So, nobody uses Java on the desktop, right?

[Fabrizio Giudici]

In this mailing list the assertion "nobody uses Java on the desktop"
frequently appears. Taken as a general statement, it is false: the
industry uses lots of Java on the desktop. But, agreeing to restrict the
assertion to the usage of end users on the web, it's reasonable. I usually
even agree with this latter interpretation.

But sometimes there's a little evidence that "nobody" is an exaggerated
term even in this interpretation. For instance, the latest bunch of Java 6
-> Java 7 upgrades is triggering problems where you won't expect: end
users on the web, the thing that shouldn't exist. For instance:

1. I've just learned that the U.S. judiciary system uses Java for
uploading documents: e.g. https://ecf.cadc.uscourts.gov/
2. Italian business consultants must use Java in various way to upload
documents, e.g. signed ones:
http://forum.commercialistatelematico.com/altri-argomenti/53284-problemi-con-firma-comunica-per-colpa-di-java.html
(sorry, this obviously is in italian)
3. I've read in a mailing list of a developer who updated to Java 7 on Mac
OS X, and then the *applet* used by his home banking service broke.

(I'm excluding partial uses, such as applets used to upload files in a
number of web sites, as they are just a minor part of the functions of the
site and usually there's a HTML-only alternative).


Clearly, these aren't tons of uses. But definitely more than nothing.
Of course, if Java keeps on creating troubles when upgrading, the number
will eventually drop to zero. But so far, it isn't.



--
Fabrizio Giudici - Java Architect @ Tidalwave s.a.s.
"We make Java work. Everywhere."
http://tidalwave.it/fabrizio/blog - fabrizio...@tidalwave.it

[Josh Berry]

On Fri, Oct 19, 2012 at 3:09 PM, Fabrizio Giudici
<Fabrizio...@tidalwave.it> wrote:
> Clearly, these aren't tons of uses. But definitely more than nothing.
> Of course, if Java keeps on creating troubles when upgrading, the number
> will eventually drop to zero. But so far, it isn't.

What are you arguing? There are people that are still using MSDOS, as
well. By and large I think most are comfortable including that group
of people in "nobody" though. The argument I've always had was
essentially, if you don't know why you have the JVM installed
(majority of folks that had it when it was a "default" item), than you
should probably not have it installed.

Just like, by and large, it is probably safe to say that games just
aren't written in Java. There are notable exceptions, but the fact
that they are notable makes them perfect for "the exception that
proves the rule."

[Ricky Clarkson]

Besides usual development tools like IDEs, I have SQuirreL open right
now, a Swing app. The support team for our group use Peregrine
Service Client, an Eclipse-based product. Cisco VPN software still
tends to be downloaded through an applet when you connect to the VPN
server.

My previous company continues to use a Swing app as the viewer for the
security cameras they produce, including recording the video an
operator looks at when handling alarms, doing a kind of binary search
to find when something appeared or disappeared, burning footage to
CD/DVD for evidence. They also use an applet based on the same
codebase and that doesn't cause as many problems as you might imagine.

I can still knock out a usable Swing application in less than half the
time and fuss it takes a typical web developer, and with some effort
it can even look good. I tend to do that for small tools that support
staff use etc., but written in such a way that it should not be
difficult to convert to a webapp if that's ever needed. I think that
compared to other desktop technologies Swing remains competitive,
though .NET tends to be slightly more responsive. The C/C++ libraries
like GTK and Qt are doing well and gaining penetration outside of
their native Linux but I wouldn't recommend them for anything that
those languages don't give any benefits for.

> --
> You received this message because you are subscribed to the Google Groups
> "Java Posse" group.
> To post to this group, send email to java...@googlegroups.com.
> To unsubscribe from this group, send email to
> javaposse+...@googlegroups.com.
> For more options, visit this group at
> http://groups.google.com/group/javaposse?hl=en.
>

[Kevin Wright]

Now would be a good time to point out the etymology of the word "prove"!

Contrary to current usage, its original meaning is "test".  So a "proof" is just a test, and not necessarily one that has been passed!

http://en.m.wiktionary.org/wiki/prove#section_1

[Fabrizio Giudici]

>> What are you arguing? There are people that are still using MSDOS, as

Well, your counter-argument seems misplaced. I've not referred to minority
groups that like to play vinyl cd for pleasure, but I cited two groups of
professionals that *have* to use Java for their daily job. I suppose there
are many people in the U.S. working in the judiciary system, right? Hell,
we've got plenty of US-made TV series where the main character is a
lawyer. :-) Do you think they comfortably fall in the "nobody" category?

[Josh Berry]

Of course, you are dodging the history of the phrase.
http://en.wikipedia.org/wiki/Exception_that_proves_the_rule

That is, the meaning is exactly as I am using it. The fact that we
consider Minecraft exceptional for being written in Java strengthens
the general rule that games are not written in Java.

[Robert Casto]

Have you seen the systems used by auto dealers? All they are is a pretty face on a green screen scraper. Those systems haven't changed in years and it doesn't look like they might ever. They are quite complicated and they seen no benefit to changing the technology used to develop the tool or make changes.

--
You received this message because you are subscribed to the Google Groups "Java Posse" group.
To post to this group, send email to java...@googlegroups.com.

To unsubscribe from this group, send email to javaposse+unsubscribe@googlegroups.com.

For more options, visit this group at http://groups.google.com/group/javaposse?hl=en.

--
Robert Casto
www.robertcasto.com
www.sellerstoolbox.com

www.lakotaeastbands.org

[Josh Berry]

On Fri, Oct 19, 2012 at 4:15 PM, Fabrizio Giudici
<Fabrizio...@tidalwave.it> wrote:
> Well, your counter-argument seems misplaced. I've not referred to minority
> groups that like to play vinyl cd for pleasure, but I cited two groups of
> professionals that *have* to use Java for their daily job. I suppose there
> are many people in the U.S. working in the judiciary system, right? Hell,
> we've got plenty of US-made TV series where the main character is a lawyer.
> :-) Do you think they comfortably fall in the "nobody" category?

Statistically, when referring to computers that are not managed by an
IT team (that is, end users), I certainly consider them nobody.

[Fabrizio Giudici]

On Fri, 19 Oct 2012 22:17:35 +0200, Robert Casto <casto....@gmail.com>
wrote:

> Have you seen the systems used by auto dealers? All they are is a pretty
> face on a green screen scraper. Those systems haven't changed in years
> and
> it doesn't look like they might ever. They are quite complicated and they
> seen no benefit to changing the technology used to develop the tool or
> make
> changes.

I know there are green screen scrapers in many places. So, what? I'm not
discussing *why* desktop people are or aren't using Java. Neither I'm
arguing that end-users *should* use Java on the desktop (it could be a
separate thread). I'm discussing *whether* there is virtually nobody or a
relatively consistent group of people.

[Robert Casto]

You turned it into a philosophical discussion which is kind of boring now.

I was adding other uses of Java to your original post. Tried to help but I see your more interested in semantics. 

[Josh Berry]

On Fri, Oct 19, 2012 at 4:22 PM, Fabrizio Giudici
<Fabrizio...@tidalwave.it> wrote:
> I know there are green screen scrapers in many places. So, what? I'm not
> discussing *why* desktop people are or aren't using Java. Neither I'm
> arguing that end-users *should* use Java on the desktop (it could be a
> separate thread). I'm discussing *whether* there is virtually nobody or a
> relatively consistent group of people.

I guess that is why I asked what are you arguing? I agree there is a
relatively consistent group of people that use Java on the desktop.
Especially consistent in that this number does not seem to be growing
significantly. I further believe that the majority of these people
use machines that are managed by a configuration team. None of this
makes me feel even slightly squeamish when I claim that nobody uses
Java on the desktop. :)

[Fabrizio Giudici]

On Fri, 19 Oct 2012 22:30:26 +0200, Robert Casto <casto....@gmail.com>
wrote:

> You turned it into a philosophical discussion which is kind of boring
> now.
>
> I was adding other uses of Java to your original post. Tried to help but
> I
> see your more interested in semantics.

Well, my first post of the thread was pretty dead on "nobody" vs
"somebody". Yes, it's about semantics.

I'm interested in other uses of Java too, of course. You don't have to
convince me that there are, and I think that the assertion is easy to
demonstrate. Ricky added more examples, and I'm working with desktop
industrial applications for years with big corporates. It's the end-user
segment that, I think, need to be clarified.

So, I figure that the US judiciary system has some requirements similar to
the one of auto-dealers you mentioned, or perhaps there's a technical
problem with HTML and signed uploads (it looks as it's the same strict
requirement for italian business consultants). Architecturally speaking,
and considering that upgrade problems aren't rare circumstances, I'd
probably prefer to see a webapp for that kind of software if I was in
charge of it. So, it might even be *bad* to have Java there, but it's
there. The presence of Java on the desktop is mostly underestimated, this
is my point.

[Kirk Pepperdine]

I use java on the desktop and ask anyone.. I'm completely unmanageable.

-- Kirk

> --
> You received this message because you are subscribed to the Google Groups "Java Posse" group.
> To post to this group, send email to java...@googlegroups.com.

> To unsubscribe from this group, send email to javaposse+...@googlegroups.com.

[Ryan Schipper]

I donated blood yesterday. The nurse was clearly using a Swing application to record patient data.

[Cédric Beust ♔]

On Fri, Oct 19, 2012 at 12:09 PM, Fabrizio Giudici <Fabrizio...@tidalwave.it> wrote:

Taken as a general statement, it is false: the industry uses lots of Java on the desktop.

Taken literally, any sentence containing the words "nobody" or "everybody" is most likely incorrect.

Practically speaking, the veracity of the statement is probably more tied to how many people would agree with it than its literal meaning. Myself, I would certainly agree with the claim that "nobody uses VHS tapes these days" even if it's technically incorrect.

-- 

Cédric

[Fabrizio Giudici]

On Sat, 20 Oct 2012 09:05:53 +0200, Cédric Beust ♔ <ced...@beust.com>
wrote:

Truth is not a matter of democracy, but of facts. There are also false
myths around which people believes in. For instance, the majority of
people use JUnit thinking that it's the best Java test tool around ;-)

It's clear that in our context "nobody" doesn't mean "0%", but "a
negligible amount".

[Casper Bang]

I get the impression your agenda is to uplift the state of desktop Java by finding niche or lock-in counter-examples. While finding evidence to document this conclusion is undoubtedly harder than when I argued Java was dead in the browser 4 years ago [http://blog.bangbits.com/2008/08/myth-java-widely-used-on-web.html], there must be a more scientific way of doing it. 

According to statowl.com, 66% of computers have a JVM installed. Let's say that 50% of those are actually being used, not just dormant/idle installs there due to OS bundling or shoehorning. Last but not least, let's assume that 10% are there to truly power a desktop application, not just to support mandatory national SSO solutions etc. That lands us down around 3%. If you lax "nobody" to mean "few people", it doesn't seem so far stretched to me - and I even think 3% is a bit on the high side (when I look around friends and family).

[Fabrizio Giudici]

On Sat, 20 Oct 2012 11:44:27 +0200, Casper Bang <caspe...@gmail.com>
wrote:

> According to statowl.com, 66% of computers have a JVM installed. Let's
> say
> that 50% of those are actually being used, not just dormant/idle installs
> there due to OS bundling or shoehorning. Last but not least, let's assume
> that 10% are there to truly power a desktop application, not just to
> support mandatory national SSO solutions etc. That lands us down around
> 3%.

The "scientific approach" would be your *assumption* of 10%, right? :-)
Why not 15%? Or 5%? Or 20%? Which is the scientific calculus that leads to
this value? 3% might reasonably qualify as nobody, but if the number is
10% or more, it wouldn't.

My agenda is not to uplift Java on the desktop, which, outside of the
industry, I agree is a minority usage (*); the relevant technology for the
masses is HTML. I'd just like to have people to back their assertions with
some solid argument, but it seems that there's no alternative to vagueness
and personal assumptions.

On the other hand, I can't make numbers too. Of course, lawyers and other
professionals could be less than 3% and your number can't be proved false
by them. Still, if you completely drop Java from the desktop it seems that
we have a few whole segments of people (outside the industry) that can't
do a relevant part of their jobs.

[Ricky Clarkson]

I missed some users I know of.. Ricoh's InfoPrint Manager, just one of whose clients is responsible for >5% of all mail in the US, has a Swing UI for administration.  There's no great advantage to it being in Swing for the user, and there is a web version but it lacks features.

[Casper Bang]

The "scientific approach" would be your *assumption* of 10%, right? :-)  

I agree, that was just inviting for trouble. :) I should have said, let's try to build an estimation model. Other than going into a 1000 random peoples house, and inspecting the installed applications, we won't ever know. And going into a 1000 random Danes' computer will more than likely result in a higher percentage than other countries, since the Danish national SSO solution is based on a Java applet (many relies on smartphone/tablet workarounds instead, since by definition no JVM is not found here anyway).

[Vince O'Sullivan]

On Friday, 19 October 2012 21:03:38 UTC+1, Ricky Clarkson wrote:

I can still knock out a usable Swing application in less than half the
time and fuss it takes a typical web developer, and with some effort
it can even look good.

As a recent convert to HTML5/CSS3/JavaScript, I'd say you're being conservative.  I could probably knock out a Swing app in well under a quarter of the time it would take to do a web app equivalent (and even after ten/fifteen years away from it I could do the same thing in VB6 in a quarter of the time again).

As far as I'm concerned as a software developer, there has been no meaningful progress in language design in the last twenty years (just different implementations of the same concepts that were already around then).

[Ricky Clarkson]

That's not at all language design.

--

You received this message because you are subscribed to the Google Groups "Java Posse" group.

To view this discussion on the web visit https://groups.google.com/d/msg/javaposse/-/1ADYicFAUvoJ.

[a.efremov]

Hello,

I see a lot of java UI on desktop.

1.) Eclipse IDE. Many java/g++ developers use it.

2.) Eclipse RCP. That's really huge. A lot of traders use SWT based app as Frontend. Delta 1, BNP.

alexander