The Artima Developer Community
Sponsored Link

Weblogs Forum
Hybridizing Java

129 replies on 9 pages. Most recent reply: Mar 18, 2011 1:41 PM by David Benson

Welcome Guest
  Sign In

Go back to the topic listing  Back to Topic List Click to reply to this topic  Reply to this Topic Click to search messages in this forum  Search Forum Click for a threaded view of the topic  Threaded View   
Previous Topic   Next Topic
Flat View: This topic has 129 replies on 9 pages [ « | 1 ... 4 5 6 7 8 9 | » ]
Tim Nash

Posts: 11
Nickname: tim99
Registered: Apr, 2005

Re: Credibility Posted: Feb 5, 2007 3:19 PM
Reply to this message Reply
Advertisement
> _really_ know if today installing a JRE via Java Web Start
> is really harder than installing a Flash/Flex runtime...

In many organizations it is.

But the browser as a platform is finally beginning to happen, thanks to Flex, XUL and WPF/E. And it has only been, oh about 12 years since the netscape IPO!

Check this out.

http://starkravingfinkle.org/blog/2006/12/xule-what-if/

Dean Schulze

Posts: 3
Nickname: dwschulze
Registered: Feb, 2007

Re: Hybridizing Java Posted: Feb 5, 2007 8:52 PM
Reply to this message Reply
Thanks for the URL. Cairngorm looks good (except for the name). I'm glad to see that Adobe consulting is taking patterns seriously.

Bruce, maybe you should focus on teaching Cairngorm instead of Flex.

Dean Schulze

Posts: 3
Nickname: dwschulze
Registered: Feb, 2007

Re: Swing on top of Flash Posted: Feb 5, 2007 8:57 PM
Reply to this message Reply
That's great, if the API is at least as well designed as Swing. That's not setting the bar too high, so I hope they didn't do worse than the designers of Swing did :-).

It will be good to see what aswing.org has to offer when their web site comes back in a few days.

Fabrizio Giudici

Posts: 14
Nickname: fgiudici
Registered: Jan, 2007

Re: Swing on top of Flash Posted: Feb 6, 2007 1:00 AM
Reply to this message Reply
> I was just illustrating Sun's reluctance to change some of their more
> dubious decisions even when there's a lot of community interest in doing so.

I don't want to drift the topic, but while nobody asserts that Sun does always everything well, it's not that just because a group of programmers complain about something we should imply that Sun did it wrong. While there are lot of programmers asking for things to be changed, usually people that like things as they are just don't say anything because they are fine (so it's even difficult to say where the majority is). Maybe you second the first group of programmers and then the latter starts complaining. I mean that the mere existence of people complaining is not enough to draw the conclusion that something has been done bad (in other words, if you are in a position like Sun that's offering a technology to millions persons, it's impossible to please everybody).

> Just about everything I've read about multi-threaded programming says
> that it's often useful, but that you should avoid it if it really isn't necessary.
> The design of Swing forces you into multi-threading even when there's
> really no good reason for it. Has this been changed?

Multi-threading is required when you have to perform a "long" task in reaction of a user gesture. No, it hasn't changed, but Sun has introduced some helper classes (such as SwingWorker) that greatly simplify the handling of this scenarios. On the other hand, the compatibility of Swing with multithreading, versus the inability of Flex to do real multithreding, is to me a definite architectural advantage of Swing over Flex if I think of a really serious client application with a perspective of some years - as I said, at least every ten days somebody publishes a new article on how we will be forced to use multithreading for dealing with multi-core computers.

Also about multithreading, can somebody shortly explain me how Flex handles the case in which the user presses a button, a remote invocation to the server must be performed and it can take seconds, and the whole GUI doesn't get blocked? Thanks.

Les Stroud

Posts: 7
Nickname: lstroud
Registered: Jul, 2003

Re: Hybridizing Java Posted: Feb 6, 2007 4:54 AM
Reply to this message Reply
I mentioned this earlier, but it didn't really get any play. What is the security story flex/flash? Does it have a verifier? Does it require certs for access to local resources? I am sure that if there were glaring holes that there would be a variety of flash viruses. So, does someone have a link to a whitepaper on it's security arch?

Hugh Grady

Posts: 1
Nickname: baseuser
Registered: Feb, 2007

Re: Hybridizing Java Posted: Feb 6, 2007 5:17 AM
Reply to this message Reply
If your suggestion is taken up can we anticipate a world full of Flash-based web sites? if so, I'm horrified - the ones I visit that are openly based on Flash content are SLOW. They inevitably have these swish (flash) intros that drag on forever unless you happen to find the dinky little 'skip intro' control hidden somewhere and are full of time-consuming images. Admittedly, my broadband isn't as broad as it could be and I understand that bandwidth is just another resource to be exploited but surely there is a better solution available than the universal adoption of a resource-hogging proprietary language? Hopefully you'll tell me that all those web sites are just the product of a deranged subset of Flash developers and the web is not going to become even slower than it is already.

Charlie Arehart

Posts: 3
Nickname: carehart
Registered: Feb, 2007

Re: Hybridizing Java Posted: Feb 6, 2007 7:20 AM
Reply to this message Reply
Yes, Les, security in the Flash Player is quite well thought out and as widely distributed as it is, surely any holes would have been exploited by now.

But yes, there are security whitepapers. The best place to find such things is the Flash Platform page at Adobe (http://www.adobe.com/platform/). There you will find two different security whitepapers, the first a summary and the second in detail:

http://www.adobe.com/platform/whitepapers/flashplatform_security_enterprise.pdf

http://www.adobe.com/devnet/flashplayer/articles/flash_player_8_security.pdf

An older version of the second one (2002, on Flash Player 7) is at:

http://www.adobe.com/devnet/flash/whitepapers/security.pdf

I'll add as well, as a help to others exploring such things: be careful while searching and reading to distinguish between the Flash Player (what runs in the browser) from the Flash IDE (what the developer uses), and Flex, which is an alternative to that.

For instance, if you use the "products" link at Adobe you're taken to http://www.adobe.com/products/flash/flashpro/, which is about the Flash IDE, where the focus is more on features of the tool, and not much on security, which could leave one wondering about its importance. The point is that its integral and doesn't require restriction but rather permission.

Charlie Arehart

Posts: 3
Nickname: carehart
Registered: Feb, 2007

Re: Hybridizing Java Posted: Feb 6, 2007 7:31 AM
Reply to this message Reply
Hugh, your concern about "bad flash apps from the past" is a common one that was addressed earlier in the comments. Most of those old "skip intro" style designs represent early bad design choices, much as some who get a word processor go crazy with fonts, or with presentation software go crazy with transitions.

Time and experience (and user backlash) has pushed a lot of that kind of Flash interface out of existence, though certainly it still exists and could still be created by one who "doesn't get it" why they're a problem.

More important, there are many ways in which Flash is being used in small segments of a web page--and people may not even notice as its so seamless.

Further, I think some would argue that Flex is itself going to lead to a difference in the likelihood of such interfaces as those bad ones of old, simply because of its approach to design. Still, it could happen, just as a Java developer could build without OO if they really missed the boat in their learning. That's where having someone like Bruce involved in Flex training will be such a blessing.

Finally, it's worth noting that as far as speed concerns go, a Flash app itself is usually very small, so should load fast. Still, that app could go back to the server for data (either on initial load or as the result of some event on the client), and so that could cause delay.

An important difference compared to a web page load, though, is that it should only be causing a load of the DATA it needs, not the entirety of the HTML (and possibly images, etc.) of a typical page refresh.

Still one more point should be made, as some may see a "Flex" app and it could load slowly. Prior to Flex 2, a Flex app was compiled on initial request on the server and then sent to the browser, which could also cause a lag.

In Flex 2, there is no longer a server compilation component or process. Instead, as Bruce explained in his article, Flex apps are compiled by the developer with the free SDK (or in the Flex Builder tool). Only the resulting SWF are placed on the server and downloaded upon request.

Don Babcock

Posts: 3
Nickname: dpbabcock
Registered: Feb, 2007

Re: Hybridizing Java Posted: Feb 6, 2007 8:07 AM
Reply to this message Reply
Bruce nailed it. I've been in this business for over 30 years and he's singing my song.

Ivan Lazarte

Posts: 91
Nickname: ilazarte
Registered: Mar, 2006

Re: Hybridizing Java Posted: Feb 6, 2007 9:23 AM
Reply to this message Reply
"If Java's weakness is its lack of multimedia support and Flash's strength is that it has solved the installation problem then what we really need is a third party library for Flash multimedia support in Java. Then convert the Java bytecodes to a Flash binaries. You get to leverage Java programming knowledge, stricter type dhecking and a language better suited to OO principles rather than a scripting language that encourages hacking. "

Dean that is a great idea. I hate relearning Actionscript every 2 years or so just to figure out what is still broken or missing.

Jeff Ratcliff

Posts: 242
Nickname: jr1
Registered: Feb, 2006

Re: Swing on top of Flash Posted: Feb 6, 2007 12:58 PM
Reply to this message Reply
> I don't want to drift the topic, but while nobody asserts
> that Sun does always everything well, it's not that just
> because a group of programmers complain about something we
> should imply that Sun did it wrong.

Of course, but likewise, just because Sun implemented something a certain way and some programmers agree, doesn't make it correct either. Obviously if the assumption is that there's nothing that needs "fixing" in Java, than this whole "do-over" issue is moot.


> Multi-threading is required when you have to perform a
> "long" task in reaction of a user gesture.

Yes, I understand the idea, but multi-threading is never required to handle such a scenario. It is often the best choice, however.

Chen Shiang Khoo

Posts: 5
Nickname: coolboykl
Registered: Feb, 2007

Flex with JSF Implementation Posted: Feb 6, 2007 11:58 PM
Reply to this message Reply
Hi

To benefits Java Web application developers on migrating to Flex, I think it's best if Adobe could provide us a Flex's JavaServer Faces (JSF) implementation (Similar to Apache MyFaces), which will render Flex application base on standard JSF Tags. Doing this, I don't have to reinvest my time on learning MXML, daya sync, and ActionScript. And our JSF configuration (i18n, page nagivation, security, transaction, error handling) stay.

Just my two cents

Cheers

Achilleas Margaritis

Posts: 674
Nickname: achilleas
Registered: Feb, 2005

Another solution would be... Posted: Feb 7, 2007 5:21 AM
Reply to this message Reply
...to make Java classes lazily downloadable and internet-aware.

By internet-aware, I mean that the programmer could import classes from URLs, like this:

import "www.sun.com/java"::com.java.swing.button;


By lazily downloadable, I mean that a class would be downloaded to the local store only when needed, i.e. if it does not locally exist or a new version is available.

What would be the benefit of this approach? well, there would be several benefits:

1) no installation of JRE
2) automatic installation of classes
3) automatic updating of classes
4) easy distribution

It would certainly make applets much easier: the only thing to initially download would be the applet itself, and then the rest of the classes on a per-need basis.

Ivan Lazarte

Posts: 91
Nickname: ilazarte
Registered: Mar, 2006

Re: Another solution would be... Posted: Feb 9, 2007 2:24 PM
Reply to this message Reply
version number would have to be included there, but i like the general idea. the only issue is if people require different versions resulting in redownloading of common components.

James Madison

Posts: 2
Nickname: madison
Registered: Feb, 2007

Legal Issues with Flex? Posted: Feb 9, 2007 9:56 PM
Reply to this message Reply
Putting aside the technical, for the moment, it would seem prudent to look at the legal framework of Flex and ancillary components/toolkits/tools/etc.

The basic question is:

What sort of licensing is involved with making, distributing, and selling Flex applications?

Obviously the only current implementation of Flex seems to be from Adobe. Obviously the tools to build Flex applications seem to be only available from Adobe and probably carry all the terms of Adobe's heavyweight EULAs.

I'm sure most of us are aware that Adobe is the company that threatened to sue Microsoft if Microsoft put PDF support (another ostensibly 'open' standard of Adobe's) into Office 2007.

With this sort of threat hanging overhead, does it make sense for any ISV to embrace Flex?

Without a clear picture of Adobe's licensing framework for the entire Flex 'stack', it would seem prudent for most ISVs, especially the smaller ones, to avoid using Adobe technology. Not many small ISVs could withstand Adobe's lawyers. And who can tell what application you build using Adobe technology that Adobe will decide to sue you for?

Bruce, since you seem to be Adobe's chosen champion to promote Flex in the developer community, perhaps you can give the developer community a clear picture of the legal situation regarding ISV/enterprise/third party developer use of Flex technologies?

Also, if you could clarify your position with Adobe, it would help potential ISVs understand why you are promoting Flex over all other possible solutions. What sort of compensation are you receiving from Adobe and what has Adobe tasked you to do for this compensation?

This latter matter is important as your technical arguments are mostly hyperbole and 'case building' against Flex competitors. This is not to say Flex itself does not have its strengths. But for many companies, even substantially more work is worth the avoidance of using a single proprietary application development/delivery stack over which they have little input and little control. Not to mention the ever-present lawsuit threat.

Lastly, I am asking about Adobe's licensing, specifically. I am not trying to make any sort of statement whatsoever about Adobe's licensing vs. Sun's licensing vs. Company X. Sun has horrible licensing and it took them many years to figure out that if they didn't open source Java that Java would die an untimely death. And Sun's licensing (and approach to building third party ecosystems) is still very immature and counterproductive.

If you can get the appropriate legal/licensing facts from Adobe and put them on the table, that would be most appreciated. If I missed an already written disclaimer regarding your position with Adobe, I apologize.

Flat View: This topic has 129 replies on 9 pages [ « | 4  5  6  7  8  9 | » ]
Topic: Scala 2.8.0 + Lift 2.1 + IntelliJ: First Step(s) Previous Topic   Next Topic Topic: Security Debt


Sponsored Links



Google
  Web Artima.com   

Copyright © 1996-2017 Artima, Inc. All Rights Reserved. - Privacy Policy - Terms of Use - Advertise with Us