Java and Mono (was Re: Managing multiple computers at home)

Bryan Sant bryan.sant at
Mon Mar 27 14:12:48 MST 2006

On 3/27/06, Michael L Torrie <torriem at> wrote:
> How is this handled on linux?  Are they going with GTK, Qt or what?


> Will it pick up my gnome themes (like Java 1.5 tried and failed to do)?


No kidding about 1.5.  I was very unimpressed with 1.5's attempt at
this.  I'm running the latest Mustang beta and it does great with the
default Ubuntu theme.  I've already had it choke on a 3rd-party
Metacity (that's the title bar) theme though.  But this is only a
problem in an MDI app.  At least what the Mustang dev team is saying
is that "native fidelity" is tippy-top priority and that Swing's GTK
emulation will be pixel perfect before Mustang ships.  We'll see.

Take a look a mustang's current rendering of GTK:

> SWT has a variety of "native" look and feels for linux including Gtk and
> Motif.  Will Java adopt a similar strategy?

Java's Swing already has a Motif Look and Feel.  So you can see that
today.  The GTK look and feel in mustang seems to be coming along too.
 So yes, Swing can do the same that SWT can do (visually) and more
(there are many other Swing look and feels that you can choose from
that SWT can't).

> Does anyone here have an opinion of SwingWT?

I'm a developer for SwingWT.  It is great.  But the project seems to
have cooled down.  Fewer people are interested because Swing is
getting to be so fast, many are asking "What's the point of having SWT
back the GUI when Swing is just as fast and there are no compatibility

If you absolutely hate Swing, then I still think that SwingWT is a great option.

> I'm not sure how I feel about the layout manipulation here.  It is kind
> of neat that you can lay out form elements without worrying about or
> messing with layout containers and still have it resize and re-lay out
> the elements properly.  I can see how this could be easier to use, but I
> wonder how well the layouts work when the UI designer isn't involved
> with setting up how the layout managers actually work.

You'll have to use it and see what you think.  I had the same
concerns.  You can still use all of the old layout managers in the GUI
builder (which I often do).  But now you can use the GroupLayout
thingy that lets everything snap together for dialogs and other things
that have been difficult to do with the old LayoutManagers.

But once you use the new GroupLayout with the Netbeans builder, you'll
realize that it "Just Works" (tm).  It does things right and the the
resizing issues are all handled.

> I somehow doubt that.  Win32 and MFC still reign supreme from what I can
> see.  Or other thin wrappers around win32.  Any article that says "...
> Swing ... has surpassed WinForms as the dominant GUI development
> toolkit" indicates to me that this is all meaningless.  The number of
> real-world apps out there actually using WinForms and .NET at this point
> is pretty minuscule.  And since .NET uptake among large windows app
> makers seems to be pretty small and immature at this point, obviously
> Swing is more widely used than WinForms.  In short the blog entry below
> is BS.  I guess what he meant to say was that in the world of Java
> vs .NET, Swing is more widely used than WinForms.

When you say "he meant to say" you're referring to Evans Data Corp
right?  This isn't some bloggers opinion.  This is the output of a
research study.  I don't know if Win32 was excluded from the study. 
However, I really doubt that much of any *new* development is done in
Win32/MFC.  Even in MFC's hay day, there was more GUI apps built with
Visual Baisc than MFC.  Now all those old VB developers are using .NET
and WinForms.  And those guys are getting beat by Swing.

> architecturally or from the programmer's point of view.  It's just the
> end-user's point of view that sucked in the past.

Amen.  I really don't care about the niceness of a GUI toolkit to the
developer (well I do care, but it's less important).  I want my users
to think that the finished product is the coolest thing ever.

> Somehow I wonder if the high IT salaries in general will last for many
> years.  I know that BYU, for example, is starting to really feel the
> effects of high IT costs.

Supply and demand.  The demand is high.  It doesn't really matter what
BYU "feels".  The reality is that IT people demand a certain price
because they themselves are in demand.  If what you're saying is that
you think that local IT folks are going to be in less demand soon, I

> I think that one would be foolish to have a skill-set exclusively in one
> language.  Being able to hop between Java, C++, Python, C#, and whatever
> language of the day is out there is probably a pretty important thing
> long-term.  I have a goal to learn one new language every year and to
> write some large program in it.  This year I'm doing python.  Next year
> may be C#.

I think a true geek is compelled to learn new languages and
technologies.  It will help their overall skillset and employability
to some extent, but I don't think that it's as necessary as you're
describing.  I happen to know several languages, but I primarily use
Java.  When Java dies (about 8 to 10 years from now), then I'll
primarily use language X.  There is some need to "hit the ground
running", but throwing a lot of effort after the lang of the day, is
mostly a waste of time when it comes to converting that skill into
dollars.  There is little-to-no market demand for many languages that
might otherwise be interesting to learn.

If you're a professional Admin, then learn all the langs you want. 
Use what works best for you.

If you're a professional software developer, learn a specific language
intimately and get those "Senior" positions.  Learning many languages
will not help your paycheck in most cases.


More information about the PLUG mailing list