level as a preverbal child..."
>
> It's not obscure enough to qualify as a koan, but it can be as
> enlightening, if you contemplate it a little.
There is a distinction between language and protolanguage
at the center of my thoughts on the matter.
In a protolanguage, you have symbols for lots of things, like
eating, and mice, and cheese. And if you throw all these
symbols together, in any order, people reach the obvious
conclusion -- that a mouse ate some cheese. This is usually
okay as long as you have nouns and verbs, although you're
already running into trouble as you have no clear notion of
which noun is the subject and which the object. (okay, with a
mouse and cheese, and the verb "eat," I think we can make an
assumption... but things are not always so clear, and we can't
communicate the surprising case if we happen to run across some
kind of bizarre mouse-eating cheese).
Most people (preverbal children) start learning languages on
the basis of protolanguage. Most adults learning a new language
foreign to them start out with protolanguage. The additon of
structure (ie, grammar) creates language from protolanguage.
Language allows more nuanced communication: we know which noun
is the subject and which the object, and we can add adjectives
like "heavy" or "old" with some confidence of them being
attached to the noun for which they were intended.
The way most people use a GUI, it's a protolanguage at best.
We can specify a noun, and we hope the system infers the
proper verb. Usually we have to specify a different noun
for the object of the sentence (or the subject, if we grabbed
the object first). And adjectives, for the most part, are beyond
us. We're operating the way preverbal children operate, pointing
at the things on the dinner table that we want and attracting the
attention of an adult with opposable thumbs and better reach, and
hoping that that person (subject) figures out that we want them
to pass (verb) the desired items (objects) to us (indirect objects).
Without linguistic advance, we cannot communicate the surprising
case. The preverbal child points at the milk, we point at a
word processing document. The typical case (give the child the
milk, open the document in its accustomed editor) usually results.
But the surprising case (where what the child actually wanted was
for the milk to be poured over the pumpkin pie, or we wanted to
use a hex editor instead of our word processor) requires language.
If we want, for example, to open up a word-processor document
in a hex editor rather than its accustomed editor, we have to
invent syntax (in most GUI's now available as drag-n-drop specifies
a verb to go with a particular object). The child who wants milk
poured over the pumpkin pie must also learn language, as such an
arbitrary request is unlikely to be inferred by others from any
prelinguistic behavior.
My point is that the "language" of GUI's, while now admitting of
a subject-verb structure via drag-n-drop and a question form
(still pretty ragged and not universally implemented) via the
right mouse button, is advancing very very slowly. By contrast,
the "language" of most shell command systems is already quite
advanced. In the shell I have grouping (globbing) and adverbs
(command-line switches) and locatives (redirections) and
adjectives(globbing plus filename discipline) already, and
they're convenient to use. I also can structure and sequence
commands, creating a hierarchy of verbs via pipe commands.
These things, and means to express these things, simply don't
exist in GUI's yet. I cannot say the things via a GUI that I can
say via the shells, because the linguistic structure of GUIs
is not yet as far advanced.
And as soon as the language of GUI's is sufficiently flexible
and general to express such thoughts, it will have also aquired
the arbitrary character and learning curve typical of spoken
or typed shell languages. If you have ever watched two people
having a conversation via sign language, you have seen an
advanced GUI language in action, and I'm willing to bet that
unless you also have put as much effort into learning that same
language as would be required to learn any foreign language, you
understood it not at all.
Bear
Tom Lord
referencing more explicitly that the current fascination with
current GUI technology is almost entirely driven by systems
of economic oppression. We are looking, my friends, at
nothing more or less than the collective output of a few 100K
"day jobs" and "heads down and do what I'm told till I get where
I'd rather be". More or less, anyway.
Today I used a GUI to query a database for a company I work
for as part of a data mining exercise. The GUI comes from a
multi-million dollar company.
I shit thee not: it has buttons to add left and right parentheses
around SQL expressions that you select from menus. You don't
get the full range of SQL mind you -- and, no, the schema aren't
documented. But you get to form expressions with handy gestures
like "click on the left paren; pick the field from a menu; pick a
relational operator from a menu; pick the value to compare to
etc......" You should see what fun is involved in nested queries.
I've been doing a certain amount of tallying things up on paper,
by hand. Big productivity gains, there.
(People say that when free software emulates the functionality
of crap like this it sucks the value out of markets. Ok. Then
they say this is bad because it robs of incentive for further
"innovation". I still say: "ok.")
-t
Brian Harvey
> If you have ever watched two people
>having a conversation via sign language, you have seen an
>advanced GUI language in action, and I'm willing to bet that
>unless you also have put as much effort into learning that same
>language as would be required to learn any foreign language, you
>understood it not at all.
This (the whole message) is very interesting and thought-provoking, but
I'm not sure it's entirely true, especially the claim quoted here.
What makes a GUI a GUI isn't just that it's gestural and pictorial -- nobody
suggests that written English is a GUI, for example. Maybe a better example
is that Grafitti on the Palm Pilot isn't a GUI; it's a keyboard replacement.
(And these days they've, alas, replaced it with the hideous thumb keyboard.)
The GUI is full of carefully designed and empirically quick-to-learn central
*metaphors*. The same is not true of American Sign Language and similar
languages in other parts of the world; they are content-free languages, ready
for any metaphor or none, like English. GUIs have the Desktop, the
black-on-white-just-like-paper windows, all that.
One of the things I like about my PowerBook is that MacOS X gives me access
to both worlds; when I'm doing the same things my mom or my kid does with a
computer, I can do it as easily as they can; but when I want to do something
harder, I can open an xterm.
We'll have an "advanced GUI language" when the Mac-like features and the
Unix-like features of the current Macs are better integrated. Right now it's
like two separate worlds inhabiting the same box. My favorite one-sentence
example is that the Unix cp command in MacOS does not copy a Mac file with
all its associated information, so the copy is not recognized by the GUI part
of the world as an application or whatever it's supposed to be. (There is a
special "CpMac" command that does the right thing, but it's a secret, and it's
not in your path unless you work at it.) And when you call up the AppleCare
number, they say "We don't support anything you do in a shell window." :-(
And also when Programming By Example is really bulletproof and flexible.
bobu...@yahoo.com
> If GUIs are so great, why are you busy teaching yourself a very powerful
> symbolic language with which to communicate with the computer?
I love symbolics, but I also love images and kinetics. I see no
conflict there. I also love, Mac and Unix and Windows. I don't see any
conflict there as well. The new Macs built on Intel processor will
support all three platforms. That I have been wanting for a long time
:-)
For me GUI is not the end of a story but a beginning. I think that
kinetics will evolve from the mouse (the only kinetics we have today),
to multi-dimensional worlds in which we can move and interact with and
get tactile feedback from. I really don't understand this talk about if
you love A you must hate B. But I accept that people think differently
and have different views and that my view is not necessarily the best
one for everyone; in fact nobodys view is best for everyone.
Bob
Geoffrey Summerhayes
<bobu...@yahoo.com> wrote in message news:1138298318.5...@g44g2000cwa.googlegroups.com...
> Pascal, Pascal, Pascal. Thank you, I haven't had such a good laughs
> since I don't know when. My wife came in and asked me if I was watching
> some comedy on my computer. You have an answer for everything. It's
> like those religious people coming to my door trying to convince me to
> believe in God. Whatever I say they have an answer. We'll have to
> accept that we have different world view and value different kind of
> things. When a new paradigm comes a long the people living in the old
> paradigm do not get convinced. The new paradigm (if it's any good) wins
> because the old paradigmers die. That's why it usually take 25 to 50
> years to a new paradigm to establish itself.
There are two kinds of fool;
One who says "This is old and therefore bad" and
One who says "This is new and therefore better."
That pretty much sums up this thread.
Both interfaces have their place, ignore one,
work harder. Me, I'm a unbelievably lazy slob,
I use both.
--
Geoff
Pascal Bourguignon
> [...] By contrast,
> the "language" of most shell command systems is already quite
> advanced. In the shell I have grouping (globbing) and adverbs
> (command-line switches) and locatives (redirections) and
> adjectives(globbing plus filename discipline) already, and
> they're convenient to use. I also can structure and sequence
> commands, creating a hierarchy of verbs via pipe commands.
> These things, and means to express these things, simply don't
> exist in GUI's yet. I cannot say the things via a GUI that I can
> say via the shells, because the linguistic structure of GUIs
> is not yet as far advanced.
I'd like to mention here, even if it seems like it's a forgotten hope,
that we could implement NLP CLI shells, with a tad of AI, which could
be much more user friendly than any GUI will ever be. And if you add
voice recognition, you still have a CLI :-)
> [...] If you have ever watched two people
> having a conversation via sign language, you have seen an
> advanced GUI language in action, [...]
Not really. It's symbolic in nature, as much as text.
--
__Pascal Bourguignon__ http://www.informatimago.com/
IMPORTANT NOTICE TO PURCHASERS: The entire physical universe,
including this product, may one day collapse back into an
infinitesimally small space. Should another universe subsequently
re-emerge, the existence of this product in that universe cannot be
guaranteed.
bobu...@yahoo.com
> There are two kinds of fool;
> One who says "This is old and therefore bad" and
> One who says "This is new and therefore better."
I think this refers to the line
>>The new paradigm (if it's any good) wins
It seems to me that the line above says that
"sometimes a new thing is better than the old thing (and if so it will
win, or at least it ought to win)"
It may look similar to
"This is new and therefore better."
but there's a huge difference. The first is an obvious truth and the
latter is an obvious lie.
Anton van Straaten
> In a protolanguage, you have symbols for lots of things, like
> eating, and mice, and cheese. And if you throw all these
> symbols together, in any order, people reach the obvious
> conclusion -- that a mouse ate some cheese. This is usually
> okay as long as you have nouns and verbs, although you're
> already running into trouble as you have no clear notion of
> which noun is the subject and which the object. (okay, with a
> mouse and cheese, and the verb "eat," I think we can make an
> assumption... but things are not always so clear, and we can't
> communicate the surprising case if we happen to run across some
> kind of bizarre mouse-eating cheese).
Eating, mice, cheese... You're obviously talking about grilled mice
with cheese!
Michael Campbell
> <bobu...@yahoo.com> wrote in message news:1138298318.5...@g44g2000cwa.googlegroups.com...
> There are two kinds of fool;
> One who says "This is old and therefore bad" and
> One who says "This is new and therefore better."
I thought it was "This is old and therefore /good/", no? Otherwise
both lines are describing the same kind of fool.
--
Michael Campbell
Geoffrey Summerhayes
"Michael Campbell" <michael....@gmail.com> wrote in message news:uirs5z...@atlmicampbell.checkfree.com...
Yeah. Serves me right for cutting-and-pasting from the web.
--
Geoff
Ray Dillinger
>> If you have ever watched two people
>>having a conversation via sign language, you have seen an
>>advanced GUI language in action, and I'm willing to bet that
>>unless you also have put as much effort into learning that same
>>language as would be required to learn any foreign language, you
>>understood it not at all.
Brian Harvey wrote:
> This (the whole message) is very interesting and thought-provoking, but
> I'm not sure it's entirely true, especially the claim quoted here.
>
> What makes a GUI a GUI isn't just that it's gestural and pictorial -- nobody
> suggests that written English is a GUI, for example. Maybe a better example
> is that Grafitti on the Palm Pilot isn't a GUI; it's a keyboard replacement.
> (And these days they've, alas, replaced it with the hideous thumb keyboard.)
> The GUI is full of carefully designed and empirically quick-to-learn central
> *metaphors*. The same is not true of American Sign Language and similar
> languages in other parts of the world; they are content-free languages, ready
> for any metaphor or none, like English.
Pascal Bourguignon wrote:
> Not really. It's symbolic in nature, as much as text.
In a way, you've both hit upon the crux of the matter. It's that
selfsame symbolic quality - the ability to talk about things not
present or say things that no one thought to prompt you or give you
the opportunity to say - in fact, the ability to talk about arbitrary
things - that gives more advanced languages their power.
The fact that language uses symbols instead of manipulating metaphors
opens up the world of discourse to any symbol known to the speaker,
rather than any metaphor present in the virtual environment. The leap
forward in expressiveness is the same as the leap forward from being
able to talk only about things you can point at to being able to talk
about anything you've got an agreed-upon symbol for.
Bear
Brian Harvey
>The fact that language uses symbols instead of manipulating metaphors
>opens up the world of discourse to any symbol known to the speaker,
>rather than any metaphor present in the virtual environment. The leap
>forward in expressiveness is the same as the leap forward from being
>able to talk only about things you can point at to being able to talk
>about anything you've got an agreed-upon symbol for.
It's not quite fair to measure GUIs against the standard of natural languages.
The relevant standard is programming languages. And in that world, the great
leap is the one wherein you can define your own procedures. After all, the
small number of primitives in (R5RS) Scheme is comparable to the small number
of icons or menu items you can put on a screen.
The typical current GUI doesn't let you define procedures, although there are
partial exceptions that allow macros of a sort. But there are visual
dataflow languages that let you build procedures by putting together virtual
jigsaw puzzle pieces; a new one out just this week is
http://education.mit.edu/starlogo-tng/
I have to say that I find such things harder to use than regular old text
based languages, but maybe novices will find them less intimidating, and maybe
GUIs will evolve that way.
But also, between current GUIs and full-scale programming languages there is a
huge middle ground into which GUIs could plausibly and usefully expand. I'd
be really happy if my non-Unix desktops had the power of pipes in the Unix
shell, along with its collection of useful little modules like grep, join,
uniq, and so on.
Ray Dillinger
> It's not quite fair to measure GUIs against the standard of natural languages.
Well... that's the standard I measure CLI's against...
> The typical current GUI doesn't let you define procedures, although there are
> partial exceptions that allow macros of a sort. But there are visual
> dataflow languages that let you build procedures by putting together virtual
> jigsaw puzzle pieces; a new one out just this week is
> http://education.mit.edu/starlogo-tng/
Interesting, but not fundamentally new, I don't think.
There've been visual-program tools around for a long
time, starting with flowcharts. This is better than
average, but it's tough to come up with actual new ground
here.
> I have to say that I find such things harder to use than regular old text
> based languages, but maybe novices will find them less intimidating, and maybe
> GUIs will evolve that way.
I want to point something out; as soon as you get to a decent
level of expressiveness, the visual language has basically the
same level of complexity and structure (and "arbitrary" or
"unintuitive" symbols) in it as does the equivalent written
command language. And (this is a guess since I haven't invested
the effort myself) probably the same kind of learning curve.
> But also, between current GUIs and full-scale programming languages there is a
> huge middle ground into which GUIs could plausibly and usefully expand.
Yup. Much of it already covered usefully by CLI's. In
fact, the line between CLI's and full-scale programming
languages can be pretty fuzzy, as languages like awk and
perl demonstrate.
> I'd
> be really happy if my non-Unix desktops had the power of pipes in the Unix
> shell, along with its collection of useful little modules like grep, join,
> uniq, and so on.
Yeah... You could invent the widgets to use for desktop
icons and draw lines from inputs of one to outputs of another
and so on. You could create the next leap forward in GUI
expressiveness ... but in the end would it be easier than
typing "|"?
Bear
Brian Harvey
>Yeah... You could invent the widgets to use for desktop
>icons and draw lines from inputs of one to outputs of another
>and so on. You could create the next leap forward in GUI
>expressiveness ... but in the end would it be easier than
>typing "|"?
Not for me, but I'm a programmer, so I don't count.
Perhaps the real future is a symbiosis of some kind. Do you know about
Mike Eisenberg's SchemePaint program? It's a standard GUI-ish painting
program with a Scheme interpreter attached. He has cute examples like a
hand-drawn bee in a Scheme-drawn honeycomb. The general principle is that
anything you can do by clicking a button you should be able to do by calling
a procedure, and for any procedure you should be able to add a button to
the GUI.
Brian Harvey
>I love symbolics, but I also love images and kinetics. I see no
>conflict there.
Every semester I show my (SICP-based) class the old Alan Kay video on user
interface history/psychology, and at the end I draw three graphs on the board:
|
| ***
| ***
| *** DOS, Unix, PDP-10, etc. (CLI)
| ***
|*******
|
|
+-------------------------
| ***
| ****
| *
| *
| * Mac Classic, Windows (WIMP GUI)
| *
| *
|***********
+-------------------------
|
|
|
| ***
| *** Alto, Lisp Machine (symbiosis)
| ***
| ***
|***********
+-------------------------
The X axis is the complexity of the task you want to accomplish, and the
Y axis is the degree of wizardliness needed to accomplish it. The idea is
that the old systems required a fair amount of wizardliness to do even the
simplest thing, but then you could grow a step at a time as your needs grow.
The pre-Mac-OS-X GUI-based systems are great if you want to do something
that was anticipated by its designers, but then you hit a wall where you
suddenly have to read 47 volumes of Inside Macintosh to continue. The
desired curve is the last one, achieved in systems that have the WIMP
interface for tasks that have been anticipated by the designers, but let you
read the code for everything, so you can relatively easily roll your own
for unanticipated tasks. Mac OS X is an interesting step in the right
direction, but (imho) not ready for prime time, because the GUI half of
the system doesn't cooperate well with the CLI half. For example, if you
cp /Applications/TextEdit myTextEdit
then the resulting file is not executable by double-clicking in the finder,
because cp doesn't copy all the information MacOS associates with files.
(Instead you have to somehow find out about the CPMac program, which isn't
in your default path, so you have to struggle to find it even after you
find out about it.)
Anyway, the point is, it's theoretically possible to have the best of both
worlds, and I won't be happy until that's what's on my desktop.
P.S. No, Gnome isn't what I want either. :-)
Raffael Cavallaro
> cp /Applications/TextEdit myTextEdit
cp -R ./TextEdit.app ~/myTextEdit.app
It may seem like picking a nit, but imho the GUI and CLI play pretty
nicely on Mac OS X but one does need to know a few things which are
intentionally hidden from naive users to make life simpler for them. In
this case, to make it possible for developers to include lots of stuff
in application bundles but still let naive users install them and move
them around by simple drag and drop without wrecking their contents
and/or internal stucture, the Finder hides the fact that they are
really directories not simple files.
regards,
Ralph
Pascal Bourguignon
GUI and CLI were better integrated in NeXTSTEP/OPENSTEP than in
MacOSX. MacOSX still hasn't caught up on OPENSTEP.
--
__Pascal Bourguignon__ http://www.informatimago.com/
"Logiciels libres : nourris au code source sans farine animale."
Raffael Cavallaro
<use...@informatimago.com> said:
> GUI and CLI were better integrated in NeXTSTEP/OPENSTEP than in
> MacOSX. MacOSX still hasn't caught up on OPENSTEP.
For better or worse (and you clearly think it's worse) NeXT/OPENSTEP
are now owned by Apple and the current release of these technologies is
called Mac OS X ;^)
regards,
Ralph