Comment Re:Assembly == SLOW ; JAVA == FAST! (Score 1) 372
What you speak of is a direct consequence of Intel holding its monopoly over everything. So while there may be a slight technical challenge, there is a much larger political challenge.
What you speak of is a direct consequence of Intel holding its monopoly over everything. So while there may be a slight technical challenge, there is a much larger political challenge.
Nope wait, wait!! That last reply, the JSR thing not correct. I'll totally admit wrong when wrong and JSR 296 isn't Swing itself but the aborted effort to build the application framework a while back. My bad, got that wrong. All shit tossing about that you want totally justified. That is all.
Well I work with Java all the time and I find Swing great to use, but when I want to build my own user interface elements, then yes, I find it very painful indeed. I've done custom widgets in Cario/GTK,
In that respect, GWT is worst. GWT makes a lot of assumptions about how you want to convert Java into JavaScript. When you want to create your own web elements, not only do you have to build it using some of the worst API I've seen, you actually have to fight the assumptions that GWT tries to make right out of the box. You are literally fighting on two fronts to just make a damn clock widget or rich editor widget. It is ridiculous.
GWT on a whole is a pretty sad state. It integrates poorly with JEE frameworks and when it does, it's literally shoe-horned. You'll find that something like JSR 356 is a waaaaaay better approach to bridging the gap between server side code and client code, and I'd hope that the Java community can keep hitting excellent projects like this. Because Google isn't doing it with their idea of interoperability.
Also since you want to be mincing words between library and framework. If you happen to head over to JSR 296 you'll see that the JSR refers to it as "The Swing Application Framework". Maybe you should head over there and educated them that they should have called it "The Swing Application Library". That being said, I believe your retort to have loss any of its credibility but seriously, this is Slashdot, does anyone think anything on this site is credible? Hope you are still laughing though.
What? Do you mean 2.6.0 RC1? 2.6 looks to be more of a clean up of the 2.5.1 stuff rather than anything new. If anything the main thing that 2.6 brings is that they brought Java 7 into the picture. I wouldn't say that Google *has* abandon GWT, but they sure are making the common gestures of getting ready for a good old fashion keelhauling.
Now for just my opinion, GWT sucks. It's a messy looking API and lacks a ton of flexibility. For example, trying to implement custom UI for your web page is painful and totally unpleasant. More so than say making the same customer UI in Java Swing (which is pretty painful in of itself). In my opinion, and you my mod me down for it, is that anything that is worst to do in (insert framework here) than it is in Java should not exist.
I think the parent has a good argument, maybe just no stated in the best of terms.
However, on the Dart site it says that Dartium, the DartVM enabled version of Chrome, will be one of the major focuses of the Dart team. Somehow, I have a sinking feeling that maybe, just maybe, Dart and NaCL are going to become *major* line items for ChromeOS and Chromebooks. Much like how ActiveX and VBScript became pretty important pillars in Microsoft's platform.
So while on the face of it, it sounds like a shrill. It actually can be rather thought provoking about the future of Chrome and Google. Just for a second think about where VBScript and ActiveX went during their lifetime and what they eventually evolved to. Granted we all now look back and see VBScript as the useless thing that it is, but in it's day, it provided a very powerful way of making offline enabled web pages and was featured heavily in WSH for admins until replaced by PowerShell. Clearly, Google hinting at Dart in the server is an indicator that Dart very may well have a life not unlike VBScript.
I think your arguments are pretty valid and I am by no stretch of the imagination a pro-Dart guy, but I believe that the "cross browser" claim comes from the olden days of cross platform languages. C/C++ had (has, just in case the past tense is a really bad choice) cross platform compilers they take C/C++ code and compile it to a language that the target platform understands. For example, C to ARM/x86/amd64/MIPS... compilers.
So my guess here, and it is just a guess, is that Google is using the same rationale to justify calling this a "cross browser" language, because the compiler can turn Dart into a language that can be understood by other browsers, much like a C compiler can compile into different paltforms. Arguments about if that is an accurate equation are totally justified and most likely will ensue hereafter. I'm just tossing up a guess as to why Google felt like that was an accurate statement.
That is correct. When writing my piece for the story I wanted to ensure that I conveyed the correct idea that the Dart2JS compiler now generates up to 40% less JavaScript than previous versions of Dart2JS, not that the JavaScript itself that it generates is 40% faster.
As far as speed goes, there is all kinds of data on that facet on the Dart website. That will no less generate large amounts of debate about the figures presented by Google about Dart's performance. Just for the comedy of it, dart2js may generate less JavaScript, but will generate at least 60% more debate about its use.
If I was not clear on the whole point by this part of the story:
The new release brings a much tighter dart2js compiler reducing overall JavaScript output up to 40%
Then it was an honest mistake. I work with computers not the English language.
They put it there for the casual on-lookers. For whatever people bang on it, when you say Linux in an interview and they ask you which distro, you say Slackware. If you know how to hold it together with Slackware, things like installing 3rd party drivers from the command line on Ubuntu or SuSE or knowing the entire purpose of everything in
Here's to the release of yet another amazing version of the best Linux distro to date.
Yes blame customers. Businesses are always going to favor vendor lock in, if that comes as a surprise or vice to anyone, then they ought not be in the biz. Customer's as you say are, "dumber than dirt." They know not what markets they are creating, nor what repercussions those choices will have. Look at the uneven level of apps vs. HTML5 on any mobile device. Heck even Facebook gave HTML5 a fair shake and they just couldn't satisfy the customers, with it.
I totally understand what you are getting at, admins suck for not moving forward enough, but at the same time, they can only move as fast as the customer because at the end of the day, they still have to pay rent. The requirements for MSIE 6/7, flash, or Java plug-in are ridiculous. But customers want camera, VoIP, desktop sharing, ability to manage a printer, reset switches, animated charting, 3D effects for presentations, and so on... Customers are impossible to satisfy, which is why companies in the tech industry keep popping up, which is why different ways of doing things keep getting made. There is always something that someone isn't doing and a company's primary goal is to do that thing and make money from it (and at times they honestly just want to make people's lives easier, but, well I don't think I really need to elaborate how often that's the case).
I totally agree with you up to a particular point. Don't get me wrong, I read your post and was nodding yes to a lot of it, but the overwhelming majority of issues are because clients want their software to do more and standards be damned in the process. There are those that will stick to the standards as much as possible, and then there are those who think standards move too slowly to really address in real-time customer demands. The ones who are breaking the rules tend to have that dot-com bubble charm in my book, but it seems to be the way to get investors and gain critical mass. However, they do it because there exists a demand for it from customers. Additionally, I've enjoyed your comments quite a bit. Rarely do I find on Slashdot someone willing to talk without resorting to some childish level, bravo.
And what I'm getting at, is that servers only serve up what the customer wants. If no one gives a flip about standards, and few do, then a company changes the server to meet customer demand. Be it HTML or some crazy binary protocol.
The number of clients out there dictate what servers are dishing up. Seeing how there is still a ton of non-standard rendering browsers out there, they still need to cover those people or just go ahead and pack the bags because they won't have a lot of customers if they can't get the website to properly render on their system.
So until we invent FUCUP (Force Upgrade to Client Universal Protocol), clients are always going to lag behind standards, just because, and there is little an admin can really do about it if they want customers to keep visiting.
That's like saying, "if you don't want to get arrested, don't do anything illegal!" Or am I the only one that got that vibe?
Which is why HTTP Client Hints were invented. User agent strings were a hack back in the day to figuring out abilities on a browser. They still are. Also, no the problem isn't the server side. That's like saying the problem with pollution is the road and not the car. "Hey look I made a car analogy! Do I win a prize!?"
"I've seen it. It's rubbish." -- Marvin the Paranoid Android