Skip to main content

Blame Sun for that?

Posted by fabriziogiudici on June 19, 2008 at 1:46 PM PDT

Warning, this is an entirely untechnical entry. Read only if you'd like to waste five minutes.

As we all know, corporates such as Sun, Microsoft or Apple are producers - we are consumers. They offer us products and we decide what we like and we don't. Customers' feedback is one of the most important mechanisms in the free market for competition and evolution, so I'm always happy when I hear people complaining about things that dont' work, or anyway they don't like, because this enhances the chances to have better and better products (I mean, I'm known to be a continuous complainer and I'm proud of it).

Nevertheless, it seems that there is an irrational component in customers' attitude to criticize. I mean, normalizing the quality of the product, there are corporates that attract more criticism, others are more immune (I'm not considering Microsoft in this picture, since it's - rightly or not - a specific case related to the figure of Bill Gates).

Let's start considering Sun. We all know the big mistakes Sun did in inventing RIAs with Applets more than ten years ago, and now finding itself in the position of chasing competitors. But in the end, in the last couple of years they have changed direction, made a lot of announcements and released new software, working or in preview, addressing the desktop problems. There is a strong commitment and there are proofs of it. J6U10 and JavaFX will be there in a matter of months and all of us will be able to understand whether they are for real or not.

Now, a couple of days ago I read a blog post with this content: hey, J6U10 can't read the proxy settings from the browser (just after installing, it pops up an annoying dialog) hence it's totally crap and it's better I start studying Flex. Now, let me say: it's a bug, it would be idiotic if it makes its way up to the final release, but we are just looking at an early access. If it was bug-free, it woldn't be an early access, right? If all the J6U10 and JavaFX stuff would prove to be a failure next Fall, I'll probably be one of the guys starting studying Flex. Now, I'm just waiting for the right moment. BTW, J6U10 already demonstrates that most of the difficult technical challenges have been addressed with success (kernel, background loading, etc...) and these are the higher risks for the failure of the project. I don't think reading the proxy setting is difficult, so at the moment the thing doesn't worry me. The correct attitude now would be: shit, that's annoying, I'm filing a bug report and watching it - Sun has better to fix it soon. Nope.

Now, the counter-examples. Well, don't let me talk about Apple. Ok, I'll talk about it. I'm not speaking only about the Java 6 infamous question: I'm pissed off by Leopard too, the most unstable Mac OS X release ever, at the same time packed with the most useless bag of new features (and before starting a flame, let's look at the Snow Leopard announce: for the first time in their history they are focusing on stability with no evident new features, this means that Leopard stability is a problem for real). But only a few people criticize Leopard (the others being too busy in saying "ahhh" at Jobs' keynotes); even most of those who criticize would never give back their Apple gear, I mean, they are not menacing to punish Apple by switching to competitors.

Ok, that's for Apple. Want another? Google: they are going to revolutionize the mobile world with Android, but so far nobody has seen a real Android phone. Don't get me wrong: I don't think it's vapourware, sooner or later it will come up and will play a major role in the scenario. But so far it's more or less in the same position as JavaFX: announced since a long time and not ready yet, but it's getting one fraction of the skepticism for this.

Want another? Mozilla. They have just released Firefox 3. One of the announced features was the better integration with the operating system. In fact, Firefox 3 on Mac OS X looks pretty cool, even cooler than Safari. I'd really like to switch to Firefox for a number of reasons, one being the fact that many sites - including my bank - don't work with Safari and I'd like to use a single browser for everything. So I tried it yesterday at lunch break. I was at a customer's office, behind a firewall. Well, Firefox didn't get the proxy settings from the system, I had to set it manually. Does this recall something from the beginning of this post? With the difference that Firefox 3 is a final release, unlike J6U10. But let's go on. Twenty minutes ago I was playing with Firefox at home and set as the default browser. I went to the "Import" menu and saw the option for importing the bookmarks, the passwords and other stuff. Pretty cool - I did it. Bookmarks are here; passwords, no way. And this is really really annoying. Furthermore - but I'm not sure - I don't think it's using the Keychain Access for storing them, but a private repository (this is NOT integrating with the o.s!). And - you bet? - I'm not seeing harsh criticism on that. On the contrary, everything is talking about the Guinness thing about the massive downloads in a single day, pretty cool, but completely useless and irrelevant in the end user perspective.

Blame Sun for that?

Comments

Right on Fabrizio.
** BEGIN RANT

I work both in the .Net and Java community and I've got to say, the Java community is the whiniest group of individuals I've ever encountered.

  • We don't have closures wha wha!
  • JavaFX isn't as cool as Flex wha wha!
  • My Swing bug hasn't been fixed wha wha!
  • They didn't use Groovy instead of JavaFX wha wha!
  • Generics suck wha wha!

    It's nitpicking at it's worst, all tools and frameworks have long standing problems that don't always get solved, I've worked in C, C++, .Net, Delphi / Pascal, Python, VB, PowerBuilder, FoxPro & Java over the last 24 years and they all have their warts.


    What Java SE gives us, is a performant, cross platform, reliable platform for building applications of any type, as well as a hugely rich ecosystem of open source, liberally licensed libraries & tools to build those applications.

    Could it be better? Yes, things can always be improved, Java has some real problems, but it is now open source, so as programmers architects & hobbyists, we can now do something about it by contributing to the OpenJDK. If there is a Swing bug that has been open for 10 years, then put your money where your mouth is and fix it!

    You have choices in the Java world, if you don't like FX, use Groovy, Flex, Ajax. If you don't like Swing then use SWT or look at something like Pivot (gui toolkit with some real potential), roll your own. If you need closures, use Groovy (not so slow any more). ** END RANT

  • What exactly was release by Sun to make web start or applets deployment better? Only Press release. 6.10 is not released, does not work fully w/ firefox2, win64 nor apple to begin w/. Once those things are fix, released (and other related bugs) and deployed by large # of users can Sun be considered. Developers need end users. No end users, no developers. Look at deployed end user here: http://www.onflex.org on right side. What are downloads for 6.10 jre (w/ fixed bugs, no one here will debug it for Sun). Sun talking about Silverlight or Flex is like a little league baseball entering a world series majors. Yes, it's baseball, but different.

    We're currently developing an enterprise application using JavaFX. We are very pleased with the level of commitment and support we have received from Sun - from the executives and the teams involved in JavaFX. We believe it is going to be a game changing technology.

    2 things Sun should do for Swing (THE Java UI component set):
    . clean the Swing mess. Don't be afraid to break code. Incorporate SwingX renderer architecture, hide to AWT plumbing and base everything on Painters.
    . give more visibility to a small selection of open source projects like JXLayer and the excellent Swing Explorer.

    OS X Leopard is a very stable and also fast operating system, but in my experience it must be installed on a fresh hard disk and you should NEVER update an older OS X version to Tiger. Format the hard disk and install Leopard from scratch. (Yes, this is now -exactly- as we know it from Windows.) You also need to have at least 2 GB of RAM in your machine, and the CPU should be no slower than a 1.67GHz G4. Is it worth it? Yes. Quick Look is one of the most useful features in the system, and the overhauled Finder and its now finally really useful and working browsing through the network is a huge step ahead. (Finder should be multi-threaded and faster, though.) I don't know what Apple did with their Java 6 implementation. Nothing seems to run on it, everything crashes or quietly disappears into Nirvana. But in all honesty, it's not even really important. Most of the world still uses Java 1.4 or even older versions, so it's hardly a show stopper. And whoever loves to bitch about Apple and their (lack of) support for Java: It is still the ONLY desktop operating system on the world that comes with a pre-installed and well integrated Java SE environment. And while Microsoft decided to go in for the kill and deploy .NET instead, Apple still develops, deploys and supports their own implementation of Java and makes it a default part of their OS. It is the only platform where you don't have deployment nightmares or headaches, and where you do NOT have to tell your users to download and install a Java VM. Talking about leaving Apple: Yes, there are enough reasons to do that. Whenever I connect my iPod Touch to my Mac and experience all the restrictions that iTunes puts on my device and my content, I am nothing but disgusted and regret the day that I paid the money for that iPod. But frankly speaking: Where do you want to go? Back to Microsoft? I remember that there were millions of reasons why I left their platform. Switching to GNU/Linux? Now that's a great joke. Almost none of the software that I use and need runs on that geek system, and there are NO acceptable or usable alternatives available. So it's either Microsoft or Apple. And all my complaints aside, I still know which system I like better. And it is not from the huge company in Redmond, Washington, USA.

    Why does this website not keep the paragraphs where I hit ENTER? Now THAT'S annoying and should be fixed.

    Well, I run Leopard on a MacBook Pro Core Duo 2GHz, I reinstalled it from scratch twice, the first time when I bought it and the last time more or less one year ago when an official update from Apple disrupted part of the file system. I'm not the only one complaining about the instability of Leopard, there is people also in Apple users' forums etc.

    For what concerns Java 6, it has been already debated do death, in any case desktop people need Java 6. _I_ need Java 6 for a lot of stuff. All the new Sun stuff for the desktop is Java 6. Mac OS X is the only operating system in the world which doesn't support Java 6 for 32 bit machines (and older PPC, which is obsolete, but still most of the installed Apple hardware). And you don't need to deploy your applications with a bundled Java also on Ubuntu or other Debian-based Linuxes since more than one year (you can easily build the distributed package to install the precisely needed version of Java automatically).

    You're right that in the end Mac OS X is still the best operating system on the whole, in fact in spite of my constant evaluation of alternatives (Windows excluded) I'm still there (BTW, if I used my laptop only for working, Linux would be pretty ok. But it lacks all the stuff I'm using for personal needs). But one point is to sadly conclude that you can't leave after thinking on it, another point is to happily stay with Apple even if it hurts you, like most people do, even those who complain.

    And yes, this site technically sucks and for what concerns me this is one of the biggest things I'm blaming Sun for.

    Well, Fabrizio as the author of the original blog post you mentioned, let me reply...in general I think most Java programmers are frustrated (putting it mildly) with both Sun's strategy (JavaFX instead of Java) and implementation (how long have we been working for Applets 2.0 already?) Look at this from my perspective: I download the beta (not some alpha nightly version, but a beta that should be fairly stable I presume) and try it. On the first site it fails right away in firefox, because no one bothered testing with a proxy server. On IE, it was even worse...i got some security notice about ActiveX, when I said I want to install it, it gave me no errors, but the applets did not display at all. So, in 5 minutes of testing the new plugin fails on both browsers and is subpar (putting it lightly) to what Flex and Silverlight offer today. I won't even go into the issue of tooling...if you compare Microsoft Expression Blend with the Notepad++ type JavaFX editor for NetBeans (or the total lack of any tools for Eclipse), it's just sad how far Sun has let Java fall behind compared to Flex and Silverlight. And for crying out loud....why JavaFX at all? Why couldn't pure Java be enhanced to allow for declarative UIs or integrated FX? It's a tall order to entice people to move from a language they are comfortable to a brand new one, especially one that has zero user base and community behind it (couldn't Sun have at least looked at Groovy or JRuby first?...at least someone is using those). Let's face it....JavaFX was a part time project from a single Sun employee that suddenly got yanked into the spotlight and made the cornerstone of Sun's RIA strategy. I guess it was too hard to step back and think of how Java itself could be enhanced so that we don't need a brand new language? In short, when it comes to RIA, Sun is failing both in terms of strategy and delivery....and it's already very late to the game anyway.

    +1 on the horrible logic of this website...why does it not respect paragraphs and just mixes all the text together?

    "You don't think it is important, but all that counts is that it was apparently important for jacek."

    That's not true, never said "it's not important". It will be a showstopper when J6U10 is final, and I agree with that; but J6U10 is not yet for production use, and that bug is not blocking any kind of experimentation. What I've said is that I don't see it so technical challenging to fear that Sun won't be able to fix it.

    The correct attitude now would be: ... I'm filing a bug report ...Good luck with that one. Those longer in the business know what filing a Java bug is good for. I must say I don't know any Java developer who any longer cares about the bug parade. All that counts is what one has in hands. ... and watching it Yes, watching paint dry.

    Let me give you a story. Many, many moons ago I took part in a Sun event regarding the introduction of Solaris 2.0. Solaris 2.0 was just freshly released. For those to young to remember, Solaris 2.0 was Sun's replacement for their rock-solid SunOS 4.1.mumble Unix. Solaris 2.0 was shocking. It was the first Unix I ever saw and used that crashed within two hours of usage. Not good, when a Unix in these days was supposed to keep up for month or years.

    OK, back to the event as I remember it. It was a bunch of presentations how great Solaris 2.0 was. And there was a Q&A session at the end. That session was run by a Sun employee who claimed to have the IMHO rather childish job title "Rabble Rouser" on his business card. Mind you, this was before the new economy with all that nonsense job titles. Maybe the title was a predecessor of the equally stupid Sun "evangelist" titles, I don't know. He was giving his spiel of how much Sun cared and how his job was to get things going within Sun.

    So, in the Q&A session someone, an early adapter, stood up and asked about a networking bug in Solaris 2.0. "Rabble Rouser" tried to play down the bug. However, the guy insisted of this being a real problem. The response was something I never forget, because "Rabble Rouser"'s response until today still pretty much describes Sun's IMHO wrong attitude to support and maintenance of code:

    Do you think your bugs are more important than everyone else's?

    To bad he told that to the admin of one of the largest universities in the admin's country who was about to role out the largest Solaris 2.0 installation in that country. But he got the only correct answer:

    Yes!

    I am telling this story because I think Sun's attitude when it comes to bugfixing Java is like this, and is still wrong after all those years. On a subjective(!) scale "my" bugs are of course more important than anyone else's bugs. Because my bugs prevent me from doing what I want to do.

    Sun, however, has tens of thousands of such bugs in the Java bug parade, because apparently Sun thinks they are not important. But they are.

    They are important for the one who took time to report them, but then just having to watch Sun to let the bugs sit there for years and then close them without action. I have the impression Sun's maintenance crew is severely understaffed (instead resources are put into the development of yet more hype features). And I have the impression that the selection which bugs to fix is based on a "oh, today I feel like taking care of this, this doesn't look to hard" attitude, instead of systematically plowing through the bugs.

    Developer's motivation to report a bug is severely limited by the impression Sun gives "your bug is not important". Developer's time is much better spend trying to find a workaround instead of reporting a bug and watching paint dry.

    And I see the same attitude regarding the JavaFX proxy problem. You don't think it is important, but all that counts is that it was apparently important for jacek. And by telling him his opinion is wrong you are walking in the footsteps of "Rabble Rouser", who couldn't wrap his mind around the fact that people have reason when they complain.

    I'd like to add an other aspect. Year ago I learned something important in user interface design. The user interface is the business card of an application, and the installer is when you first hand over that business card. Would you hand over a dirty, crumbled business card and then still expect to be taken serious by the receiver? No. So really, why does Sun do it and gives out software that acts up at startup?

    Hi Jacek. Sorry for not linking your post, but I wasn't able to retrieve the link. Back to the point, I _personally_ don't see any big advantage from JavaFX itself. I'm fine with the Java language. But JavaFX is bringing also Java new features, as part of the core of libraries (SceneGraph, JAva MEdia Components, etc..). For a marketing point of view, I understand completely the choice of introducing a new language, since we must confrontate with specific languages on the other side (e.g. ActionScript). About the JavaFX language itself, it could have been better or worse, I don't know. I've started some playing with it, but the goodness of a language can be understood only at the proof of a real context. The binding concept in JavaFX, in any case, is extremely functional and powerful, and I've seen it's fascinating quite some people who's currently in the Flax/Flesh world - this is a point.

    @jancarel: what are your doubts about Sun's policies? Yes, what we have now is mostly unfinished - But they have clearly stated some deadlines for this Summer / Fall. This is not a deadline of one or two years, it's a matter of months. Soon it will be easy to understand whether they are up to their promises or not.

    I am grateful to Sun for providing my small company and myself with free software from which we have profited without paying SUN a cent.
    I have a big wish list and criticism, but what bugs me most is that I cannot figure out what SUN's current main policies are in regard with Java development.
    It looks to me that lots of hypes are chased and very little is finished.

    That's a very honest post. Having been a Java developer for almost a decade, I can sense that Sun is changing to one that is more open and receptive perhaps due to the CEO. We should give Sun a thumb up for continue to innovate in Java amidst the rise of many competing technologies and languages. Java has come a long way and still growing and that itself is testimony of how Sun treat Java. I am glad that Sun opens many interesting technologies for early preview and the developer can get a glimpse of forthcoming stuff despite increasing receiving criticism and negativism from those who make the judgement from early pre-release of the product or from lack of product's update in Java's blog. If JavaOne is the indication, Sun has certainly changed from a company that seeks to underpromise and overdeliver. JWebPane and Java Media Component are excellent examples that Sun is capable of innovating while continue to work behind the scene. We should continue to give Sun the benefit of doubt that it will deliver the goods utlimately because this is what Sun does best.

    "J6U10 and JavaFX will be there in a matter of months and all of us will be able to understand whether they are for real or not."

    They've been saying that for something like a year now...

    "at the same time packed with the most useless bag of new features "

    Just like Java 6.

    "Why does this website not keep the paragraphs where I hit ENTER? Now THAT'S annoying and should be fixed. "

    Nope. The editor allows you to use html tags though (not that those always come across correctly... php...

    "It looks to me that lots of hypes are chased and very little is finished. "

    The development of the Java platform has been driven by marketing and egos rather than technology and business requirements for several years now. And it's set to get worse with Java 7.

    re: twe post on Sun and bugs: http://lopica.sourceforge.net/faq.html That was bugs on webstart before there was Flex. Majority of the webstart bugs are not fixed yet. .V

    The problem with JavaFX is that Java 6 Update N is only available on Windows and Linux, whether or not Apple will release an update remains to be seen.
    • There are no real demos that show how to write real apps using JavaFX. All I've seen is Hello World and some basic examples, nowhere we can find how to build a business application or a rich consumer application with video support.
    • documentation is still sparse.
    Java is going in the right direction with Java 6 Update N, if this would have been done many years ago Flash wouldn't have been where it is now. I also would have preferred to enhance Swing instead of reinventing a new language.

    "Java has been the victim of incredible over-hype in the past."

    Right. But hasn't been Apple too? And Google? And, yes, I agree that often the harsh critics come from "heart-broken lovers", not merely trolls. I find myself often in this attitude (*), but - hell - there's should be a limit. There is a boundary between constructive and destructive criticism.

    (*) Initially I criticized Apple just because I loved them, hoping that they would listen. Now that I've realized they don't and never will, I just hate them, even though I feel I can't leave them (yet).

    For some reason people like to predict nothing but doom and gloom for Java. It may be a reaction to hype. Java has been the victim of incredible over-hype in the past. That can have a backlash if expectations are not managed correctly. It's also possible that the poster wants to "scare Sun" into taking them seriously and so goes for the drama of mentioning a competitor. Apple, Google and Mozilla are (rightly or wrongly) considered the underdog and different rules apply to them becouse of that perception. It's not fair but there it is.

    Right on, especially about Apple. I was at JavaOne when his Steveness got up and proclaimed that OS X was going to be the best Java development OS. Of course, that was when Apple was actually at JavaOne. On the basis of that, and because their computers are pretty cool, I convinced my employer to purchase 2 Dual G5 PowerMacs, and later 2 Mac Pros and a Mac Book Pro. And, to Apple's credit, with Tiger and JDK 5 I think they did it.

    Unfortunately, then Apple became hugely successful and Steve's megalomania returned. Now it's all Not Invented Here and "Nobody wants that Java crap". Who knows if Snow Leopard will ever contain Java 7?

    Oh, and Apple, nice job on selling out WWDC. Congratulations! You are now a third of the size of JavaOne. And what a great company it is that makes you sign an NDA to attend their conference. Truly, Apple and the RIAA are made for each other.