emacs Vs Eclipse?

T

Tor Iver Wilhelmsen

Andrew Thompson said:
You must be referring to Sun's 1.1.4 which I have never
used personally (started coding Java with 1.1.8) whereas
the MS 1.1.4 is an interesting beast.

Yes, but it had a tool for generating a classes.zip from the strangely
named archives:

clspack -auto

would generate a classes.zip for you.
 
B

Bent C Dalager

Just because an application "offers" keyboard shortcuts doesn't
mean it has good ones.

True. Most of the time, you have _at least_ the ones on the actual
menu buttons. These tend to be the first letter of the displayed text.

In a typical Windows application, the equivalent of emacs'
string-rectangle command (which I have on the default C-x r t binding)
might be on the "_E_dit/_R_ectangle/_R_eplace" menu which you could
trigger by pressing Alt-E R R, not much less intuitive nor much more
cumbersome than the emacs binding.

Cheers
Bent D
 
G

Galen Boyer

True. Most of the time, you have _at least_ the ones on the
actual menu buttons. These tend to be the first letter of the
displayed text.

I'm well aware of the supposed keyboard interface that windows
offers.
In a typical Windows application, the equivalent of emacs'
string-rectangle command (which I have on the default C-x r t
binding) might be on the "_E_dit/_R_ectangle/_R_eplace" menu
which you could trigger by pressing Alt-E R R, not much less
intuitive nor much more cumbersome than the emacs binding.

I don't see how one could ever try to say that windows keyboard
interface is even remotely as usable as Emac's, but you are
entitled to your opinion. I mean, are you truly trying to say
that you navigate a windows machine by your keyboard, or are you
trying to find some semblance of an example where windows might
have a useful keystroke? Windows just doesn't hold a candle to
Emacs in its keystroke interface, period.

In your example, can you also do Alt-E Ctrl-h or can you do Alt-D
R Ctrl-h? Can you do, Ctrl-h k Alt-E R R? Its not only that the
keystrokes are much easier to remember in Emacs, but the help you
can recieve from Emacs is fantastic. Emacs helps one learn and
ingrain the keyboard strokes while windows just puts them there
as a convenience.
 
?

=?ISO-8859-1?Q?Per_R=F8nne?=

Galen Boyer said:
In your example, can you also do Alt-E Ctrl-h or can you do Alt-D
R Ctrl-h?

When I press alt-E in emacs I get é. Are you sure you don't mean meta-E?
^H? And why go to the end to get help? And meta-D doesn't seem to mean
anything - neither does alt-D.

BTW, the meta key is linked to the Mac's cmd-key.
 
G

Galen Boyer

When I press alt-E in emacs I get é. Are you sure you don't
mean meta-E? ^H?

Alt-E is a windows keystroke while Ctrl-h is an Emacs keystroke.
I'm giving an example of what one cannot do in windows not the
keystrokes available in Emacs.

For Emacs help on keystrokes, lets use the `C-x r' prefix key as
an example. Try, `C-x r C-h'. See how Emacs shows you what you
have available from the prefix key?
 
A

Alan Mackenzie

Oliver Scholz said:
(e-mail address removed) writes:
Emacs could definitely do with being easier to use.
Hardly. Emacs is phenomenally easy to use. It's just difficult to
learn how to use.
And this is an important distinction to make!
There are plenty of tools (emacs being one) that are not particularly
novice-friendly, but are quite expert-friendly. (To paraphrase the
old joke about Unix, emacs *is* user-friendly -- it's just choosy about
its friends.) [...]

I am personally not that convinced that this distinction is that
important. I agree that it might become important, if it should happen
that making some functionality more friendly to beginners would also
make it more tedious to use in the long run. That would be bad UI
design.
But I do think that Emacs could see waste improvements here, ....

You mean improvements in the garbage collection algorithm? ;-)
..... without sacrificing the benefits of the interface it currently
provides. Assistants ("wizards") could be one such improvement for
customization. A better GUI (-framework) than a meagre toolbar and an
unwieldy menu another one.

Why, oh why a GUI? Emacs is a program for manipulating _text_. Surely a
TUI is wanted, not a GUI.
 
K

Kai Grossjohann

Galen Boyer said:
In your example, can you also do Alt-E Ctrl-h or can you do Alt-D
R Ctrl-h?

After Alt-E, the application will show you the Edit menu. This menu
contains a list quite similar to the list obtained from Emacs via C-x
r C-h, say, except that the Edit menu might have submenus whose
content is not listed.

What these Windows apps really need is a way to list the keystrokes
that are NOT associated with menus. And they need a way to find
commands by name. And they need all kinds of other stuff.

Kai
 
B

Bent C Dalager

I don't see how one could ever try to say that windows keyboard
interface is even remotely as usable as Emac's, but you are
entitled to your opinion. I mean, are you truly trying to say
that you navigate a windows machine by your keyboard

I typically don't find that particularly problematic.
, or are you
trying to find some semblance of an example where windows might
have a useful keystroke? Windows just doesn't hold a candle to
Emacs in its keystroke interface, period.

It's certainly not as rich. Something along the lines of M-x apropos
would be very welcome in Windows (I don't generally like the Windows
help system all that much) along with a multitude of other emacs-y
features.
In your example, can you also do Alt-E Ctrl-h or can you do Alt-D
R Ctrl-h? Can you do, Ctrl-h k Alt-E R R? Its not only that the

The Windows equivalent is the fact that typing Alt-E or Alt-D R will
display the actual menu in question, and it will show all possible
sub-choices.
keystrokes are much easier to remember in Emacs,

I'm not sure why they are easier to remember. Can you elaborate?
but the help you
can recieve from Emacs is fantastic. Emacs helps one learn and
ingrain the keyboard strokes while windows just puts them there
as a convenience.

I'm not convinced of this. I think the major hurdle in Windows is the
users' inclination rather than the GUI's keyboard support.

Cheers
Bent D
 
G

Galen Boyer

I typically don't find that particularly problematic.

I find it particularly hard to be a keyboard only user in
windows. I spend alot of time in my Start Menu trying to make
that interface easier to navigate.
It's certainly not as rich. Something along the lines of M-x
apropos would be very welcome in Windows (I don't generally
like the Windows help system all that much) along with a
multitude of other emacs-y features.

Yes, I agree.
The Windows equivalent is the fact that typing Alt-E or Alt-D R
will display the actual menu in question, and it will show all
possible sub-choices.

Can you explain the keystrokes again? I can't get the "submenu"
to show up again. The only thing I can do is see the menu
drop-down after I type Alt-E. It sounds like I can get at a
whole different display of menus and help.
I'm not sure why they are easier to remember. Can you
elaborate?

Emacs is all about modes. They can be viewed as "apps" in
windows. Each "mode" author will define default keystrokes for
the particular mode. But, what they make sure of is that similar
operations in one mode have the same keystroke in other modes.

For example: Lets talk about the keystroke `n'.

In any buffer, `n' should go to the "next something". In a
directory editor (ie, windows explorer for Emacs) it is tied to
dired-next-line. In the speedbar, it is tied to speedbar-next,
in the newsreader, it is tied to gnus-summary-next-unread-article.

But, as a user, you would probably think that if you type `n',
you would get the character `n'. Emacs understands this and in
these editable buffer modes, Emacs then did the next best thing
and tied `C-n' to the "next something". But, `n' or `C-n' should
somehow get you to the "next something".

So, the point about `n' is that once I understand that `n' means
"next something" I sort of have the `n' keystroke remembered for
all of Emacs.

Each keystroke combination is thought of in that manner.

I'm not convinced of this. I think the major hurdle in Windows
is the users' inclination rather than the GUI's keyboard
support.

I would just ask my example question again. Why does a windows
user not exclusively use the mouse to cut-n-paste? Its because
the combination of two things. They know how much FASTER Ctrl-c
Ctrl-v are and they REMEMBER that those keystrokes accomplish
what they need done. If windows made it much less cumbersome to
highlight with the keyboard, then the entire operation of
cut-n-paste would normally be done with the keyboard. But it
becomes a bit more cumbersome to move one's hands from the
keyboard and use Shift-ArrowKeys to highlight. Because of that,
most will highlight with the mouse and then hit Ctrl-c Ctrl-v.

Here is another example question highlighting that the windows
user unknowingly gravitates towards keystroke driven interfaces.
How many times does an Excel user hit the F2 key?
 
B

Bent C Dalager

Can you explain the keystrokes again? I can't get the "submenu"
to show up again. The only thing I can do is see the menu
drop-down after I type Alt-E. It sounds like I can get at a
whole different display of menus and help.

I'm not sure what you're getting at. When I type Alt-E, I get to see
all direct descendents of that partial keyboard shortcut. It is
generally evident from the menu texts what the different sub-choices
do and one character tends to be underlined for each choice to
indicate its mnemonic. If any of the sub-choices have application-wide
shortcuts, that is also shown. There is ample opportunity here both to
browse the available commands and to learn the shortcuts needed to
activate them using the keyboard.
Emacs is all about modes. They can be viewed as "apps" in
windows. Each "mode" author will define default keystrokes for
the particular mode. But, what they make sure of is that similar
operations in one mode have the same keystroke in other modes.

Seeing emacs modes as Windows applications, there is much the same
effect in Windows with explicit standardization for a number of
features (e.g., F1 for help) and de facto standardization for a number
of others (Ctrl-N for "new", Ctrl-O for "open", etc.)

Now, I do agree that emacs does this _better_, but I'm not sure that
Windows does it poorly.
For example: Lets talk about the keystroke `n'.

(...)

So, the point about `n' is that once I understand that `n' means
"next something" I sort of have the `n' keystroke remembered for
all of Emacs.

This particular case would be covered by arrow keys in Windows. Of
course, you can either love or hate the arrow keys ...
Each keystroke combination is thought of in that manner.

If I understand you correctly, you are saying that if you've learned
one emacs mode, then you've learnt them all (to an extent).

Windows doesn't give the keyboard user a lot of that transferrable
learning, true. There's a few of the more prominent features (some of
which I mentioned above) but by and large Windows depends on the GUI
to do this job. This, of course, means that the first times you use a
new app, you are left to browse the menus to find the stuff you're
looking for and it might mean that you need to learn widely different
keyboard shortcuts for doing the same thing in different apps.
I would just ask my example question again. Why does a windows
user not exclusively use the mouse to cut-n-paste?

I am sure many do. The few that have bothered to learned the faster
way probably do use the keyboard.
Its because
the combination of two things. They know how much FASTER Ctrl-c
Ctrl-v are and they REMEMBER that those keystrokes accomplish
what they need done. If windows made it much less cumbersome to
highlight with the keyboard, then the entire operation of
cut-n-paste would normally be done with the keyboard. But it
becomes a bit more cumbersome to move one's hands from the
keyboard and use Shift-ArrowKeys to highlight. Because of that,
most will highlight with the mouse and then hit Ctrl-c Ctrl-v.

I believe the shift-arrowkeys aren't used much mostly because what
people need to copy is often found in places were you don't actually
have a cursor. Therefore, they use the mouse because it _always_
works, whether it's in Word, or a web browser, or wherever.

When I use Windows-based editors such as Word or JBuilder I find
shift-arrowkeys vastly superior to using the mouse so I doubt the
reason is that the mouse is less cumbersome. I think it's because it
is more versatile for this particular use.
Here is another example question highlighting that the windows
user unknowingly gravitates towards keystroke driven interfaces.
How many times does an Excel user hit the F2 key?

That really depends how knowledgable he is. Those that have bothered
to learn the F2 key probably use it a lot, but I have known some very
heavy Excel users that just looked at me in amazement when I showed it
to them.

Cheers
Bent D
 
O

Oliver Scholz

Alan Mackenzie said:
You mean improvements in the garbage collection algorithm? ;-)

Uhmph, I think I meant something like "huge". I have no idea of which
word I was thinking when I wrote "waste" ...
Why, oh why a GUI? Emacs is a program for manipulating _text_. Surely a
TUI is wanted, not a GUI.

Hmm, you say that as if "manipulating text" were a constrained area.
As I see it, the fact that Emacs is for manipulating text basically
means that it is not for image or video manipulation.

Sure, a "TUI" is wanted, and a good one at that. But I don't see why
this should make it impossible to have a good GUI /in addition/. A
good (!) GUI presents the most important commands visibly and---to a
certain degree---allows for exploring them by experimenting ("What
happens when I click this?"). A CLI on the other hand, even a good
one, demands that you know it in advance.

I am not exactly a newbie with Emacs and in fact I touch the rodent
rarely. But even /I/ would sometimes wish to have the option to switch
temporarily to a more graphical user interface: whenever I use Elisp
packages that I use rarely. ediff and artist-mode would be examples
for this in my case, because I use each of them about once a year.
Rather than invoking `C-h m' or `C-h a' and studying the output until
I find that damn comand that I used half a year ago to draw a
rectangly, I'd prefer to interact with a GUI. A CLI or a TUI shows
its strength only there were you use it very often. Other than that
it is cumbersome.

And I do think that there is some value in itself in being friendly to
newcomers. When presented with a good (!) GUI, they get the chance to
move to the more powerful TUI gradually. And if they don't, they
don't, but they still get to use Emacs. Some of those newbies could be
future package authors; if they are attracted to Eclipse, because it
looks nicer and it's easier to get started with it (I don't know,
whether the latter is the case), they will add java extensions to
Eclipse instead of writing Emacs Lisp libraries as the gods intended.

Besides that, the TUI could also profit from good GUI design in a few
places. Providing several parameters to a command via the minibuffer
is cumbersome, when you decide that you want to go back and make a
change to what you previously typed. Such commands would better use
some sort of form (which would also work on a console).

Then there is the issue that Emacs hardly makes use of its ability to
use proportional fonts of varying sizes. Despite everything that has
been achieved, Emacs still looks and feels a bit
tty-ish, unnecessarily so. This is not just a matter of eyecandy
(though I do think that eyecandy is also important). But it is a
matter of guiding the eye and of screen estate economy.


That said, I also think that most existing GUIs suck.

Oliver
 
O

Oliver Scholz

Kai Grossjohann said:
After Alt-E, the application will show you the Edit menu. This menu
contains a list quite similar to the list obtained from Emacs via C-x
r C-h, say, except that the Edit menu might have submenus whose
content is not listed.

What these Windows apps really need is a way to list the keystrokes
that are NOT associated with menus.

Especially because those keystrokes pull the menus down, which
is very distracting.
And they need a way to find commands by name. And they need all
kinds of other stuff.

Yeah, it is a matter of the overall design and of a lot of small
details. I couldn't work without isearch used for navigation, for
example. Or without cursor movement commands that move over
syntactical units in source code.

Oliver
 
D

David Kastrup

Oliver Scholz said:
Uhmph, I think I meant something like "huge". I have no idea of which
word I was thinking when I wrote "waste" ...

Vell, yu are a deutshmann, so it vud be a gut gedank sett yu vas
sinking "vast".
 
G

Galen Boyer

I'm not sure what you're getting at. When I type Alt-E, I get
to see all direct descendents of that partial keyboard
shortcut. It is generally evident from the menu texts what the
different sub-choices do and one character tends to be
underlined for each choice to indicate its mnemonic. If any of
the sub-choices have application-wide shortcuts, that is also
shown. There is ample opportunity here both to browse the
available commands and to learn the shortcuts needed to
activate them using the keyboard.

From the first description, I thought I was going to get a place
where the available choices were all summarized.
Seeing emacs modes as Windows applications, there is much the
same effect in Windows with explicit standardization for a
number of features (e.g., F1 for help) and de facto
standardization for a number of others (Ctrl-N for "new",
Ctrl-O for "open", etc.)

Now, I do agree that emacs does this _better_, but I'm not sure
that Windows does it poorly.

Well, when coming from Emacs, it seems quite poor.
This particular case would be covered by arrow keys in
Windows. Of course, you can either love or hate the arrow keys
...


If I understand you correctly, you are saying that if you've
learned one emacs mode, then you've learnt them all (to an
extent).

Yes, this was what the example was trying to explain and "to an
extent" is a correct characterization.
Windows doesn't give the keyboard user a lot of that
transferrable learning, true.

My bigger gripe is that windows doesn't go out of its way to make
the keyboard easy to use, ie, make it a prominent interface.
Emacs has done such a good job with the keyboard that the mouse
is an afterthought.
There's a few of the more prominent features (some of which I
mentioned above) but by and large Windows depends on the GUI to
do this job. This, of course, means that the first times you
use a new app, you are left to browse the menus to find the
stuff you're looking for and it might mean that you need to
learn widely different keyboard shortcuts for doing the same
thing in different apps.

This is exactly how I usually learn what options are available in
a new mode.
I am sure many do. The few that have bothered to learned the
faster way probably do use the keyboard.


I believe the shift-arrowkeys aren't used much mostly because
what people need to copy is often found in places were you
don't actually have a cursor.

Well, thats part of the point.
Therefore, they use the mouse because it _always_ works,
whether it's in Word, or a web browser, or wherever.

When I use Windows-based editors such as Word or JBuilder I
find shift-arrowkeys vastly superior to using the mouse so I
doubt the reason is that the mouse is less cumbersome. I think
it's because it is more versatile for this particular use.


That really depends how knowledgable he is. Those that have
bothered to learn the F2 key probably use it a lot, but I have
known some very heavy Excel users that just looked at me in
amazement when I showed it to them.

I would say that this is the fault of windows. It should be part
of the users experience that she is constantly looking for the
keystrokes to accomplish things, to the point that she's
frustrated when the keystrokes are so cumbersome that she resorts
to the mouse.
 
K

Kai Grossjohann

Galen Boyer said:
My bigger gripe is that windows doesn't go out of its way to make
the keyboard easy to use, ie, make it a prominent interface.
Emacs has done such a good job with the keyboard that the mouse
is an afterthought.

Emacs does have some mouse features, however, that I haven't seen in
Windows apps. For example, in Emacs you can mark text without moving
the cursor by holding down the Meta/Alt key. (You'll also need the
Meta/Alt key for pasting, too.) I haven't seen that feature
elsewhere.

Also, marking a parenthesized expression by double-clicking on its
opening or closing parenthesis is quite convenient -- do Windows apps
offer this kind of feature?

Kai
 
B

Bent C Dalager

On Mon, 18 Oct 2004, (e-mail address removed) wrote:

From the first description, I thought I was going to get a place
where the available choices were all summarized.

The basic difference between emacs and Windows in this regard is that
emacs shows a flat view of all subcommands while Windows gives you a
hierarchical interface to browse all subcommands. I am not sure that
one is superior to the other, they are just different.
Well, when coming from Emacs, it seems quite poor.

That is just because, coming from emacs, you have unnaturally high
standards :)
Yes, this was what the example was trying to explain and "to an
extent" is a correct characterization.

I have a feeling that this effect largely stems from the fact that
Windows applications have a wider range of different functionality
than what emacs modes do. In emacs, all you do is manipulate text and
so it is within the realm of possibility to define some common
commands that will always be useful one way or the other. In Windows,
you manipulate all sorts of information that is presented in all sorts
of ways and so defining such common functionality is more
difficult. Now, it can be seen as a misfeature of Windows that it
insists on showing anything that conceivably _can_ be shown as
graphics, as graphics. It remains, however, that some things are just
that much more convenient to work with in some non-textual
representation (e.g., video editing) and I'm not sure how well a
common set of commands would translate between such applications.
I would say that this is the fault of windows. It should be part
of the users experience that she is constantly looking for the
keystrokes to accomplish things, to the point that she's
frustrated when the keystrokes are so cumbersome that she resorts
to the mouse.

This is what I refer to as the lack of inclination to use the
keyboard. My theory is that since people are incredibly lazy (*) and
since Windows offers a bloody-obvious-but-inefficient interface,
people just keep using that inefficient interface. The fact that
Windows _also_ has an easy-and-efficient interface is completely lost
on them as they simply don't care enough to find out about it. By and
large, the shortcut keys are there in plain sight for them to see
(Excel's F2 key being a notable exception) and so I can only blame the
users for not making use of them.

Alternatively, the blame can be put on whoever trained them and forgot
to mention what the underlined characters mean and what the "Ctrl-x"
to the right of the menu text means.

I don't really think a super-duper emacs-like type of keyboard support
would help these people at all because if they can't be bothered to
learn that F3 means "search" then they certainly can't be bothered to
find out that Alt-E Ctrl-H means "tell me about all commands starting
with Alt-E".

(*) Lazy in a bad "can't be bothered to find out" way rather than lazy
in a good "got to find out how to use this with the least bother" way.

Cheers
Bent D
 
G

Galen Boyer

Emacs does have some mouse features, however, that I haven't
seen in Windows apps.

The actual mouse functionality is actually better in Emacs than
windows, but it is still an afterthought. For me, what is better
in windows is the "mouse only" interface. I find it difficult to
navigate Emacs with just the mouse alone.
For example, in Emacs you can mark text without moving the
cursor by holding down the Meta/Alt key. (You'll also need the
Meta/Alt key for pasting, too.) I haven't seen that feature
elsewhere.

I didn't know this. That could come in quite handy. Kai, as
always, I learn something new from you.

Also, marking a parenthesized expression by double-clicking on
its opening or closing parenthesis is quite convenient -- do
Windows apps offer this kind of feature?

No, and this is one of my main usages of the mouse.
 
A

Alan Mackenzie

David Kastrup said:
+0200: [...]
But I do think that Emacs could see waste improvements here, ....
You mean improvements in the garbage collection algorithm? ;-)
Uhmph, I think I meant something like "huge". I have no idea of which
word I was thinking when I wrote "waste" ...
Vell, yu are a deutshmann, so it vud be a gut gedank sett yu vas
sinking "vast".

Das claw bitch outch.

That post of mine was unkind. Sorry, Oliver.
 
O

Oliver Scholz

Alan Mackenzie said:
David Kastrup said:
+0200:
[...]
But I do think that Emacs could see waste improvements here, ....
You mean improvements in the garbage collection algorithm? ;-)
Uhmph, I think I meant something like "huge". I have no idea of which
word I was thinking when I wrote "waste" ...
Vell, yu are a deutshmann, so it vud be a gut gedank sett yu vas
sinking "vast".

Das claw bitch outch.

That post of mine was unkind. Sorry, Oliver.
[...]

Never mind. I thought it was funny.

I was rather a bit surprised that I could not remember the word I was
thinking of. Now I am a bit surprised that obviously went by sound,
although I am sure that I have never heard nor spoken either "vast" or
"waste". (And that remark might tell you that I believed they were
pronounced the same ...)

Having little occassion to practice speaking of English, my
pronounciation is indeed incredibly bad. Think of a hard German
accent in a bad movie and double it. I zometimes zink zat I should
reflect zat in my orthograffy in order to betray nobuddy hier.

Oliver
 
K

Kai Grossjohann

Oliver Scholz said:
I am not exactly a newbie with Emacs and in fact I touch the rodent
rarely. But even /I/ would sometimes wish to have the option to switch
temporarily to a more graphical user interface: whenever I use Elisp
packages that I use rarely. ediff and artist-mode would be examples
for this in my case, because I use each of them about once a year.
Rather than invoking `C-h m' or `C-h a' and studying the output until
I find that damn comand that I used half a year ago to draw a
rectangly, I'd prefer to interact with a GUI. A CLI or a TUI shows
its strength only there were you use it very often. Other than that
it is cumbersome.

For ediff, just hit ?. That will show you a little cheat sheet of
things to do.

What is artist-mode?

Kai
 

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments. After that, you can post your question and our members will help you out.

Ask a Question

Members online

No members online now.

Forum statistics

Threads
473,982
Messages
2,570,185
Members
46,738
Latest member
JinaMacvit

Latest Threads

Top