How to download and install Emacs
Girish T Deshpande
I am a newbie to Java and programming. I have seen some of the
postings in Java group, everybody talk about Emacs.
I too want to use this editor since I am frustrated with the notepad.
I tried to download and install but couldnt overcome the stuff it
asks(which machine,precompiled,compile.........).
Will anybody please give me the exact URL where I can find download
link, click and ready to use.
I tried myself from gnu site, there were some 2500 files no install
file and many readme files(it takes me months to install and use the
editor if I try to do myself).
Also please suggest me some instruction of how to use it since I didnt
even seen how the editor looks.
I am using Intel, Win XP(professional) machine.
Thanx in advance
Girish
Roedy Green
Deshpande) wrote or quoted :
>I am a newbie to Java and programming. I have seen some of the
>postings in Java group, everybody talk about Emacs.
>
>I too want to use this editor since I am frustrated with the notepad.
>I tried to download and install but couldnt overcome the stuff it
>asks(which machine,precompiled,compile.........).
EMACS is probably the worlds more difficult editor. I suggest
something simpler. see http://mindprod.com/jglosside.html
--
Available for tutoring, problem solving or contract
programming for $50 US per hour. The Java glossary is at
http://www.mindprod.com/jgloss.html
or http://64.251.89.39/jagg.html
-
canadian mind products, roedy green
Chris Verges
> *snip*
If you really want to see a war, ask which is better: notepad, vi,
pico, or emacs.
Anyway, I was feeling compassionate today, so I zipped up my copy and
posted it to http://www.headnut.org/emacs/. It has the JDEE precompiled
into it. Extract the file to C:\ and make sure you use the existing
folder names!!!! If you don't, you will have LOTS of crap in your root
directory.
For help in using emacs, search google. That is a slightly daunting
task to ask people to undergo via newsgroups....
Enjoy.
chris
--
squi...@mail.headnut.org
http://www.headnut.org
"How can you have any pudding if you don't eat your meat?!"
Chris Verges
> EMACS is probably the worlds more difficult editor. I suggest
> something simpler. see http://mindprod.com/jglosside.html
While I do admit Emacs can be very daunting at times, if you stick with
the simple commands, it can be very easy.
Definition of "simple commands":
C-x C-f Find a file (or create a new one)
M-x jde-gen-console With JDEE installed, creates a template for a
Java class
C-x C-c Exits Emacs
C-x <number here> Splits the Emacs window into <number here>
parts
M-% Query/Replace (Search/Replace)
C-s Search based on regex. Press again to move to
the next instance found.
C-g Cancel command
With those, I can accomplish 99% of everything I need. The C- part
means "Control", eg hold down the control key while pressing ... , and
the M- means "Meta" key. On some systems, this key is Alt or Esc. Try
both until you figure yours out.
Chris Verges
> C-s Search based on regex. Press again to move to
> the next instance found.
Correction:
C-u C-s is for regex searching. C-s is for incremental searching.
Le Wang
> I am a newbie to Java and programming. I have seen some of the
> postings in Java group, everybody talk about Emacs.
In which Java group do they talk about Emacs all the time? I want to go see
what they are saying. Have you seen JEdit? (www.jedit.org)
If learning is not part of your goal, I'd go for a much simpler editor/ide.
-Le
Henrik Enberg
> EMACS is probably the worlds more difficult editor. I suggest
> something simpler. see http://mindprod.com/jglosside.html
No it's not. Granted, it's not the worlds easiest to learn, but once
you know it it is _very_ easy to use.
--
Booting... /vmemacs.el
Henrik Enberg
> Will anybody please give me the exact URL where I can find download
> link, click and ready to use.
This is the file you want.
<ftp://ftp.gnu.org/gnu/windows/emacs/21.2/emacs-21.2-fullbin-i386.tar.gz>
Note that winzip is broken and may corrupt the files. It is
recommended that you use a better program for unpacking the files.
<ftp://ftp.gnu.org/gnu/windows/emacs/utilities/i386/djtarnt.exe>
> I tried myself from gnu site, there were some 2500 files no install
> file and many readme files(it takes me months to install and use the
> editor if I try to do myself).
Once you unpack it to some suitable directory, go to the bin/ subdir and
run the program addpm.exe. It will set up various registry entries.
After that it is simply a matter of running the runemacs.exe program in
the same dir.
> Also please suggest me some instruction of how to use it since I didnt
> even seen how the editor looks.
Emacs comes with an extensive user manual. Type "C-h i" when you're
in Emacs to see it.
--
Booting... /vmemacs.el
bruce ingalls
> ...EMACS is probably the worlds more difficult editor...
I recently released EMacro 2.2; that may help your installation.
If you do try EMacro, you will want to unpack all of the JDEE Java
support libraries in the ~/emacs/packages directory, and you should
be good to go.
Before you try EMacro, you will need to unpack Emacs or XEmacs,
install a C compiler, and do
configure
make install
on the source. Else, grab a package, such as an RPM; depends on OS.
Phillip Lord
Roedy> On 20 Jun 2002 14:09:13 -0700, giris...@hotmail.com
Roedy> (Girish T Deshpande) wrote or quoted :
>> I am a newbie to Java and programming. I have seen some of the
>> postings in Java group, everybody talk about Emacs.
>>
>> I too want to use this editor since I am frustrated with the
>> notepad. I tried to download and install but couldnt overcome
>> the stuff it asks(which machine,precompiled,compile.........).
Roedy> EMACS is probably the worlds more difficult editor. I
Roedy> suggest something simpler. see
Roedy> http://mindprod.com/jglosside.html
Roedy, I know you like to moan about emacs, and I guess that this is
okay, but doing so on gnu.emacs.help is really little short of
trolling.
For the record you can get emacs at...
ftp://ftp.gnu.org/gnu/windows/emacs/latest/emacs-21.2-fullbin-i386.tar.gz
To install it you just untar the package. If you want an icon, and a
start menu item as well. then you need to run addpm.exe. This process
is documented.
All of the documentation for emacs can be a little daunting, if you
want to run it on windows. But its ability to present a familiar user
interface on any operating system, from big to small, from new to old,
is one of its many useful characteristics.
Phil
Girish T Deshpande
I downloaded and installed and its working fine now.
The real battle starts now :)
Girish T Deshpande
I downloaded and installed and its working fine now.
The real battle starts now :)
Girish
Kenny Chaffin
says...
> In comp.lang.java.programmer Chris Verges <squi...@mail.headnut.org> wrote:
> > C-s Search based on regex. Press again to move to
> > the next instance found.
>
> Correction:
>
> C-u C-s is for regex searching. C-s is for incremental searching.
>
> chris
>
See.
KAC
--
Kenny A. Chaffin
KAC Website Design - http://www.kacweb.com
Custom/Contract Programming, Graphics, Design
Poetry Page: http://www.kacweb.com/poems/
Lute Kamstra
> In article <aetn2h$dum$1...@news-int.gatech.edu>, squi...@mail.headnut.org
> says...
> > In comp.lang.java.programmer Chris Verges <squi...@mail.headnut.org> wrote:
> > > C-s Search based on regex. Press again to move to
> > > the next instance found.
> >
> > Correction:
> >
> > C-u C-s is for regex searching. C-s is for incremental searching.
> >
> > chris
> >
>
> See.
Yea right. I suppose you know all key bindings of ultraedit by heart
and never make a mistake. Or do you use the menu for every editing
command?
Lute.
--
Lute Kamstra
CWI department PNA4
Dale King
news:87adppi...@enberg.org...
Which is pretty much what Roedy said. Personally I think VI is even harder
to learn. Neither are user friendly (more like user hostile). That says
nothing about their power.
--
Dale King
Phillip Lord
Dale> "Henrik Enberg" <henri...@enberg.org> wrote in message
Dale> news:87adppi...@enberg.org...
>> Roedy Green <ro...@mindprod.com> writes:
>>
>> > EMACS is probably the worlds more difficult editor. I suggest
>> something >simpler. see http://mindprod.com/jglosside.html No
>> it's not. Granted, it's not the worlds easiest to learn, but
>> once you know it it is _very_ easy to use.
Dale> Which is pretty much what Roedy said. Personally I think VI is
Dale> even harder to learn. Neither are user friendly (more like
Dale> user hostile). That says nothing about their power.
No I don't agree. There is a general confusion between "easy to use"
and "easy to learn". Often these concepts are conflated into one
thing. What does user friendly mean? To me, Emacs is very friendly,
and very easy to use. Although I remember in the past, it was a bugger
to learn.
The OP didn't said what he wanted. As he said that he was frustrated
with notepad, perhaps ease of learning was not the issue (notepad is,
after all, very easy to learn, but very hard to use).
Either way Emacs is a perfectly reasonable suggestion for people to
use as an editor. Roedy likes to snipe at it, and describe people who
like it as dogmatic fanatics. But as anyone who has listened to Roedy
expound at length Bali, or SCID's will know, this criticism from him,
is, well, a little rich...
Phil
AlexV
editing/development, with almost transparent access to
linux/unix resources and support for several languages:
For example:
My friend was reading emails, debuging perl and java
programs, and watch for server real-time output
from the same emacs (and in the same time!)...
Nevertheless, in graphical environment (and with remote xserver)
going towards specific language (for example - java)
specialized java editiors/environments offer
MUCH more (for java development) than universal vi/emacs.
(IMHO) AlexV.
"Phillip Lord" <p.l...@russet.org.uk> wrote in message
news:vf3cvci...@rpc71.cs.man.ac.uk...
Chris Smith
> Which is pretty much what Roedy said. Personally I think VI is even harder
> to learn. Neither are user friendly (more like user hostile). That says
> nothing about their power.
He he... I teach a storage hardware config class for Compaq (/HP), and
part of one of the labs contains an instruction that's something like
"now type 'vi /etc/fstab' and add a line that looks like this: ..." No
further instructions.
Needless to say, those who are new to vi in the class are hopelessly
lost. Generally, they break out laughing as I give them instructions:
"okay, now press i, and then type the line... oh you made a mistake? no
problem, just hit Escape and go back, and then hit x until it's gone, and
then i again... okay good, now Escape and two capital Z's.. and there you
are!" Here I'm even relying on cursor keys working... which is
supposedly a sign of bad vi style.
Chris Smith
Phillip Lord
AlexV> Emacs (and VI) have its niche - non graphic (often remote)
AlexV> editing/development, with almost transparent access to
AlexV> linux/unix resources and support for several languages:
Its one of my main reasons for using Emacs. Java, perl, sh, make,
ant (and other XML), tex, all in one. Add in news, and email, and this
makes about 90% of my normal day.
AlexV> For example: My friend was reading emails, debuging perl and
AlexV> java programs, and watch for server real-time output from the
AlexV> same emacs (and in the same time!)...
Sounds fairly normal to me.
AlexV> Nevertheless, in graphical environment (and with remote
AlexV> xserver)
The ability to use Emacs over VT100 is nice of course, but I suspect
that most people use it under X.
AlexV> going towards specific language (for example - java)
AlexV> specialized java editiors/environments offer MUCH more (for
AlexV> java development) than universal vi/emacs.
So I have heard, although I am yet to be convinced, that learning a
new environment which only works for a single language, or other form
of editing, is actually worth while. I suppose if I was working in a
more homogeneous environment this might not be the case, but for me
the high powered, but general purpose text editor works much better
than any "for Java" IDE.
Phil
Dale King
>
> Dale> "Henrik Enberg" <henri...@enberg.org> wrote in message
> Dale> news:87adppi...@enberg.org...
> >> Roedy Green <ro...@mindprod.com> writes:
> >>
> >> > EMACS is probably the worlds more difficult editor. I suggest
> >> something >simpler. see http://mindprod.com/jglosside.html No
> >> it's not. Granted, it's not the worlds easiest to learn, but
> >> once you know it it is _very_ easy to use.
>
>
> Dale> Which is pretty much what Roedy said. Personally I think VI is
> Dale> even harder to learn. Neither are user friendly (more like
> Dale> user hostile). That says nothing about their power.
>
>
> No I don't agree. There is a general confusion between "easy to use"
> and "easy to learn". Often these concepts are conflated into one
> thing. What does user friendly mean? To me, Emacs is very friendly,
> and very easy to use. Although I remember in the past, it was a bugger
> to learn.
Then it was not user friendly. User friendly comprises both ease of use, and
ease of learning. Emacs has a very steep learning curve. It also requires
you to remember a larg number of arcane key combinations so is not easy to
use.
> The OP didn't said what he wanted. As he said that he was frustrated
> with notepad, perhaps ease of learning was not the issue (notepad is,
> after all, very easy to learn, but very hard to use).
I was not responding to the OP, so I don't really care what the OP said. I
am not criticizing his choice of Emacs. When I was in the Unix world I used
Emacs quite well. I was adept at it, but it was never user friendly.
> Either way Emacs is a perfectly reasonable suggestion for people to
> use as an editor. Roedy likes to snipe at it, and describe people who
> like it as dogmatic fanatics. But as anyone who has listened to Roedy
> expound at length Bali, or SCID's will know, this criticism from him,
> is, well, a little rich...
I wasn't criticizing Emacs. Emacs simply is not user friendly. I was
criticizing the defense that Emacs is user friendly becuase it is powerful
and that once you learn it it is easy to use. The fact that it is powerful
and can be learned do not contradict the notion that it is not user
friendly.
--
Dale King
Henrik Enberg
You're confusing user friendly with easy to learn. They are not the
same thing.
--
Booting... /vmemacs.el
Ingalls
>>>EMACS is probably the worlds more difficult editor.
It is *possible* for Emacs to be difficult, but it ain't necessarily so.
As well, Unix can be difficult, but the Mac OS brand X of Unix acts in
an easy way.
It is also possible for me to make EMacro
http://emacro.sf.net/
easier. Are people willing to send me money, for a clean, commercial
distribution?
I believe that EMacro combined with cua.el & Emacs are pretty easy to learn.
In the EMacro + cua system, common keybindings work, menus are helpful,
and configuration is easy, IMHO.
OTOH, you do need these skills:
You need to know how to connect to the internet, and use it. Most people
reading this, have mastered that part.
You need to know how to use an unzip program such as
http://7-zip.sf.net/
You /might/ need to know how to run 'make'
You /might/ need to search EMacro's resources links for downloads, and know
how to copy them into ~/emacs/packages/
It /helps/ to read EMacro's docs, which I believe are concise &
clear. There is a ~/emacs/bin/??/e-install as well as e-refresh
script that removes ~/emacs/preferences/*cache.el, before
changing packages
However, this is still more of a learning curve, than going to the local
store, purchasing a CD, and sticking it into a drive, and under the
right conditions, AUTORUN.INF will tell you how to proceed.
Let me know if there are reasonable suggestions, to make Emacs v21.x +
EMacro + cua.el any easier.
If your suggestions are of an organizational nature, (e.g. convince the
FSF to bundle EMacro); or cost me money (e.g. purchase InstallShield);
or take much work, then I need a commitment of volunteers and/or money,
for these to go any where.
However, I will listen to users willing to do more than complain.
Jon A. Cruz
> Nevertheless, in graphical environment (and with remote xserver)
> going towards specific language (for example - java)
> specialized java editiors/environments offer
> MUCH more (for java development) than universal vi/emacs.
I've yet to see such claims backed up.
I've been working with Java professionally for several years now, and Emacs
with JDEE does just about everything I can think of that I need it to do.
Actually, it does a lot more, but I don't always take advantage of every
feature.
And I moved to emacs straight from being a power DevStudio user (with
custom toolbars, menus, VBscripts all over, etc.)
--
Jon A. Cruz
What's this? It's almost as though the fates were tempting me to...
Bad fates. Baaad... Naughty fates...
Kai Großjohann
> Then it was not user friendly. User friendly comprises both ease of
> use, and ease of learning. Emacs has a very steep learning curve. It
> also requires you to remember a larg number of arcane key
> combinations so is not easy to use.
I think easy to use should be defined by the user, not by you. What
about people who are happy with the key combinations? Maybe Emacs is
just the right thing for them!
I think ease of learning and ease of use should be separated. For
some programs, one is more important, for another program, the other
is more important. For programs that are used only once, easy to
learn is more important. For programs that are used often, easy to
use is more important. And how often it is used depends on the user.
kai
--
A large number of young women don't trust men with beards. (BFBS Radio)
Kai Großjohann
> Nevertheless, in graphical environment (and with remote xserver)
> going towards specific language (for example - java)
> specialized java editiors/environments offer
> MUCH more (for java development) than universal vi/emacs.
Emacs has niftier mouse commands than many other programs. Many
people might not realize, but it's true.
Now Opera and Galeon support mouse gestures, but strokes.el provides
this for Emacs, as well, and is older.
Sven Utcke
> "Phillip Lord" <p.l...@russet.org.uk> wrote in message
> news:vf3cvci...@rpc71.cs.man.ac.uk...
> > No I don't agree. There is a general confusion between "easy to
> > use" and "easy to learn". Often these concepts are conflated into
> > one thing. What does user friendly mean? To me, Emacs is very
> > friendly, and very easy to use. Although I remember in the past,
> > it was a bugger to learn.
>
> Then it was not user friendly.
Well, it _is_ very friendly to it's users. current emacs, as opposed
to 18.x, is also quite friendly to the beginner, but if Phillip
started out with any console-based emacs, then yes, that was a bugger
to learn...
> User friendly comprises both ease of use, and ease of
> learning. Emacs has a very steep learning curve. It also requires
> you to remember a larg number of arcane key combinations so is not
> easy to use.
But it doesn't. Emacs comes with menus, just like any other modern
application (granted, in pre-21.x times the menus were structured
differently from most other applications). So it's perfectly possibly
to use Emacs without ever resorting to _any_ arcane key combination
--- it's just so much easier to _not_ use the mouse, but arcane key
combinations instead, and that is why we all do it...
> I wasn't criticizing Emacs. Emacs simply is not user friendly.
Well, it _is_ very freindly to me --- maybe it's just selective about
its friends?
> I was criticizing the defense that Emacs is user friendly becuase it
> is powerful
But it _is_ userfriendly exactly because it is powerful --- it allows
me to do many things faster and easier than any other given tool (ok,
most tasks could be handled as well by some other tool, but as
different tools are needed for different tasks it is, overall, a lot
easier for me to just use Emacs for everything. And giving me that
power and ease of use I find very friendly indeed).
> The fact that it is powerful and can be learned do not contradict
> the notion that it is not user friendly.
Well, the notion that it wasn't easy to learn, once upon a time,
doesn't exactly contradict the claim that it is very _user_ friendly
either, or did I miss something?
Sven
--
_ __ The Cognitive Systems Group
| |/ /___ __ _ ___ University of Hamburg
| ' </ _ \/ _` (_-< phone: +49 (0)40 42883-2576 Vogt-Koelln-Strasse 30
|_|\_\___/\__, /__/ fax : +49 (0)40 42883-2572 D-22527 Hamburg
|___/ http://kogs-www.informatik.uni-hamburg.de/~utcke/home.html
Marco Schmidt
Sven Utcke wrote:
[...]
>But it _is_ userfriendly exactly because it is powerful --- it allows
>me to do many things faster and easier than any other given tool (ok,
>most tasks could be handled as well by some other tool, but as
>different tools are needed for different tasks it is, overall, a lot
>easier for me to just use Emacs for everything. And giving me that
>power and ease of use I find very friendly indeed).
So, bascially, as soon as _you_ can handle a powerful program well it
must be declared user-friendly?
>Well, the notion that it wasn't easy to learn, once upon a time,
>doesn't exactly contradict the claim that it is very _user_ friendly
>either, or did I miss something?
For you, user-friendly seems to mean that you can get things to work
well, no matter how much time you spent learning it and not taking
into consideration that you are an advanced user (of computer systems
in general). Many (most?) others would see that differently. I found
the following definition at
<http://www.computeruser.com/resources/dictionary/definition.html?lookup=5488>:
Definition for: user-friendly
Easy to learn and use, especially for people who are not experts.This
term is most often used to describe software.
This is exactly the contrary.
Regards,
Marco
--
Please reply in the newsgroup, not by email!
Java programming tips: http://jiu.sourceforge.net/javatips.html
Other Java pages: http://www.geocities.com/marcoschmidt.geo/java.html
Dale King
news:vaf6607...@lucy.cs.uni-dortmund.de...
> "Dale King" <Ki...@TCE.com> writes:
>
> > Then it was not user friendly. User friendly comprises both ease of
> > use, and ease of learning. Emacs has a very steep learning curve. It
> > also requires you to remember a larg number of arcane key
> > combinations so is not easy to use.
>
> I think easy to use should be defined by the user, not by you. What
> about people who are happy with the key combinations? Maybe Emacs is
> just the right thing for them!
But we were not talking about ease of use. We were talking about user
friendliness. Ease of use is part of user friendliness. All you are saying
is that one you muddle past its lack of user friendliness that it can be
used easily, which does not contradict that it is not user friendly.
> I think ease of learning and ease of use should be separated.
From each other, yes. From user friendliness, no. They are an integral part
of user friendliness.
> For
> some programs, one is more important, for another program, the other
> is more important. For programs that are used only once, easy to
> learn is more important.
And for those that are used infrequently.
> For programs that are used often, easy to
> use is more important. And how often it is used depends on the user.
I think the problem is that you think saying that it is not user friendly is
a kiss of death. Saying it is not user friendly does not mean that emacs is
evil, that it should necesarily be avoided, that you should not try it. It
is simply a matter of fact that it is not user friendly. Can that lack of
user freiendliness be overcome? Sure. That still doesn't make it user
friendly.
--
Dale King
Dale King
Nor are they completely separate things. Ease of learning is one facet of
user friendliness.
--
Dale King
Miles Bader
> Nor are they completely separate things. Ease of learning is one facet of
> user friendliness.
It seems like the problem is that you're using the buzzword definition
of `user friendliness', whereas others are using the common-sense
definition of `friendly to its users'.
-Miles
--
o The existentialist, not having a pillow, goes everywhere with the book by
Sullivan, _I am going to spit on your graves_.
Phillip Lord
Marco> fup2 gnu.emacs.help - really not a Java discussion
Marco> Sven Utcke wrote:
Marco> [...]
>> But it _is_ userfriendly exactly because it is powerful --- it
>> allows me to do many things faster and easier than any other
>> given tool (ok, most tasks could be handled as well by some other
>> tool, but as different tools are needed for different tasks it
>> is, overall, a lot easier for me to just use Emacs for
>> everything. And giving me that power and ease of use I find very
>> friendly indeed).
Marco> So, bascially, as soon as _you_ can handle a powerful program
Marco> well it must be declared user-friendly?
Emacs is easy to use for a variety of reasons. Of course its true that
its powerful. But, for instance, the much derided key control
sequences make it very rapid and easy to use. Of course they take time
to learn though. These days emacs has a full menu system though, which
I use for commands I use less frequently.
Marco> I found the following definition
Marco> at
Marco> <http://www.computeruser.com/resources/dictionary/definition.html?lookup=5488>:
Marco> Definition for: user-friendly Easy to learn and use,
Marco> especially for people who are not experts.This term is most
Marco> often used to describe software.
Of course you are free to have your definition of "user-friendly". But
this conflates easy to use and easy to learn, which I think is a
problem. I don't think any one would argue that Emacs is easy to
learn. It is not (although I suspect its easier now than it used to
be). It is easy to use however.
Phil
Robert Uhl <ruhl@4dv.net>
>
> Nor are they completely separate things. Ease of learning is one
> facet of user friendliness.
By your definition. I would argue that there are two qualities:
user-friendliness and learner-friendliness. The former is important
for an app which is used often, while the latter is important for an
app which is use but little. Obviously the ideal is an app which has
both qualities. ISTR an article which argued that this is probably
not possible in a large application: either it'll be easy to learn,
but a chore to use, or difficult to learn, and a delight to use. The
former is exemplified by Windows administration; the latter by emacs.
--
Robert Uhl <ru...@4dv.net>
All I want is a warm bed and a kind word and unlimited power.
--Ashleigh Brilliant
AlexV
"Jon A. Cruz" <jon...@geocities.com> wrote in message
news:3D17E7E3...@geocities.com...
> AlexV wrote:
>
> > Nevertheless, in graphical environment (and with remote xserver)
> > going towards specific language (for example - java)
> > specialized java editiors/environments offer
> > MUCH more (for java development) than universal vi/emacs.
>
> I've yet to see such claims backed up.
>
> I've been working with Java professionally for several years now, and
Emacs
> with JDEE does just about everything I can think of that I need it to do.
> Actually, it does a lot more, but I don't always take advantage of every
> feature.
>
> And I moved to emacs straight from being a power DevStudio user (with
> custom toolbars, menus, VBscripts all over, etc.)
I was not power user of emacs, but did code a bunch (of java) with it.
I still love it (powerful/flexible code highline/format, powerbar,
vt100 and so on). But, let me list some features found, for example,
in jbuilder ($$$) and eclipse (!!!):
---refactoring (various)
---templates (from simplest out>>System.out to
overriding all methods upon inheritance)
---almost transparent version control (diff, branching, tagging)
---debugging, watching
---visual builders
---wizards (from simplest to complex project management)
---backing of powerful (java-concern) companies that insures
up-to-date IDE
---uml <--> code
---various views, filters vs. rather static powerbar
note 1: not all of this features are necessary everyday
(or necessary at all !?!)
note 2: sure it is possible to extend emacs to cover
may be half of this. The question is if somebody doing this?
note 3: I am claiming that specialized IDE can _offer_ more
for java programmer. The result depends on
programmer anyway.
AlexV.
Phillip Lord
AlexV> I was not power user of emacs, but did code a bunch (of java)
AlexV> with it. I still love it (powerful/flexible code
AlexV> highline/format, powerbar, vt100 and so on). But, let me list
AlexV> some features found, for example, in jbuilder ($$$) and
AlexV> eclipse (!!!):
AlexV> ---refactoring (various) ---templates (from simplest
AlexV> out>>System.out to overriding all methods upon inheritance)
AlexV> ---almost transparent version control (diff, branching,
AlexV> tagging) ---debugging, watching ---visual builders ---wizards
AlexV> (from simplest to complex project management) ---backing of
AlexV> powerful (java-concern) companies that insures up-to-date IDE
AlexV> ---uml <--> code ---various views, filters vs. rather static
AlexV> powerbar
AlexV> note 1: not all of this features are necessary everyday (or
AlexV> necessary at all !?!)
AlexV> note 2: sure it is possible to extend emacs to cover may be
AlexV> half of this. The question is if somebody doing this?
Not all of your points are entirely clear. Many of those that are, are
covered. templates, yes, version control, yes, debugging yes, various
views, filters, yes. Although here its not clear what you mean.
UML, no, visual builders, no, wizards no (of the sort you describe
anyway, no).
AlexV> note 3: I am claiming that specialized IDE can _offer_ more
AlexV> for java programmer. The result depends on programmer anyway.
Eclipse looks nice, to be honest. The ethos is very similar to
emacs...extensibility. I haven't played with it yet. Once it reaches a
somewhat higher level of maturity it might be worth a closer look. It
would also be nice if their website stopped crashing netscape....
Phil
Stefan Monnier <foo@acm.com>
Could people get a bit more concrete ?
I mean, it can be fun to talk about whether Emacs is or is not
user-friendly, but I feel like it would be a lot more fun if people could
back up their arguments with facts, such as horror stories and (better yet)
suggestion for how to make it better.
This is not just because it can be fun, but also because as an Emacs
maintainer who first of all cares about "Emacs for power users", I also
have an interest in Emacs being easier to learn.
I've already given the example of the split-window, so I'll throw in another
this time: how to jump to the definition of a function. Emacs has
reasonably good support for such functionality, but it's not really trivial
to learn: there are several packages each one of them doing something
slightly different. `imenu' is not bound to anything by default. `etags'
requires building the TAGS file manually, some major modes have their own
ad-hoc support, ... and none of it is advertised anywhere, so users have
to actively look for that functionality.
Now, being an Emacs bigot, I have somewhere between -0.0 and +0.0
seconds of exposure to other all-singing-and-dancing editors, so I don't
know how other editors advertise that functionality or make it obvious
to the user how it works. Can someone help ?
Stefan
Robert Uhl <ruhl@4dv.net>
>
> ---almost transparent version control (diff, branching, tagging)
Emacs has this. Its CVS functionality is really quite nice.
> ---debugging, watching
Emacs has this as well, for many languages. I'm not certain if Java
is supported, though.
> ---visual builders
> ---wizards (from simplest to complex project management)
I'm not certain that these are assets...
> ---backing of powerful (java-concern) companies that insures
> up-to-date IDE
Emacs is backed by its users--as long as there are folks who use the
same functionality you do, you can be pretty certain that it will be
improved. And even if there are not, _you_ can improve it.
--
Robert Uhl <ru...@4dv.net>
One doesn't expect governments to obey the law because of some
higher moral development. One expects them to obey the law because
they know that if they don't, those who aren't shot will be hanged.
--Michael Shirley
Stefan Monnier <foo@acm.com>
Can someone describe precisely the kind of refactoring that are offered ?
> ---templates (from simplest out>>System.out to
> overriding all methods upon inheritance)
I don't believe in templates as far as "creating code from scratch".
And this despite my being a slow typist. Typing code is really not
the most time-consuming part of programming. I'd rather see progress
in the realm of "code syncing" as in "here is a class, here is an interface,
please do what is obviously needed so that the class implements the
interface". I guess it can be considered as "templates where the output
is pre-existing and should be adapted". Maybe it's generally understood
as being part of refactoring ?
> ---almost transparent version control (diff, branching, tagging)
VC's support for branching is not too hot, but the rest seems OK.
Can you expand on what jbuilder does additionally or better ?
> ---debugging, watching
I hear someone is working on a better GUD for gdb (with separate windows
showing the current stack trace, the set of breakpoints, the currently
selected list of displayed variables, ...). What other things were
you thinking about ?
> ---various views, filters vs. rather static powerbar
You mean `speedbar' ?
What are the various views/filters exactly ?
> note 1: not all of this features are necessary everyday
> (or necessary at all !?!)
But if they are easily available, they can be very helpful.
Stefan
Joona I Palaste
on comp.lang.java.programmer:
> A large number of young women don't trust men with beards. (BFBS Radio)
Who do they trust with beards then?
--
/-- Joona Palaste (pal...@cc.helsinki.fi) ---------------------------\
| Kingpriest of "The Flying Lemon Tree" G++ FR FW+ M- #108 D+ ADA N+++|
| http://www.helsinki.fi/~palaste W++ B OP+ |
\----------------------------------------- Finland rules! ------------/
"Products like that make me wish I could menstruate."
- Andy Richter
AlexV
"Stefan Monnier <f...@acm.com>"
<monnier+gnu.emacs.help/news/@flint.cs.yale.edu> wrote in message
news:5lelev6...@rum.cs.yale.edu...
> >>>>> "AlexV" == AlexV <avek_...@videotron.ca> writes:
> > ---refactoring (various)
>
> Can someone describe precisely the kind of refactoring that are offered ?
(from eclipse)
---rename, modify parameters, moving
members , functions with
global synchronization including javadocs.
---creating set/getters implement interfaces,
abstract methods
---extracting code fragment to method with proper parameters
---cover fragment by try/catch with automatically detected
Exception types
---import suggestion and "beatification"
Most of refactoring show you pre-view of changing
with all altered areas colored.
And, of course, powerful help, code assist, code completition....
> > ---templates (from simplest out>>System.out to
> > overriding all methods upon inheritance)
> I don't believe in templates as far as "creating code from scratch".
> And this despite my being a slow typist. Typing code is really not
> the most time-consuming part of programming. I'd rather see progress
> in the realm of "code syncing" as in "here is a class, here is an
interface,
> please do what is obviously needed so that the class implements the
> interface". I guess it can be considered as "templates where the output
> is pre-existing and should be adapted". Maybe it's generally understood
> as being part of refactoring ?
It is just __faster__. The generated code is no better or worst than
accurate hand coding...
> > ---almost transparent version control (diff, branching, tagging)
>
> VC's support for branching is not too hot, but the rest seems OK.
> Can you expand on what jbuilder does additionally or better ?
I just do not know if emacs is able to
get//create new project from/to cvs by
couple of clicks.
> > ---debugging, watching
>
> I hear someone is working on a better GUD for gdb (with separate windows
> showing the current stack trace, the set of breakpoints, the currently
> selected list of displayed variables, ...). What other things were
> you thinking about ?
...looks like a lot of thing arrive to emacs in last 1.5 year.
But can you (in emacs), stopping on breakpoint navigate
to any argument and to read values of ANY object
(without specifically set watch ?) Eclipse even offers
(limited) hot code replacement in the middle of
debugging
> > ---various views, filters vs. rather static powerbar
>
> You mean `speedbar' ?
> What are the various views/filters exactly ?
(sorry, of course "speedbar")
Views can show you the same workspace from
different angle:
-- only java files and jars
-- all files ( including, for example .cvsignore, that are important but
annoing by themself)
-- cvs view, where you can see only branches and tags and
see big picture of progress
-- debug set of windows
-- filters can just hide, for example, all CVS directories
> > note 1: not all of this features are necessary everyday
> > (or necessary at all !?!)
>
> But if they are easily available, they can be very helpful.
! this my point too !
AlexV.
> Stefan
AlexV
> >>>>> "AlexV" == AlexV <avek_...@videotron.ca> writes:
..............................
> Eclipse looks nice, to be honest. The ethos is very similar to
> emacs...extensibility. I haven't played with it yet. Once it reaches a
> somewhat higher level of maturity it might be worth a closer look. It
> would also be nice if their website stopped crashing netscape....
They are couple week before major 2.0 release, RC2 (they call GM2)
is there. There (eclipse newsgroup) was some discussion
about some plug-in for emacs-style coding... :-)
AlexV.
P.S. Please, see my post in response to
Stefan Monnier, there is some more info about
specific features.
Thien-Thi Nguyen
> It is simply a matter of fact that it is not user friendly. Can that lack of
> user freiendliness be overcome? Sure. That still doesn't make it user
> friendly.
if it were a matter of fact we could verify it independently (first specifying
the method of verification, which would also need to be able to be validated
independently, and so on -- where you ground out depends on how rigourous your
science is).
"user friendly" is a matter of user definition (basically, user preference).
it's difficult to know everyone's preferences and sometimes difficult to know
your own for what they are.
thi
Jon A. Cruz
> I was not power user of emacs, but did code a bunch (of java) with it.
> I still love it (powerful/flexible code highline/format, powerbar,
> vt100 and so on). But, let me list some features found, for example,
> in jbuilder ($$$) and eclipse (!!!):
>
> ---refactoring (various)
JDEE (some)
>
> ---templates (from simplest out>>System.out to
> overriding all methods upon inheritance)
JDEE
> ---almost transparent version control (diff, branching, tagging)
Yes. Implemented _very_ well.
And integration with many different types of version control including CVS and
StarTeam.
> ---debugging, watching
Yup. And cross-debugging for JNI and debugging of C/C++
> ---visual builders
Ewww. Hate them. Most professionals I've worked with find that they actually
slow down projects, as the up-front time saved is more that taken over by
later refactoring needed to get things robust.
> ---wizards (from simplest to complex project management)
JDEE.
> ---backing of powerful (java-concern) companies that insures
> up-to-date IDE
Backing of programmers from the begining. Written by programmers for
programmers, and no marketing types or other executives getting in the way.
> ---uml <--> code
Don't do that myself. Probably should at some point, but I've been mainly on
stuff that XP works well on.
> ---various views, filters vs. rather static powerbar
So far the few things I've needed done that can't be hit by a simple *shell is
easily accomplished by a little elisp.
Also, splitting the window and viewing data/programs/output in differerent
ways is fairly simple (you don't have to rely just on the Speedbar)
> note 2: sure it is possible to extend emacs to cover
> may be half of this. The question is if somebody doing this?
Some.
> note 3: I am claiming that specialized IDE can _offer_ more
> for java programmer. The result depends on
> programmer anyway.
Ahhh... but if you look at say eclipse, and then compare it to Emacs, you end
up with something that's just about the same.
Emacs is a developer's tool that integrates other tools, has
'plugin/extensibility' support, works with a plethora of languages, has
third-party integration support for things such as SGML, XML, HTML, commercial
source control (e.g. StarTeam), and is internationalized significantly:
Support for many languages and their scripts, including all
the European ``Latin'' scripts, Russian, Greek, Japanese,
Chinese, Korean, Thai, Vietnamese, Lao, Ethiopian, and some
Indian scripts. (Sorry, Mayan hieroglyphs are not supported.)
[From gnu.org's "What is Emacs"]
Some _existing_ extensions for Emacs include programmer's language-specific
support for lisp, GNU Makefiles, C/C++, Java, FORTRAN, Pascal, Objective-C,
Perl, Tcl, PHP, TeX, Haskell, Eiffel, POV-Script, Eqn, Mathematica, Maple,
Ada, Python, Q, Miranda, SQL-Plus, Shakespeare...
Whew! And all that just from my first page of Google results.
Once eclipse gets mature, I have a feeling that it will have a similiar level
of support.
Oh, and since when I first ran Emacs after upgrading to RedHat 7.3, it even
started up with graphical buttons in a toolbar across the top (needless to
say, I turned that option off right-quick!). Looking at both Emacs and Eclipse
from a functional or abstract viewpoint, they seem to be almost the same.
Jon A. Cruz
> AlexV wrote:
>
> > I was not power user of emacs, but did code a bunch (of java) with it.
> > I still love it (powerful/flexible code highline/format, powerbar,
> > vt100 and so on). But, let me list some features found, for example,
> > in jbuilder ($$$) and eclipse (!!!):
> >
> > ---refactoring (various)
>
> JDEE (some)
And there are other refactoring tools for Emacs, at least one commercial.
Michel Schinz
> >>>>> "AlexV" == AlexV <avek_...@videotron.ca> writes:
> > ---refactoring (various)
About refactoring, since nobody pointed it out yet, I just wanted to
say that something (commercial) exists for Emacs: Xrefactory
[http://www.xref-tech.com/xrefactory/].
I used it for some time (the free demo version) and didn't really like
the way it (doesn't) integrate with Emacs, but the functionalities
provided are interesting.
[...]
> > ---templates (from simplest out>>System.out to
> > overriding all methods upon inheritance)
>
> I don't believe in templates as far as "creating code from scratch".
> And this despite my being a slow typist. Typing code is really not
> the most time-consuming part of programming.
I think it really depends what kind of programming you do. I have the
same opinion as you concerning templates, but that's because I'm
fortunate enough to write mostly "non-trivial" programs. I think that
for some kind of programs (GUI, Enterprise JavaBeans, etc.) the amount
of "boring" code one has to write makes templates attractive. Or at
least some kind of automatic code generation.
> I'd rather see progress in the realm of "code syncing" as in "here
> is a class, here is an interface, please do what is obviously needed
> so that the class implements the interface".
That's definitely useful (and available for Emacs, for Java at least,
through JDEE [http://jdee.sunsite.dk/]).
Michel.
Kai Großjohann
> But we were not talking about ease of use. We were talking about user
> friendliness.
What is user friendliness?
Phillip Lord
AlexV> And, of course, powerful help, code assist, code
AlexV> completition....
JDEE does code completion, and javadoc look up.
>> > ---templates (from simplest out>>System.out to overriding all
>> methods upon >inheritance) I don't believe in templates as far as
>> "creating code from scratch".
AlexV> It is just __faster__. The generated code is no better or
AlexV> worst than accurate hand coding...
There are two arguments here. I think over reliance on code wizards is
very bad. The point is that the way that you have built the
application is not obvious from the code.
On the other hand Java particularly is fairly verbose, and some sort
of wizard support is useful. Being able to say "write template methods
to implement this interface", or "write a get/set pair", or "delegate
this interface to this object" is useful. Emacs will do all of this.
>> > ---almost transparent version control (diff, branching,
>> tagging) VC's support for branching is not too hot, but the rest
>> seems OK. Can you expand on what jbuilder does additionally or
>> better ?
AlexV> I just do not know if emacs is able to get//create new
AlexV> project from/to cvs by couple of clicks.
If you are adding to an existing repository module, then yes. If you
are creating a new module, then you need to do this from the command
line. This stems directly from CVS. The first import requires a little
bit of playing.
Eclipse seems to store all of its code in a single place called a work
bench. If you copied this notion with emacs, you'd have to add your
work bench once (or check out the shared work bench already in the
repository), and then it just works...
Phil
Sven Utcke
I would. It takes time, no doubt, and few people, if any, could claim
to have mastered all aspects of Emacs --- but then again, who would
need that (especially as it would mean that you have mastered, e.g., 4
or 5 different Email-clients build into Emacs --- to give just one
example).
But apart from it taking time, it _is_ easy. When you start Emacs
( at least up to 20.x) it will tell you:
Get a tutorial C-h t
(it also tells you what this means, and if invoked on X11 tells you to
use the menu). You do just that (which takes anything from 30 minutes
to 2 hours), ignore anything it says about C-f and it's ilk, and you
are up and running with an editor at least as powerfull as the early
WordStar based IDEs. Once you got so far, you simply add to your
knowledge as you go along, and as needed. Yes, this requires you to
read something like a manual, and yes, it takes some time, but it is
perfectly _easy_ --- nothing to hard for anyone who is
a) able to read (and why else would you use an editor?)
b) equiped with an IQ > 80 (which you probably are if you managed a).
This is certainly easier than learning to drive a car, never mind
riding a bike --- and I at least spend a lot more time in Emacs than
on a bike (and more on a bike than in a car) --- so learning Emacs was
time well invested.
Phillip Lord
Sven> Phillip Lord <p.l...@russet.org.uk> writes:
>> >>>>> "Marco" == Marco Schmidt <marcos...@geocities.com>
>> writes:
Marco> Definition for: user-friendly Easy to learn and use, Marco>
>> especially for people who are not experts.This term is most
>> Marco> often used to describe software.
>>
>> Of course you are free to have your definition of
>> "user-friendly". But this conflates easy to use and easy to
>> learn, which I think is a problem. I don't think any one would
>> argue that Emacs is easy to learn.
Sven> I would.
Sven> But apart from it taking time, it _is_ easy. When you start
Sven> Emacs ( at least up to 20.x) it will tell you:
I think this is stretching the definition of "easy" a little too
far. I would say that something that takes a long time is not
particularly easy.
From my own memory I remember learning emacs to be a frustrating
experience.
Sven> a) able to read (and why else would you use an editor?)
Ask someone with a strong visual impairment. Fortunately you can use
emacs without being able to read.
Sven> This is certainly easier than learning to drive a car, never
Sven> mind riding a bike
This is a different point. Cars and bikes are safety critical. They
are not that hard to use, but you have to learn it well as the
consequences of screwing up are possibly fatal.
Sven> so learning Emacs was time well invested.
On this I agree.
Phil
Dale King
news:%K3S8.13286$r54.5...@weber.videotron.net...
>
> "Stefan Monnier <f...@acm.com>"
> <monnier+gnu.emacs.help/news/@flint.cs.yale.edu> wrote in message
> news:5lelev6...@rum.cs.yale.edu...
> > >>>>> "AlexV" == AlexV <avek_...@videotron.ca> writes:
> > > ---refactoring (various)
> >
> > Can someone describe precisely the kind of refactoring that are offered
?
>
> (from eclipse)
> ---rename, modify parameters, moving
> members , functions with
> global synchronization including javadocs.
> ---creating set/getters implement interfaces,
> abstract methods
> ---extracting code fragment to method with proper parameters
> ---cover fragment by try/catch with automatically detected
> Exception types
> ---import suggestion and "beatification"
>
> Most of refactoring show you pre-view of changing
> with all altered areas colored.
You forgot extract local variable, inline local variable, and
self-encapsulate field. The list continues to grow. The refactoring supports
undo and redo.
> And, of course, powerful help, code assist, code completition....
>
> > > ---templates (from simplest out>>System.out to
> > > overriding all methods upon inheritance)
> > I don't believe in templates as far as "creating code from scratch".
> > And this despite my being a slow typist. Typing code is really not
> > the most time-consuming part of programming. I'd rather see progress
> > in the realm of "code syncing" as in "here is a class, here is an
> interface,
> > please do what is obviously needed so that the class implements the
> > interface". I guess it can be considered as "templates where the output
> > is pre-existing and should be adapted". Maybe it's generally understood
> > as being part of refactoring ?
>
> It is just __faster__. The generated code is no better or worst than
> accurate hand coding...
It is very intelligent templates. For example I have a collection and need
to iterate over it. I can type for hit ctrl-space and select "for - iterate
over collection". It will insert the following:
for (Iterator iter = collection.iterator(); iter.hasNext();)
{
type element = (type) iter.next();
}
But it doesn't stop there I can for instance change the name of the local
variable iter in the declaration and it will automatically change the
others. The same with the type. There are several templates like that. And
many times it will fill in the variable and type with a nearby declaration
(for iterate over array usually).
--
Dale King
Dale King
> >>>>> "Marco" == Marco Schmidt <marcos...@geocities.com> writes:
>
> Marco> fup2 gnu.emacs.help - really not a Java discussion
> Marco> Sven Utcke wrote:
>
> Marco> [...]
>
> >> But it _is_ userfriendly exactly because it is powerful --- it
> >> allows me to do many things faster and easier than any other
> >> given tool (ok, most tasks could be handled as well by some other
> >> tool, but as different tools are needed for different tasks it
> >> is, overall, a lot easier for me to just use Emacs for
> >> everything. And giving me that power and ease of use I find very
> >> friendly indeed).
>
> Marco> So, bascially, as soon as _you_ can handle a powerful program
> Marco> well it must be declared user-friendly?
>
> Emacs is easy to use for a variety of reasons. Of course its true that
> its powerful. But, for instance, the much derided key control
> sequences make it very rapid and easy to use. Of course they take time
> to learn though. These days emacs has a full menu system though, which
> I use for commands I use less frequently.
Which is simply another way of saying it is not user friendly.
> Marco> I found the following definition
> Marco> at
> Marco>
<http://www.computeruser.com/resources/dictionary/definition.html?lookup=548
8>:
>
> Marco> Definition for: user-friendly Easy to learn and use,
> Marco> especially for people who are not experts.This term is most
> Marco> often used to describe software.
>
>
> Of course you are free to have your definition of "user-friendly". But
> this conflates easy to use and easy to learn, which I think is a
> problem. I don't think any one would argue that Emacs is easy to
> learn. It is not (although I suspect its easier now than it used to
> be). It is easy to use however.
Which once again is another way of saying it is not user friendly.
--
Dale King
Dale King
> "Dale King" <Ki...@TCE.com> writes:
> >
> > Nor are they completely separate things. Ease of learning is one
> > facet of user friendliness.
>
> By your definition. I would argue that there are two qualities:
> user-friendliness and learner-friendliness. The former is important
> for an app which is used often, while the latter is important for an
> app which is use but little. Obviously the ideal is an app which has
> both qualities.
And thus is why we have the term user friendly which encompasses both of
them.
> ISTR an article which argued that this is probably
> not possible in a large application: either it'll be easy to learn,
> but a chore to use, or difficult to learn, and a delight to use. The
> former is exemplified by Windows administration; the latter by emacs.
And that is sheer nonsense. I use Eclipse which is extremely powerful and
very easy to use.
--
Dale King
Dale King
news:kk9fzzb...@glug.org...
No it is not a matter of personal preference. User friendliness is not about
a single person it is about its ease of learning and ease of use for a
population. It could be objectively measured and I'm sure human factors
people have ways of doing so.
--
Dale King
Stefan Monnier <foo@acm.com>
idiot, it would help ;-]
>>>>> "AlexV" == AlexV <avek_...@videotron.ca> writes:
> ---rename, modify parameters, moving members , functions with global
> synchronization including javadocs.
Renaming variables with proper scope is something that I'd like to see
in Emacs, indeed. I'm not sure what you meant by the rest, tho.
> ---creating set/getters implement interfaces, abstract methods
The languages I use don't have any such things, so I don't have
a clear idea of how useful it'd be. It sounds like templates
more than refactoring, tho. Unless you mean that it automatically
changes all the code that directly accesses the variables to
use the getter/setter functions instead ? That would be cool.
> ---extracting code fragment to method with proper parameters
Cool. Does it guarantee that the semantics of the code are 100%
preserved or does it just do a pretty good job that you have to
double-check ?
> ---cover fragment by try/catch with automatically detected Exception types
> ---import suggestion and "beatification"
Not sure what those are.
>> I don't believe in templates as far as "creating code from scratch".
>> And this despite my being a slow typist. Typing code is really not
>> the most time-consuming part of programming. I'd rather see progress
[...]
> It is just __faster__.
Sure, but I rarely notice the difference. I agree with Michel, that it
might be due to the fact that my programs don't suffer from the same kind of
redundancy and/or verbosity that tends to plague Java and/or some kinds
of applications.
I actually believe that verbosity is bad. And boring code should
simply not have to be seen at all (i.e. auto-generated by the compiler,
either through the use of macros or templates or higher-order functions
or whichever other kind of abstraction mechanism you prefer).
After all, I still haven't seen a rebuff of the study that showed that the
overall coding speed measured in lines of code is pretty much independent
from the language used, so a less verbose program will be more quickly done
and have fewer bugs (since most of the time actually came from bug fixing
rather than writing the code itself).
>> > ---almost transparent version control (diff, branching, tagging)
>>
>> VC's support for branching is not too hot, but the rest seems OK.
>> Can you expand on what jbuilder does additionally or better ?
> I just do not know if emacs is able to get//create new project from/to
> cvs by couple of clicks.
Indeed, Emacs does not offer much in terms creating a new CVS project
or of checking out a CVS project. Creating a CVS project is rare enough
that I don't find it to be a problem. But I agree that a "project browser"
where you can browse CVS repositories and do checkouts would be a nice
addition. Problem is that it is painful to do if you stick to the `cvs'
command (i.e. you pretty much have to access the repository either directly
or through the client/server protocol).
>> > ---debugging, watching
>>
>> I hear someone is working on a better GUD for gdb (with separate windows
>> showing the current stack trace, the set of breakpoints, the currently
>> selected list of displayed variables, ...). What other things were
>> you thinking about ?
> ...looks like a lot of thing arrive to emacs in last 1.5 year.
I said "is working", so feel free to assume it's part of the future rather
than the past ;-)
> But can you (in emacs), stopping on breakpoint navigate to any argument
> and to read values of ANY object (without specifically set watch ?)
Isn't that the basic functionality of any debugger ?
If so, yes of course. Otherwise, please explain.
I think there's even a tooltips mode for gud where the value of variable
appears in tooltips as you move the mouse over the variable. I never tried
it, tho.
> Eclipse even offers (limited) hot code replacement in the middle of
> debugging
This would be considered as being outside the scope of Emacs and thus
depends on the debugger you're using. But GUD (the interface between Emacs
and the underlying debugger) currently has no special support for such
features in a debugger, so even if GDB supports it, GUD would supports it
only via the interactive command-line it offers.
> -- cvs view, where you can see only branches and tags and
> see big picture of progress
What does this look like. Can you direct me to some snapshots ?
Does it present it on a file-by-file basis or on the whole project ?
I tried to write such a thing, but stumbled on the fact that tags/branches
in CVS are defined on a file-by-file basis (even though they are usually
present in all the files of a project and used more or less consistently
across those files) and that it's not easy to merge the data into
a coherent view while still doing something reasonable in the case where
some things are not coherent (a branch is missing from one of the files,
or a tag has been moved to another branch on some of the files, ...)
Stefan
Phillip Lord
Dale> "Phillip Lord" <p.l...@russet.org.uk> wrote in message
>> Emacs is easy to use for a variety of reasons. Of course its true
>> that its powerful. But, for instance, the much derided key
>> control sequences make it very rapid and easy to use. Of course
>> they take time to learn though. These days emacs has a full menu
>> system though, which I use for commands I use less frequently.
Dale> Which is simply another way of saying it is not user friendly.
No. The key bindings are hard to learn, but easier to use compared to
the menu items. This is why emacs has both.
>> Of course you are free to have your definition of
>> "user-friendly". But this conflates easy to use and easy to
>> learn, which I think is a problem. I don't think any one would
>> argue that Emacs is easy to learn. It is not (although I suspect
>> its easier now than it used to be). It is easy to use however.
Dale> Which once again is another way of saying it is not user
Dale> friendly.
No. As I said it depends on what you mean by "user friendly". As has
been said before in this thread, if we drop this rather foolish
buzzword, and ask questions about "easy to learn", and "easy to use"
then the discussion would be easier.
Phil
Kai Großjohann
> It is very intelligent templates. For example I have a collection and need
> to iterate over it. I can type for hit ctrl-space and select "for - iterate
> over collection". It will insert the following:
>
> for (Iterator iter = collection.iterator(); iter.hasNext();)
> {
> type element = (type) iter.next();
> }
>
> But it doesn't stop there I can for instance change the name of the local
> variable iter in the declaration and it will automatically change the
> others. The same with the type. There are several templates like that. And
> many times it will fill in the variable and type with a nearby declaration
> (for iterate over array usually).
It's fairly easy to build skeletons to do that. The skeleton would
ask for the collection to iterate over or the name of the iterator
variable.
With a little more work, the collection variable could be glorked
from context.
There are some code generation wizards in JDEE already, but such an
iterator construct does not show up there. Any takers?
Eli Zaretskii
>
> I would say that something that takes a long time is not
> particularly easy.
Alternatively, it might be that there's much to learn, i.e. that Emacs is
simply a large package with many features.
> From my own memory I remember learning emacs to be a frustrating
> experience.
I rather enjoyed it.
Jon A. Cruz
AlexV
> AlexV wrote:
.......
> Some _existing_ extensions for Emacs include programmer's
language-specific
> support for lisp, GNU Makefiles, C/C++, Java, FORTRAN, Pascal,
Objective-C,
> Perl, Tcl, PHP, TeX, Haskell, Eiffel, POV-Script, Eqn, Mathematica, Maple,
> Ada, Python, Q, Miranda, SQL-Plus, Shakespeare...
>
> Whew! And all that just from my first page of Google results.
>
>
> Once eclipse gets mature, I have a feeling that it will have a similiar
level
> of support.
>
> Oh, and since when I first ran Emacs after upgrading to RedHat 7.3, it
even
> started up with graphical buttons in a toolbar across the top (needless to
> say, I turned that option off right-quick!).
:-)))
> Looking at both Emacs and Eclipse
> from a functional or abstract viewpoint, they seem to be almost the same.
Menu > Refactoring > Implement abstract viewpoint > Enter ... :-)))
Have a nice day,
AlexV
AlexV
.....
> The languages I use don't have any such things, so I don't have
> a clear idea of how useful it'd be. It sounds like templates
> more than refactoring, tho. Unless you mean that it automatically
> changes all the code that directly accesses the variables to
> use the getter/setter functions instead ? That would be cool.
>
> > ---extracting code fragment to method with proper parameters
>
> Cool. Does it guarantee that the semantics of the code are 100%
> preserved or does it just do a pretty good job that you have to
> double-check ?
Well... eclipse even compiles changes rigth-away! But human check
is anyway unavoidable (may be someone wants add error checking
inside new method or just throw an exception....)
> > ---cover fragment by try/catch with automatically detected Exception
types
> > ---import suggestion and "beatification"
>
> Not sure what those are.
For example 10 lines of code have several calls that throw
SQLException, IOException and CustomizedException.
Try/catch refactoring will create
try{
......
}
catch(SQLException sqle){}
catch(IOException ioe){}
catch(CustomizedException ex){}
> >> I don't believe in templates as far as "creating code from scratch".
> >> And this despite my being a slow typist. Typing code is really not
> >> the most time-consuming part of programming. I'd rather see progress
> [...]
> > It is just __faster__.
>
> Sure, but I rarely notice the difference. I agree with Michel, that it
> might be due to the fact that my programs don't suffer from the same kind
of
> redundancy and/or verbosity that tends to plague Java and/or some kinds
> of applications.
>
> I actually believe that verbosity is bad. And boring code should
> simply not have to be seen at all (i.e. auto-generated by the compiler,
> either through the use of macros or templates or higher-order functions
> or whichever other kind of abstraction mechanism you prefer).
>
> After all, I still haven't seen a rebuff of the study that showed that the
> overall coding speed measured in lines of code is pretty much independent
> from the language used, so a less verbose program will be more quickly
done
> and have fewer bugs (since most of the time actually came from bug fixing
> rather than writing the code itself).
My friend wrote two lines of perl replacing 33 java lines !!!!!!! :-))))))
........
> > -- cvs view, where you can see only branches and tags and
> > see big picture of progress
>
> What does this look like. Can you direct me to some snapshots ?
The best link would be http://www.eclipse.org/downloads/index.php :-)
There is no good pictures of all features because docs are a bit
behind - they are busy with 2.0 release.
> Does it present it on a file-by-file basis or on the whole project ?
> I tried to write such a thing, but stumbled on the fact that tags/branches
> in CVS are defined on a file-by-file basis (even though they are usually
> present in all the files of a project and used more or less consistently
> across those files) and that it's not easy to merge the data into
> a coherent view while still doing something reasonable in the case where
> some things are not coherent (a branch is missing from one of the files,
> or a tag has been moved to another branch on some of the files, ...)
Merging (upon conflict) is human responsibility anyway, but eclipse
gives very friendly visual environment instead of
cvs diff -r B_1_2_20 -r B_1_2_19
AlexV.
> Stefan
Miles Bader
> User friendliness is not about
> a single person it is about its ease of learning and ease of use for a
> population. It could be objectively measured and I'm sure human factors
> people have ways of doing so.
Though true to form, the marketing people will then claim that the
measurements apply universally...
-Miles
--
Somebody has to do something, and it's just incredibly pathetic that it
has to be us. -- Jerry Garcia
Jim Sculley
Don't forget to mention that you can tab from field to field in the
template.
Jim S.
Thien-Thi Nguyen
> No it is not a matter of personal preference. User friendliness
> is not about a single person it is about its ease of learning
> and ease of use for a population. It could be objectively
> measured and I'm sure human factors people have ways of doing
> so.
that you say this shows you know how to express your personal
preference (of this easily-confused term). the population in
question is not homogenous in their preference, learning and usage
patterns, so what single term would suffice?
it all comes down to the single person -- hopefully those who
deign to speak for others are able to handle this simple truth and
incorporate it into their thinking (perhaps during backtrack from
some faux solipsistic conclusion).
thi
Michiel Konstapel
need
> > to iterate over it. I can type for hit ctrl-space and select "for -
iterate
> > over collection". It will insert the following:
> >
> > for (Iterator iter = collection.iterator(); iter.hasNext();)
> > {
> > type element = (type) iter.next();
> > }
> >
> > But it doesn't stop there I can for instance change the name of the
local
> > variable iter in the declaration and it will automatically change the
> > others. The same with the type.
>
> Don't forget to mention that you can tab from field to field in the
> template.
Cool, I didn't know that :)
More ammo in the NetBeans vs. Eclipse war at work. Just kidding of course,
it's more like friendly rivalry with a college of mine. "Ah, but can
NetBeans do X?" "Damn, no! But it does do Y!" :)
Michiel
Jon A. Cruz
> Merging (upon conflict) is human responsibility anyway, but eclipse
> gives very friendly visual environment instead of
> cvs diff -r B_1_2_20 -r B_1_2_19
I just right-click and pick "merge" from the popup menu.
http://www.twobarleycorns.net/tkcvs/tkcvs-work.gif
--
Lee Sau Dan
Eli> Phillip Lord wrote:
>> I would say that something that takes a long time is not
>> particularly easy.
Eli> Alternatively, it might be that there's much to learn,
Eli> i.e. that Emacs is simply a large package with many features.
You don't have to learn all at once! Emacs's packages work very
independently from one another. You just need to learn what you want
to learn. You don't have to learn all before you can do anything
useful. That's the point behind it!
Who dares to claim that he has mastered ALL Emacs packages?
>> From my own memory I remember learning emacs to be a
>> frustrating experience.
Eli> I rather enjoyed it.
Me too! That's why I keep on learning and trying new packages. Emacs
makes learning an anjoyable experience. :)
--
Lee Sau Dan 李守敦(Big5) ~{@nJX6X~}(HZ)
E-mail: dan...@informatik.uni-freiburg.de
Home page: http://www.informatik.uni-freiburg.de/~danlee
Lee Sau Dan
Dale> No it is not a matter of personal preference. User
Dale> friendliness is not about a single person it is about its
Dale> ease of learning and ease of use for a population.
How do you weigh the importants of the individuals in the population?
Do you put a user who uses Emacs 10 hours a day on equal weights as a
user who would only use Emacs for 10 minutes a week?
Dale> It could be objectively measured and I'm sure human factors
Dale> people have ways of doing so.
Given that most heavy_ Emacs users are very satisfied with Emacs, I
don't think you have any grounds for saying it's not "user"-friendly.
Lee Sau Dan
>>>>> "AlexV" == AlexV <avek_...@videotron.ca> writes:
Stefan> I don't believe in templates as far as "creating code from
Stefan> scratch". And this despite my being a slow typist.
Stefan> Typing code is really not the most time-consuming part of
Stefan> programming.
Ever tried M-/ ? It's really COOL! Now, I don't mind using long but
descriptive function and variable names. M-/ removes the penalty in
typing loooooooooooooooooooooooooooooooooooong names! :D
Lee Sau Dan
Sven> But apart from it taking time, it _is_ easy. When you start
Sven> Emacs ( at least up to 20.x) it will tell you:
Phillip> I think this is stretching the definition of "easy" a
Phillip> little too far.
But I did started using Emacs from the Tutorial. Nobody else taught
me how to use Emacs. All I know about Emacs so far (after 10 years)
are done by reading on-line documentations (esp. Info). I don't find
it anything difficult at all.
Phillip> I would say that something that takes a
Phillip> long time is not particularly easy.
Sleeping takes 8 hours for average adults. So, that's not easy?
Do you mean it takes a long time TO LEARN? Well... but nobody has
learnt ALL aspect of English. And English takes years to learn (even
for native speakers). So, English must be more difficult than Emacs.
Phillip> From my own memory I remember learning emacs to be a
Phillip> frustrating experience.
Maybe you didn't do it properly. Secret: It requires patience!
I was patient enough to sit in front of the computer to practise. I
was patient enough to read what's on the *start up screen* (which most
people are, for strange reasons, impatient to read). I discovered C-h
t there. So, I pressed C-h t and the tutorial started. I had the
patience to go through the tutorial. And I had the patience to really
try using Emacs. And I had the patience to learn how to use dir-ed,
ange-ftp, compile, gud, vc, Rmail, GNUS, Gnus, ... all by reading the
Infos.
It's sad that people nowadays no longer have any patience.
Sven> a) able to read (and why else would you use an editor?)
Phillip> Ask someone with a strong visual impairment. Fortunately
Phillip> you can use emacs without being able to read.
Ever heard of Emacspeak?
Sven> This is certainly easier than learning to drive a car, never
Sven> mind riding a bike
Phillip> This is a different point. Cars and bikes are safety
Phillip> critical.
So are computers nowadays, esp. with security-ignorant e-mail clients
such as LookOut!
Phillip> They are not that hard to use,
They are! Many people cannot ride bikes. It's a very difficult thing
for those people.
Phillip> but you have to learn it well as the consequences of
Phillip> screwing up are possibly fatal.
You can ride a bike slowly for safety. Why hurry?
The point is that some people cannot ride a bike or swim at all, even
though it's to others as simple a skill as walking.
Sven> so learning Emacs was time well invested.
Phillip> On this I agree.
Absolutely! I'll work 10 times slower without Emacs.
Lee Sau Dan
>>>>> "AlexV" == AlexV <avek_...@videotron.ca> writes:
>> ---refactoring (various)
Stefan> Can someone describe precisely the kind of refactoring
Stefan> that are offered ?
>> ---templates (from simplest out>>System.out to overriding all
>> methods upon inheritance)
Stefan> I don't believe in templates as far as "creating code from
Stefan> scratch". And this despite my being a slow typist.
Stefan> Typing code is really not the most time-consuming part of
Stefan> programming. I'd rather see progress in the realm of
Stefan> "code syncing" as in "here is a class, here is an
Stefan> interface, please do what is obviously needed so that the
Stefan> class implements the interface".
Why has nobody mentioned JDE (implemented in E-lisp) yet? I used to
use it during 1999--2001 for intensive Java development. Its got this
"code syncing" feature. Given that Java is quite popular and many
people use Emacs to develop Java programs, JDE must have evolved a lot
in the past year.
>> ---almost transparent version control (diff, branching,
>> tagging)
Stefan> VC's support for branching is not too hot, but the rest
Stefan> seems OK.
I find M-x cvs-annotate very amazing. Has any other VC tool come
cloned this feature? And the pcl-cvs package (now comes with standard
Emacs 21) is a pretty good interface for CVS.
Phillip Lord
>>>>> "Phillip" == Phillip Lord <p.l...@russet.org.uk> writes:
Sven> But apart from it taking time, it _is_ easy. When you start
Sven> Emacs ( at least up to 20.x) it will tell you:
Phillip> I think this is stretching the definition of "easy" a
Phillip> little too far.
Lee> But I did started using Emacs from the Tutorial. Nobody else
Lee> taught me how to use Emacs. All I know about Emacs so far
Lee> (after 10 years) are done by reading on-line documentations
Lee> (esp. Info). I don't find it anything difficult at all.
Then you are obviously far more intelligent than I. I bow done to
you. I learnt in the same way. I found it hard.
Phillip> I would say that something that takes a long time is not
Phillip> particularly easy.
Lee> Sleeping takes 8 hours for average adults. So, that's not
Lee> easy?
Pretty crap argument all things considered.
Lee> Do you mean it takes a long time TO LEARN? Well... but nobody
Lee> has learnt ALL aspect of English. And English takes years to
Lee> learn (even for native speakers). So, English must be more
Lee> difficult than Emacs.
Yes. English is very difficult. Your point?
Phillip> From my own memory I remember learning emacs to be a
Phillip> frustrating experience.
Lee> Maybe you didn't do it properly. Secret: It requires patience!
I did it by trial and error. As there was no one around to tell me off
a better way, this is how I did it.
Sven> a) able to read (and why else would you use an editor?)
Phillip> Ask someone with a strong visual impairment. Fortunately
Phillip> you can use emacs without being able to read.
Lee> Ever heard of Emacspeak?
"you can use emacs without being able to read". Yes I have heard of
it.
Sven> This is certainly easier than learning to drive a car, never
Sven> mind riding a bike
Phillip> This is a different point. Cars and bikes are safety
Phillip> critical.
Lee> So are computers nowadays, esp. with security-ignorant e-mail
Lee> clients such as LookOut!
For most people if their computer gets virused to hell no one
dies. Computers are not generally safety critical.
Phil
Sven Utcke
> >>>>> "Sven" == Sven Utcke <utcke...@informatik.uni-hamburg.de> writes:
>
> Sven> But apart from it taking time, it _is_ easy. When you start
> Sven> Emacs ( at least up to 20.x) it will tell you:
>
> I think this is stretching the definition of "easy" a little too
> far. I would say that something that takes a long time is not
> particularly easy.
Hmm. But that certainly is a strange definition --- does that imply
that watching (and understanding) a movie is more difficult than
reading (and understanding) a parabola by Solzhenitsyn, simply because
the movie takes more time? Does it mean that watching a movie is more
difficult then solving a system of differential equations (provided
you can do it in under 90 minutes :-)? I don't think so!
Again, learning Emacs is _not_ fast, it does take some time (although
less than it will save you later on). But it _is_ easy.
> Sven> a) able to read (and why else would you use an editor?)
>
> Ask someone with a strong visual impairment. Fortunately you can use
> emacs without being able to read.
Well, what I meant of course was the requirement for literacy (that's
the right word? Being able to read and write, although now of course
you are going to claim that I not only exclude the visually impaired,
but also many people with other bodily impairments stopping them from
writing).
> Sven> This is certainly easier than learning to drive a car, never
> Sven> mind riding a bike
>
> This is a different point. Cars and bikes are safety critical. They
> are not that hard to use,
You mean it took you longer to learn how to write in Emacs than how to
ride a bike? Or maybe time isn't as important any more as you made it
out to be above --- so you mean it actually took you more effort and
concentration to work through the Emacs-tutorial than to learn how to
ride a bike? Never mind the frustrations incurred, or the injuries...
Well, I can only say that _I_ found learning Emacs a lot easier than
learning how to ride a bike, or how to swim. Cars might indeed be a
different story, but it certainly takes a lot more time to learn how
to drive a car, and it was more stressful --- but that might indeed be
a reflection of the dangers connected with it more than on the actual
difficulties.
> Sven> so learning Emacs was time well invested.
>
> On this I agree.
Well, that's something :-)
Sven Utcke
> "Phillip Lord" <p.l...@russet.org.uk> wrote in message
> news:vf7kknz...@rpc71.cs.man.ac.uk...
> >
> > Emacs is easy to use for a variety of reasons. Of course its true
> > that its powerful. But, for instance, the much derided key control
> > sequences make it very rapid and easy to use. Of course they take
> > time to learn though. These days emacs has a full menu system
> > though, which I use for commands I use less frequently.
>
> Which is simply another way of saying it is not user friendly.
Let's just see whether I got this right:
* If a program has menus only, requiring you to fumble for the mouse
anytime you want to do something more complicated than entering
text, it is user-friendly.
* If a program has menus, just as above, but also key control
sequences to be used a shortcuts, so that I do not need to remove
my hand from the keyboard, and will not suffer Carpal Tunnel
Syndrom, than it is _not_ user friendly by (your) definition.
Did I get that right? Sorry, I think I'll just ignore any further
posts by you an this subject...
Oliver Scholz
> "Phillip Lord" <p.l...@russet.org.uk> wrote in message
[...]
>> Emacs is easy to use for a variety of reasons. Of course its true that
>> its powerful. But, for instance, the much derided key control
>> sequences make it very rapid and easy to use. Of course they take time
>> to learn though. These days emacs has a full menu system though, which
>> I use for commands I use less frequently.
>
> Which is simply another way of saying it is not user friendly.
This is not clear to me. Admittedly some of the more advanced key
control sequences in Emacs are rather obscure. But, heck, the number
of keys on a keyboard is limited. If you have a huge functionality
that you want to make accessible via keystrokes also -- not only via
menus -- then you _have_ to resort to some not-so-uncomplicated
keystrokes. I don't know eclipse, but I dare say: from a structural
point of view there a only three possibilities:
1. an application has only simple and limited capabilities
(cf. Notepad). It doesn't need any complex key command sequences.
2. an application provides a lot of functionality, but not everything
is available via key commands.
3. an application provides a lot of functionality _and_ provides
keystrokes for almost everything.
With (3) it is not avoidable to have complex keystrokes. However, (3)
has no drawback conferred with (1) and (2). Either use the keystrokes
or don't use them.
---
I think the buzzword "user-friendly" is suited only for advertising
and marketing. Or for flame wars, for that matter. I am not interested
in any of these.
"Easy to learn" and "intuitive UI" are another issue. But first of all
I'd like to state:
There is _no_ intuitive user interface. Everything is learned.
I experience this every time I try to teach to some
computer-illiterate the basic usage of MS Windows or MS Word. Many
people that use a computer frequently are accustomed to the MS
standard interface (or the similar Macintosh interface). They tend to
forget that they had to learn the usage of this interface once,
because they have internalized it habitually.
But, mind you, on Unix-like systems most of the standard programs
provide Emacs-like keystrokes like `C-a', `C-f', `C-k', `C-y'
etc. where applicable. So for people with a Unix background the basic
Emacs interface is "intuitive", while `C-x' for killing text or `C-v'
for yanking it is fairly "unintuitive".
There is no good reason to prefer the one over the other. And,
frankly, I don't really think that this is a problem. One gets used to
either the one or the other very quickly. It can be a problem for
people that have to switch between C-x/C-v-style and C-w/C-y-style
applications very often. They can use cua.el which allows
them the C-x/C-v style with Emacs. No, this isn't
"user-friendly". But, again, there are only two possibilities: a) you
can get used to the defaults an application provides for you, b) you
can customize the application to your liking (if it allows you to do
so).
This is the case with Emacs, too. There is a constant discussion about
reasonable defaults for Emacs. That I am not satisfied with the
defaults is something you can measure on the number of lines in my
configuration file. However, even if _I_ were in charge, I would _not_
suggest _my_ customization as defaults for Emacs. The maintainers
strive hard to provide reasonable defaults and in general they do a
very good job at this. So if you don't want to bother yourself with
customization, stick to the defaults. This is the choice _every_ good
application leaves to you (the not-so-good ones don't allow
customization).
Now "easy to learn": indeed, there are a few "strange" things with
Emacs that make it hard for some people accustomed to the MS Word-like
interface to get used to it. We have a very interesting discussion
currently here on gnu.emacs.help about how and to which extend it is
possible to ease the cultural shock for those people. Unfortunately
some (maybe not all) of the difficulties are tied to (powerful)
_features_. Trading features for similarity to other programs is a
no-go. This isn't an option for _any_ non-trivial program. There are a
few points that have to be balanced against each other:
* similarity to other programs against unique features
An example that we discussed in gnu.emacs.help is the way in which
Emacs treats buffers. Here we can do nothing but provide good
documentation and tutorials and say: Get used to it. Once you are,
you will think that everything else is a very strange
limitation. However, there might be other things I am not aware of
where the decision is perhaps not so clear.
* power and ability against obvious simplicity
An example that comes to my mind here: When selecting a file in the
minibuffer, the user doesn't have to kill the default pathname
inserted there, but can start to type an absolute pathname
immediately. I learned this only after I used Emacs for more than a
year. Perhaps I would have discovered it if the default path were
removed immediately as soon as the user starts to type an absolute
path name (as it is the case with XEmacs). So it is a good thing to
remove the default path. However, OTOH a user should be able to
change her decision. So it is a good thing that the default path
stays there. Personally I think the Emacs solution is better then
the XEmacs solution.
* easing the cultural shock for new users against not hiding features
The goal must be to make it easy for new users to become comfortable
with Emacs. This may not be achieved by hiding key features from
them. The learning curve, so to say, should be smoothed, but must
not be flattened.
Those things have to be _balanced_ against each other. There is no
easy decision in one direction or another. IMO the Emacs maintainers
did a very good job at balancing all this. Although Emacs is, of
course, by no means perfect. (Who or what is perfect?)
So, if you can name a few things where Eclipse is better at balancing
those things: please tell us! You could help us to improve Emacs. I am
absolutely sure that the Emacs community would really appreciate some
well reasoned thoughts on how to make Emacs a better program.
-- Oliver
--
9 Messidor an 210 de la Révolution
Liberté, Egalité, Fraternité!
Stefan Monnier <foo@acm.com>
>> preserved or does it just do a pretty good job that you have to
>> double-check ?
> Well... eclipse even compiles changes rigth-away! But human check
> is anyway unavoidable (may be someone wants add error checking
> inside new method or just throw an exception....)
I understand that you generally want to do further modifications.
Is my question more clear now ? I want to know whether the code generated
by Eclipse when extracting a code fragment into a new method is
guaranteed to be 100% equivalent to the original code, no matter how
twisted the code is.
>> > -- cvs view, where you can see only branches and tags and
>> > see big picture of progress
>> What does this look like. Can you direct me to some snapshots ?
> The best link would be http://www.eclipse.org/downloads/index.php :-)
I didn't see any picture of what the CVS view looks like, sorry.
Anything else ?
>> Does it present it on a file-by-file basis or on the whole project ?
>> I tried to write such a thing, but stumbled on the fact that tags/branches
>> in CVS are defined on a file-by-file basis (even though they are usually
>> present in all the files of a project and used more or less consistently
>> across those files) and that it's not easy to merge the data into
>> a coherent view while still doing something reasonable in the case where
>> some things are not coherent (a branch is missing from one of the files,
>> or a tag has been moved to another branch on some of the files, ...)
> Merging (upon conflict) is human responsibility anyway, but eclipse
> gives very friendly visual environment instead of
> cvs diff -r B_1_2_20 -r B_1_2_19
No, you misunderstood. I said "merge the data" not "merge changes".
I'm talking about merging all the revision-and-tag trees of all the files of
a project into a single tag tree. Nothing to do with diff3-style merging.
Stefan
Stefan Monnier <foo@acm.com>
> sequences to be used a shortcuts, so that I do not need to remove
> my hand from the keyboard, and will not suffer Carpal Tunnel
> Syndrom, than it is _not_ user friendly by (your) definition.
Huh? I though that arms movement was good (slows you down, but reduces
the strain).
Stefan
Kai Großjohann
> Merging (upon conflict) is human responsibility anyway, but eclipse
> gives very friendly visual environment instead of
> cvs diff -r B_1_2_20 -r B_1_2_19
Is it better than ediff? M-x ediff-revisions RET...
Kai Großjohann
> Don't forget to mention that you can tab from field to field in the
> template.
ELSE mode for Emacs has this. (Using some other key than TAB, I
guess, but who cares...)
Jon A. Cruz
> Why has nobody mentioned JDE (implemented in E-lisp) yet? I used to
> use it during 1999--2001 for intensive Java development. Its got this
> "code syncing" feature. Given that Java is quite popular and many
> people use Emacs to develop Java programs, JDE must have evolved a lot
> in the past year.
JDE is know known as JDEE, which I had mentioned earlier.
Jon A. Cruz
> Ever tried M-/ ? It's really COOL! Now, I don't mind using long but
> descriptive function and variable names. M-/ removes the penalty in
> typing loooooooooooooooooooooooooooooooooooong names! :D
Actually... M-/ is why I never bothered to get going with any 'intellisense' stuff in Emacs. Not only does
it complete words, but it will then go onto the next most likely word, based on your existing buffers. I
often get a lot of code written very quickly that way.
Stefan Monnier <foo@acm.com>
>> the template.
> ELSE mode for Emacs has this.
And so does tempo and so does skeleton.... I guess I just can't think
of a template package that doesn't offer this functionality.
But admittedly, skeleton and tempo do it rather poorly (it only works
for the most recent template inserted, for instance).
Stefan
AlexV
> AlexV wrote:
>
> > Merging (upon conflict) is human responsibility anyway, but eclipse
> > gives very friendly visual environment instead of
> > cvs diff -r B_1_2_20 -r B_1_2_19
>
> I just right-click and pick "merge" from the popup menu.
>
> http://www.twobarleycorns.net/tkcvs/tkcvs-work.gif
>
Yes, I played with tkcvs for some time - very nice.
(it was tkdiff companion also, rigth?)
unfortunatelly , I can not find links to eclipse cvs pictures.
Eclipse cvs is of the same power level.
AlexV.
AlexV
> Lee Sau Dan wrote:
>
> > Ever tried M-/ ? It's really COOL! Now, I don't mind using long but
> > descriptive function and variable names. M-/ removes the penalty in
> > typing loooooooooooooooooooooooooooooooooooong names! :D
>
> Actually... M-/ is why I never bothered to get going with any
'intellisense' stuff in Emacs. Not only does
> it complete words, but it will then go onto the next most likely word,
based on your existing buffers. I
> often get a lot of code written very quickly that way.
Emacs-style code assist for Eclipse:
http://lunar-eclipse.sourceforge.net/
AlexV
Stefan Monnier <foo@acm.com>
> Eclipse cvs is of the same power level.
That means that Eclipse can only show the tags-and-revisions tree of
a file but not of a project, is that correct ?
Stefan "still looking for a tool that can show the CVS
tags-and-revisions tree of a whole project at a time"
Dale King
news:3D1A1DD3...@abraxis.com...
> Dale King wrote:
> >
> > But it doesn't stop there I can for instance change the name of the
local
> > variable iter in the declaration and it will automatically change the
> > others. The same with the type.
>
> Don't forget to mention that you can tab from field to field in the
> template.
Ooh, I didn't know that!
Unfortunately, while Eclipse keeps getting better and better, the
documentation is lagging behind a bit (not that I would actually read
documentation).
--
Dale King
Dale King
news:vafu1nq...@lucy.cs.uni-dortmund.de...
> "Dale King" <Ki...@TCE.com> writes:
>
> > It is very intelligent templates. For example I have a collection and
need
> > to iterate over it. I can type for hit ctrl-space and select "for -
iterate
> > over collection". It will insert the following:
> >
> > for (Iterator iter = collection.iterator(); iter.hasNext();)
> > {
> > type element = (type) iter.next();
> > }
> >
local
> > variable iter in the declaration and it will automatically change the
And
> > many times it will fill in the variable and type with a nearby
declaration
> > (for iterate over array usually).
>
> It's fairly easy to build skeletons to do that. The skeleton would
> ask for the collection to iterate over or the name of the iterator
> variable.
>
> With a little more work, the collection variable could be glorked
> from context.
>
> There are some code generation wizards in JDEE already, but such an
> iterator construct does not show up there. Any takers?
Well obviously it can be done since Eclipse does it. I'm sure it could be
done in Emacs as well. Note I was answering a specific question about
Eclipse, in particular the notion that templates weren't very useful. They
are very useful when made this intelligent.
--
Dale King
Dale King
news:vd0ptyd...@kogs12.informatik.uni-hamburg.de...
> "Dale King" <Ki...@TCE.com> writes:
>
> > "Phillip Lord" <p.l...@russet.org.uk> wrote in message
> > news:vf7kknz...@rpc71.cs.man.ac.uk...
> > >
> > > Emacs is easy to use for a variety of reasons. Of course its true
> > > that its powerful. But, for instance, the much derided key control
> > > sequences make it very rapid and easy to use. Of course they take
> > > time to learn though. These days emacs has a full menu system
> > > though, which I use for commands I use less frequently.
> >
> > Which is simply another way of saying it is not user friendly.
>
> Let's just see whether I got this right:
>
> * If a program has menus only, requiring you to fumble for the mouse
> anytime you want to do something more complicated than entering
> text, it is user-friendly.
>
> * If a program has menus, just as above, but also key control
> sequences to be used a shortcuts, so that I do not need to remove
> my hand from the keyboard, and will not suffer Carpal Tunnel
> Syndrom, than it is _not_ user friendly by (your) definition.
>
> Did I get that right? Sorry, I think I'll just ignore any further
> posts by you an this subject...
I am exiting this thread as I know how pointless it is to argue with Emacs
zealots. Emacs is great and powerful, it isn't user friendly however.
Let's see whether I got this right:
Any program that takes a long time to be proficient in and one where anytime
you have to do something new you have to go figure it out is user friendly.
It makes no difference how long it takes to become proficient as long as
someday you can become proficient.
By this defintion command line interfaces are user friendly as well.
--
Dale King
Ole Arndt
> "Dale King" <Ki...@TCE.com> writes:
>
> It's fairly easy to build skeletons to do that. The skeleton would
> ask for the collection to iterate over or the name of the iterator
> variable.
>
> With a little more work, the collection variable could be glorked
> from context.
>
> There are some code generation wizards in JDEE already, but such an
> iterator construct does not show up there. Any takers?
It's there. See jde-gen-cflow-for-iter.
--
Ole Arndt
-------------------------------------------------------
I waited and waited and when no message came I knew it must be from you.
Alan Shutko
> I am exiting this thread as I know how pointless it is to argue with Emacs
> zealots.
Correct. It's much more useful to suggest specific improvements, or
specific areas in need of improvement. They have a very high
possibility of being implemented, while arguing about the definition
of "user friendliness" is a waste of time.
--
Alan Shutko <a...@acm.org> - In a variety of flavors!
"I didn't know it was impossible when I did it."
Henrik Enberg
> I am exiting this thread as I know how pointless it is to argue with Emacs
> zealots. Emacs is great and powerful, it isn't user friendly however.
Funny that. I was just thinking how the "If it isn't done like
Apple/Microsoft GUI guidelines says it isn't user friendly" zealots is
utterly incapable of understanding that _all_ interaction with a
computer is learned.
The last time I had the misfortune of using Microsoft Word, I kept
popping up the print dialog 8 times a minute because it is bound to the
key I reach for instinctively when I want to move the cursor up one
line. Does this mean that Word is objectively user unfriendly?
> Let's see whether I got this right:
>
> Any program that takes a long time to be proficient in and one where anytime
> you have to do something new you have to go figure it out is user friendly.
No, what makes you think that?
> It makes no difference how long it takes to become proficient as long
> as someday you can become proficient.
Of course it does.
> By this defintion command line interfaces are user friendly as well.
Some are, but I know that it is hard to grasp for one-track minds.
--
Booting... /vmemacs.el
Robert Uhl <ruhl@4dv.net>
>
> I am exiting this thread as I know how pointless it is to argue with Emacs
> zealots. Emacs is great and powerful, it isn't user friendly however.
It's user friendly; it's just not learner-friendly. There is a
difference. As long as the learning period is short relative to the
using period, I'll take user-friendly over learner-friendly. As long
as the using period is short relative to the learning period, I'll
take learner-friendly.
A utility used once a year better have a nice, simple idiot-proof
interface. Something I use every day had better get out of my way and
let me do my thing. Emacs does not have an idiot-proof interface.
But it gets out of the way when I use it. And it offers me
capabilities that _no_ idiot-proof interface does, or can.
But when _I'm_ the idiot, I like the idot-proof interface...
> By this defintion command line interfaces are user friendly as well.
They are. Back when I was a Mac zealot, I never thought it could be
true, but command line interfaces can really be nice. Probably what
is needed is a fusion of the two, but that's another flamewar:-)
--
Robert Uhl <ru...@4dv.net>
The Lord of the Rings is a million times more interesting than mere
literature, which is why mere literary critics cannot get to grips with
it. --Robert McNeil
Miles Bader
> I am exiting this thread as I know how pointless it is to argue with Emacs
> zealots.
Ah, `retreat and declare victory'...
> Emacs is great and powerful, it isn't user friendly however.
Actually, it is.
Cheers,
-Miles
--
Is it true that nothing can be known? If so how do we know this? -Woody Allen
Galen Boyer
> Who dares to claim that he has mastered ALL Emacs packages?
I haven't mastered _one_ Emac's package, because that would have meant
that I would have mastered all of Emacs.
--
Galen deForest Boyer
Sweet dreams and flying machines in pieces on the ground.
Galen Boyer
> A utility used once a year better have a nice, simple idiot-proof
> interface. Something I use every day had better get out of my way and
> let me do my thing. Emacs does not have an idiot-proof interface.
> But it gets out of the way when I use it. And it offers me
> capabilities that _no_ idiot-proof interface does, or can.
>
> But when _I'm_ the idiot, I like the idot-proof interface...
Very succinct. Well said.
AlexV
"Alan Shutko" <a...@acm.org> wrote in message
news:87elesn...@wesley.springies.com...
> "Dale King" <Ki...@TCE.com> writes:
>
> > I am exiting this thread as I know how pointless it is to argue with
Emacs
> > zealots.
>
> Correct. It's much more useful to suggest specific improvements, or
> specific areas in need of improvement. They have a very high
> possibility of being implemented, while arguing about the definition
> of "user friendliness" is a waste of time.
I strongly support suggestion to make positive statements:
Emacs is great powerful __universal__ environment !!
Java community has several powerful java-oriented
environments that __offer more__ for java programming.
And lets leave issue about user-friendlyness (sic-?)
to everybody's personal opinion.
For me, Emacs is hard to learn, easy to use, and
Eclipse/JBuilder better for java development.
AlexV.
Galen Boyer
> I know how pointless it is to argue with Emacs zealots.
Give it a break. Your as much more an anti Emacs zealot than Emacs
users are zealots.
The Emacs guys that help me learn how to learn this software want to
hear how to improve the software, not argue semantics. Who gives a damn
about your definition of user-friendly. Say how to make Emacs easier to
use in your eyes and you might see it implemented, but don't bring in
some silly buzzword, user-friendly, and then assume the definition is
found in your interfaces. What makes your interfaces easier to use than
Emacs is the question. How to make Emacs easier to use than your
interfaces so you will use it, is the challenge that you could put forth
and if your interfaces are actually easier, it will be accepted.
Thien-Thi Nguyen
> I am exiting this thread as I know how pointless it is to argue with
> Emacs zealots. Emacs is great and powerful, it isn't user friendly
> however.
the point of arguing w/ emacs zealots (some of whom hack emacs) is to
develop your own emacs zealotry. you have it in you (the zeal part, at
least) -- just a matter of practice.... good luck!
thi
Lee Sau Dan
Galen> On 27 Jun 2002, dan...@informatik.uni-freiburg.de wrote:
>> Who dares to claim that he has mastered ALL Emacs packages?
Galen> I haven't mastered _one_ Emac's package, because that would
Galen> have meant that I would have mastered all of Emacs.
Neither have I. I just use the coolest parts of every package I'm
using. And that's already far more than what I can do with those
eye-calories (caution: candies carry calories!). :)
--
Lee Sau Dan 李守敦(Big5) ~{@nJX6X~}(HZ)
E-mail: dan...@informatik.uni-freiburg.de
Home page: http://www.informatik.uni-freiburg.de/~danlee
Lee Sau Dan
Jon> Lee Sau Dan wrote:
>> Ever tried M-/ ? It's really COOL! Now, I don't mind using
>> long but descriptive function and variable names. M-/ removes
>> the penalty in typing loooooooooooooooooooooooooooooooooooong
>> names! :D
Jon> Actually... M-/ is why I never bothered to get going with any
Jon> 'intellisense' stuff in Emacs. Not only does it complete
Jon> words, but it will then go onto the next most likely word,
Jon> based on your existing buffers. I often get a lot of code
Jon> written very quickly that way.
Yeah. That's why I now write programs with
longLongLongLongIdentifiersWithoutWorryingHowToTypeThem. And I'm
really beginning a bad habit of making the first 2 or 3 letters of
them different, so that I can quickly type them with the M-/ feature.
And I really wonder how my colleagues show are working on the same s/w
project with me and still not using Emacs could tolerate this! I'm a
touch-typist with a peak rate of 60wpm, whereas they aren't!
These days, I'm have been writing Prolog programs, where the "culture"
is to use cryptic and short variable names, such as "append(X,Y,Z) :-
...". I just find this stupid, and hence I wrote "append(Part1,
Part2, Appended) :- ...". It isn't harder to type (thanks to Emacs),
but much easier to read!
The downside: when typing documents (LaTeX), I'm also tempted to type
long words with M-/. That's normally not a problem. However, when
I've made a spelling mistake, it would "propagate" through M-/. :(
Lee Sau Dan
Lee> But I did started using Emacs from the Tutorial. Nobody else
Lee> taught me how to use Emacs. All I know about Emacs so far
Lee> (after 10 years) are done by reading on-line documentations
Lee> (esp. Info). I don't find it anything difficult at all.
Phillip> Then you are obviously far more intelligent than I. I bow
Phillip> done to you. I learnt in the same way. I found it hard.
That's not intelligence, but diligence.
Everyone can do it. Just a matter of the learning attitude. If you
don't want to learn something, it will never be easy. If you want to
learn something, it will never be hard.
Lee> Do you mean it takes a long time TO LEARN? Well... but
Lee> nobody has learnt ALL aspect of English. And English takes
Lee> years to learn (even for native speakers). So, English must
Lee> be more difficult than Emacs.
Phillip> Yes. English is very difficult. Your point?
So, please shout: "Make English easier, please! It's no point taking
years to learn this language. Please make it easy to learn, so that
anyone can use it in an hour!"
The point is: Emacs is difficult to "newcomers", but not so for those
use are fluent in it. There is not much intention among the Emacs
community to make it easier, esp. when doing so would penalize the
productivity of existing power users.
Phillip> From my own memory I remember learning emacs to be a
Phillip> frustrating experience.
Lee> Maybe you didn't do it properly. Secret: It requires
Lee> patience!
Phillip> I did it by trial and error. As there was no one around
Phillip> to tell me off a better way, this is how I did it.
I also did it by trial-and-error. But apparently, my trial-and-error
method is different from yours. When I make mistakes, I learn from
them! I go to read documentation, study the subject, and come out
with the solution. I don't just sit there crying, just to get
attention from the adults, hoping that they'd help me. I go to LEARN
more to find a solution.
Lee> Ever heard of Emacspeak?
Phillip> "you can use emacs without being able to read". Yes I
Phillip> have heard of it.
So, please try Emacspeak. That's designed (primarily) to serve the
visually impaired or the blind population. Of course, you can use it
even when you have perfect vision.
Lee> So are computers nowadays, esp. with security-ignorant e-mail
Lee> clients such as LookOut!
Phillip> For most people if their computer gets virused to hell no
Phillip> one dies. Computers are not generally safety critical.
When people where spreading hypes about how the Bluetooth technology
would let you control your fridge, washing machine, stove,
blahblahblah from your PC, do you still not think it's totally
unrelated to safety?
Thien-Thi Nguyen
> The downside: when typing documents (LaTeX), I'm also
> tempted to type long words with M-/. That's normally
> not a problem. However, when I've made a spelling
> mistake, it would "propagate" through M-/. :(
maybe a new root set would help. put this in your ~/.emacs:
(find-file-noselect "/usr/share/dict/words")
munge path to taste.
thi