I read with interest
and
some amusement David Orme's blog.
With regards to his comment about me being "hot and bothered" - I'm not
hot and
|
|
bothered - after all I'm not the one who has to fix all the SWT
bugs.
And as for Mike Milinkovich's attributed comments - I'm sorry he is
disappointed in me - however it is a misplaced disappointment - as most
of the comments in my blog come directly from Eclipse developers...so
his disappointment should probably rest with the current state of SWT and with Eclipse's lack of some key features. Perhaps his position precludes him from voicing that
disappointment.
As anyone that has been watching the blogs lately can attest to - many
Eclipse developers who
have looked at NetBeans 4.1 Beta have a lot of positive things to
say. This culminated yesterday with a very nice review from an
Eclipse developer/author who wrote a review
of NetBeans 4.1 Beta and gave it the following high praise :
| All
arrows point towards go on this new revision of NetBeans. Increased
performance, cool new features, integrated Ant building, support for
Java 5.0—the list goes on. Unless you are one of those
faithful-to-the-end patriots, any Eclipse user—or IntelliJ IDEA user,
for that matter—should waste no time in evaluating this new version.
That said, it is important to keep in mind that Eclipse still has a
huge marketshare, is a decent IDE, has a host of plug-ins available and
a huge mass of SWT applications that have been developed. So, try it
for yourself, compare, and let's talk about it. |
Please go read the entire review here,
it really gets to the heart of what is happening with many Eclipse
developers and why they are trying NetBeans 4.1. Please don't believe me
and I encourage you to discount David's flawed arguments -
read their words. Shortly before this
appeared - another Eclipse developer focusing on GUI development and
wrote the following report
which is definitely worth reading. In the report he stated that
NetBeans 4.0 was the best free/open source integrated solution and
wrote that Eclipse Visual Editor was not ready for prime time. Other
Eclipse developers have already bolted and are today using NetBeans
instead of Eclipse. None of this is a fabrication - you can go
check it out here.
There are a significant number of Eclipse developers which have come
out and now said they are also using NetBeans for some of their
work.
David's blog
has tried to call what I wrote into question - not with any
counter-examples - rather simply by offering up questions and providing
an unfortunate marketing quote by the Director of Eclipse.
Let's look
at some of the questions :
| They
have claimed that the switch to NetBeans was from Eclipse
when the original blog author said nothing of the sort! (See here and here.) |
Please note that the author of Coding.Mu mentioned that he was a
(NetBeans) convert here.
I did something that the blog writer posing the question obviously did
not do, I went back before I wrote the original blog entry and
looked at what he was using. The author of coding.mu
mentions that he has
is running Eclipse clear back here
and here he speaks of his
use of MyEclipseIDE on Eclipse and here he mentions his use
of TruStudio plugin on Eclipse. He ends this by pointing to
the switch site and mentioning that he is a convert.
Feel free to make up your own minds by examining the entries - he was
using both Eclipse and
NetBeans prior to apparently making up his mind. So David, I guess,
failed to take in the complete picture - and jumped to his erroneous
judgement.
David's blog continues :
They post articles about “Why Eclipse Developers Are Moving to NetBeans”
with no market research (or even download) numbers to support that the thesis is true to begin with!
All this when only six individuals are listed on their
"Switch to NetBeans :
Real Stores” web site (compared with nearly
every major industry company joining Eclipse).
|
Get a grip! I started this by giving you a context - Eclipse developers
are reporting and authoring papers that they like what they see in
NetBeans. Instead of attacking the words of many Eclipse developers you
should be listening to them. One more point the last survey done showed
NetBeans growing at a hefty 35%.
That was before NetBeans 4.1 and before 4.0 was released! My blog
entry describes what
appears to be a relatively new trend. You may disagree with this
but it is very difficult to disagree that Eclipse has attracted
considerable genuine dissatisfaction among some Eclipse
developers who have been very vocal on Eclipse Bug reports, blogs,
their evaluations of NetBeans, JavaLobby discussions, and some
have now announced they are switching to
NetBeans (the NetBeans switch site) and there are other anecdotal
points of
references. Are all these Eclipse developers astroturfing as well
? Complaining about lack of market studies is silly - because
if there had been a new one - then David might discount it
anyway. And the one that does exist confirms that
Eclipse, NetBeans and IDEA all grew last year. The reality is that most
of my blog was quotes from Eclipse
developers - not from me. This is perhaps the most damning
aspect of Eclipse's problem. It is what I and others call the toppling-over point.
Eclipse has grown at a very fast and at an unsustainable rate.
The
reason it is unsustainable is because reality is now catching up to
marketing. Some Eclipse users that remember Eclipse 2.1 are
now complaining about sluggish performance, lack of scalability,
incompatibilities between platforms, poor look and feel with the native
toolkit and usability issues. Eclipse has morphed into a
different form that has considerable issues. SWT has large
incompatibilities
between platforms (SWT-AWT on MacOS X to name one) and has very few
apps. The RCP locks-in developers into a structure that uses
exclusively Eclipse.org forms that leverage a non-standard
toolkit. By the way - the NetBeans Switch site
is a new site and is only one
place to look - there are
developers that are blogging about their efforts with NetBeans and I
have shown you two Eclipse developers that have authored two articles
which are extremely positive about NetBeans features.
On the issue of some companies joining Eclipse.org- I think this
deliniates the Eclipse.org dilemma. As different members with
conflicting interests join it is clear Eclipse has a big
challenge. This will become the problem that Eclipse.org will
have to come to grips with. With SWT becoming increasingly a
liability and the source of some of Eclipse's problems - it will be
interesting
to see how companies deal with some of the most basic issues.
Will they re-invent Swing and Java2D completely and at the same time
bumping up against many of the problems with native platform issues
that cause SWT code to be unportable ? As both Eclipse and SWT
grow it really isn't a surprise that many
developers that use Eclipse are now beginning to recognize that there
are many problems - Eclipse has become slower
since 2.1, SWT has some serious cross-platform problems and the
dilemma is that as more
features need to be added to the platform the larger and more bloated
it gets. All of this just to be able to compete with
the strong Swing-based IDEs that are now out-performing it and offering
a better out-of-box experience. It is ironic that many of the
advertisements for SWT are coming back to haunt it. Reality is
now intruding into the marketing of SWT and Eclipse.
Now let's examine Mike Milinkovich's (Director - Eclipse.org)
attributed comments :
Sun is making a transparent and somewhat desperate attempt to
counter
all of the great news that came out at EclipseCon. With BEA,
Borland, Wind River,
Sybase and Computer Associates joining as
Strategic Developers, the industry has spoken.
To paraphrase Bill
Clinton:
it’s the ecosystem, stupid.
Desperate people do strange things. However, it is disappointing that
Sun has chosen the low road.
|
Spoken in true marketing-speak.
I think the empty comment "the industry has spoken" is actually
over-used and often used in silly ways. However, since he
likes
the comment all I can say is the JCP has 876 members - most of them are
the same
large companies Mike listed. The actions of the of JCP results in
real
standards for J2SE, J2EE and J2ME as well as API standards such as
Swing. The Java industry has not only spoken but it is activily
implementing Java standards through the JCP.
Eclipse.org may be the biggest marketing hallucination of the moment -
one that serves primarily to benefit.... IBM. It is also a
hallucination that
Eclipse.org can afford to avoid making Eclipse a fast performing,
feature-rich and a
compatible IDE and still be successful. The same phenomena that
allowed Eclipse to grow can also crush it. All the basic
Eclipse.org
assumptions collapse if NetBeans or for that matter any other Java IDE
offers a better developer experience. And let me paraphrase
something Bill Clinton said back to Mike : It's the developer, stupid.
Eclipse today is behind NetBeans in
offering J2SE 5.0, full J2EE, GUI builder, a better mobility
solution, an advanced profiler, an intuitive UI and fast
performance - out -of-the-box. In addition, Eclipse
developers are also concerned about locking their applications to a
non-standard SWT toolkit with limited portability. I can
understand why Mike
would prefer to turn to talking about "low roads" instead of addressing
the difficult problems (like SWT-GTK performance, Motif L&F
rendering or SWT-AWT on the MacOS X) that Eclipse.org faces.
However, in retrospect, I
think the low-road was taken awhile back by the attempt to fracture the
Java community with SWT in the first place. Both Swing and
NetBeans are both growing. Swing even on the Eclipse IDE.
NetBeans by last count at a 35% growth rate. This is hardly the
desperate situation Mr Milinkovich would like
people to believe. Quite the contrary. Before NetBeans 4.0
was released and prior to NetBeans 4.1 Beta's release - NetBeans was
growing at 35%. It isn't an accident
that NetBeans now has an Import Eclipse Project plugin - it was created
to make the transition easier - and Eclipse developers are using
it. NetBeans is now ready to grow much faster. |
|
