Re: Oracle buys Sun, owns Java

cbossens73@xxxxxxxx wrote:
On Apr 21, 2:26 am, Arved Sandstrom <dces...@xxxxxxxxxxx> wrote:
...
I'm sure I can think of more. Point of the exercise being, if Eclipse
3.4 or NetBeans 6.5 (which are the versions I use at present)
disappeared tomorrow, I would not be inconvenienced. I don't mean not
_seriously_ inconvenienced, I mean at all.

Well, from a "text editor" point of view, I wouldn't
miss much.

Actually when I'm using IDEA I feel that the "text editor"
is really lame compared to Emacs and from time to time
I open a .java file from Emacs to do some "Emacs magic"
on them.

Before I answer some (or all) of the points you make below, let me reiterate that I was posing a thought experiment. It is strictly applicable only to me, because I know what I do with IDEs. For someone else it may well be that some things they do in IDEs are not so convenient otherwise.

However what I'd completely miss are things like:

- IDEA telling me "xxx is unlikely to have useful
semantics" or "condition xxx is always true/false" (when
making dumb/obvious mistake) or "@NotNull annotated method
xxx can return null"... IDEA really shines for real-time
code analysis (especially on incomplete/non-compilable
.java file) and this is very powerful. You cannot easily
replicate this (how would you go about analyzing a source
file that you can't even compile ?).

I use FindBugs myself, as often if not more often outside Eclipse than in the IDE. As for trying to static analyze a source file that doesn't even compile yet, I can't say I've ever wished to do that.

- all the "programming by intention" that IDEA has (in
version 8 it just got better).

I am not familiar with the IntelliJ interpretation of "programming by intention". I _am_ reasonably familiar with what it means in XP, although to be honest I don't know why it's associated with XP, since I would like to think that most of what is in "programming by intention" is done by any programmer doing a decent job.

- integrated Emma code coverage. How would I go about
replicating this in, say, Emacs or vi or JEdit?

I've used Emma in conjunction with JUnit for test coverage, all run through Ant under the control of Hudson. So I'm reasonably familiar with it. My first question would be, why does it matter what initiates the execution of instrumented code? You're going to get Emma code coverage results no matter what...

- integrated Dependency Structure Matrix, integrated
cyclic dependencies
finder, I'd miss these and not know where to look to
replicate this.

Point taken. I don't use this stuff, you do.

- powerful GUI editor hiding boilerplate Swing code for me,
yet allowing me to easily work on the important stuff

Point taken. I haven't done non-web Java GUI work since maybe 2001-02. The only non-web GUI work I've done since has been on the M$ side, and quite frankly I wouldn't want to give up Visual Studio for that.

Some of this can probably be made to work by using this and
that technology, but with a good IDE it's all conveniently
integrated.

I'd miss that.

This is absolutely my own opinion, but to me a full-fledged IDE is integrated in the same way that a Walmart is integrated compared to a dozen stores in a shopping plaza...I have to walk as far, my individual purchases are still independent, etc. If I were to set up my terminal windows and text editors as I saw fit I'd have the shopping plaza.

The operations I do in NB or Eclipse are usually quite independent. The "Integration" in IDE simply means that I can access operations through a single application. But that's usually all it means.

Not having real-time code analysis on .java file
I'm currently editing (and which hence are often in a
"non-compilable" state) and not having all the "programming
by intention" features that IDEA offers would be like
going back to the stoneage for me.

They market it as "Intelligent Coding Assistance" and
"On-the-fly code analysis" which look like buzzwords.
But they really aren't.

It works and it works darn well.

These features were made by very smart people, are
incredible time savers (and for time is money, people
are willing to pay big bucks to have these features),
and there's no way I'm replicating this with any grep/find
command-line-fu

:)

I believe you when you assert that these IDEA features help you. I'm not arrogant enough to figure that what I need or find helpful is the same for everyone.

AHS
.

Relevant Pages

  • Re: RosAsm disassembler output vs. IDA Pro
    ... IDE it seems quite a good thing, so we may take some ideas into LuxAsm, if it ... But what I mean is, that when I'm used to a particular Editor, maybe I don't ... But you still have the integration. ... >features they have faster learned than forget. ...
    (alt.lang.asm)
  • Re: Downgrading to D7
    ... I'm not worried about IDE or VCL features, ... to spend so much money on it and being unable to used it. ... I don't buy clichet like "...But you wont be helping the delphi community, industry, ...
    (borland.public.delphi.non-technical)
  • Re: Questions about the DLR
    ... References, declarations, etc. were all indexed, and there was even a macro that would refer to the source management database and show you the exact check-in and diff where the line was added or changed. ... I can see how the features you're talking about would be useful. ... Even the stuff that can be measured in seconds, I'm just not compiling often enough for it to be a big deal. ... obviously this is as much about .NET as it is about a specific IDE. ...
    (microsoft.public.dotnet.framework)
  • Re: C# vs VB.Net
    ... In VB.NET you have to ~remember to put in "namespace ... Also, in VB 6, you could have the IDE ... Status unknown - I haven't used the serialize features of dotNET. ...
    (microsoft.public.dotnet.framework.aspnet)
  • Re: C# runtime version
    ... >> 2) Visual Studio Team System provides many features which are ... All I want is for the IDE to be able to target a compiler for one ... I would take the code and try to compile it in the 1.1 ...
    (microsoft.public.dotnet.languages.csharp)