Follow Slashdot stories on Twitter

 



Forgot your password?
typodupeerror

Comment Re:Natty uses Wayland? (Score 1) 179

I don't know what windows does, but there is nothing close to "remote desktop" in the linux world. It simply is that good, and I'm a linux guy. They are somehow able to make a hardware accelerated desktop, and also provide a very snappy remote view into it which works very well over a tiny net connection. VNC (and even nxclient) are not in the same class... so that understanding that somehow the current X implementation is needed to provide a remote view of some sort is simply not true. Windows does it, and does it extremely well.

Comment Re:What about internal vs. external sound card? (Score 1) 520

Yes, USB 2 is more than acceptable. I have a motu microbook which even draws it's power from the usb source, and the sound quality is second to none. Using ASIO drivers (windows) there is absolutely no latency (there will certainly be less latency then using an onboard sound card without ASIO drivers).

Comment Re:Finder (Score 1) 312

Can't agree with you there... return means a lot more than just "new line". It also means "complete action"; for example, hitting return in the google search box will perform the search and show the results. There may be an argument for taking other actions on hitting enter, but I myself certainly would never choose something so odd as "rename file".

Comment Re:I doubt this is Oracle's motive (Score 1) 341

. Remember that java is not competing in the kernel / driver space; it's designed for networked / business / enterprise applications. In that space, developers playing with pointers means that I can crash an app by uncovering a bug in the toolbar (happened to me the other day in a kde app; click on a button, entire application goes down with an incredibly helpful "segfault"). Again, I have no idea what you were trying to do in your application that you benchmarked, but trust me, if it's 50 times slower then the c version (like you claim), then you better think about hiring a competent java developer, because you're most certainly doing it wrong.

Stop showing that you don't know what you're talking about. Pointer arithmetic is done by coders in c for all sorts of apps, not just kernel space. You really don't know what you're talking about, and your ignorance shows.

Now you're just putting words in my mouth. Please re-read what I actually wrote.

We interviewed "competent java programers" - including one of the lead devs from netapps - 10 years experience and he couldn't pass the interview because he could only think in terms of java ways to solve problems. Algorithms in c that let you do things quickly just aen't available in java. The language is crap when it comes to high performance.

But that's okay - people like you can't be counted on to manage their own memory to begin with, which is why you have toy languages like java.

So now Java is a toy... wow, not many java haters have the balls to say that, knowing that they're going to end up looking like an ignorant tramp.

Python and Ruby are perfectly valid comparisons; at the very least, it shows that the jvm is not "slow / bloated" in comparison to other technologies, and in the second place, you're still running an application in the jvm vs running an application in native code (the application being an interpreter running a script), and in these cases the jvm version performs better. It's really not that hard to understand.

They're not valid comparisons because they are ALSO SLOW - especially Ruby. The jvm will never beat a good coder in straight c. Ever. The best it could possibly do in theory is, after many runs, come close.

What part of a java application beating the c version of the same app don't you understand? (It is nice to hear you admit that 2 very famous c apps are slow; maybe there is hope for you yet... though I would have laid money on the fact that you would be one of these typical /. clowns who trash java but praise python; I myself am quite happy using both.)

Regarding firefox, the compressed image only holds true for the first run; subsequent runs better be almost instantaneous (because, hey, it was written in c, right?), which it's not... unless firefox is already running and open in another window, it's always in the order of several seconds to launch it. And besides, you're trying to dodge the argument on a technicality; I just gave you a bunch of examples, proving that using C and friends does not mean your application will be blazing fast; that is the point here.

No dodging - I pointed out that you're running a compressed file system - and if you did any checking, you'd notice that even after you load an instance, not all the code is actually in memory. But you wouldn't know that. Learn what the various columns mean when you look at the output from "top". Oh,, wait, what am I saying - I'm talking to someone who never learned a real computer language.

Java stopped being slow at about version 1.2, 1.3.

Bullshit. Everyone admits that swing is slow, bloated, and crap.

(Are you even capable of writing a swing app? How will you survive without pointer arithmetic?)

If a swing app is slow, 99/100 times it's because the developer was doing "stuff" in the event dispatch thread; something that is warned about in the swing docs, and mentioned all over the internet. Again, competent programmers and all that nice stuff...

Final hint: There is a reason java currently owns the enterprise space.

Really? So all those enterprises run operating systems written in Java?

Sigh... No. enterprise OS's themselves are not Java, and Java was never meant for systems programming. But statisically, a metric shit-ton of the middleware / backend stuff is built in java.

Stop the stupidity. Walk away from the keyboard, because you can't win this one repeating the advertising slogans of the Javanistas.

I will indeed walk away, this is my last post on the subject, but that's because I'm arguing with a zealot.

Not when you just showed that you don't even understand that not all of a program is resident in memory when the program is running (your firefox example).

Are you really that stupid, or are you trying extra hard especially for me? Firefox starts slow, runs slow, looks like shit, pisses all over the available memory in my system (though they have been working on this in recent updates). Feel free to replace firefox with openoffice in the previous sentence as well. Both C apps. Not that I hold that against the devels though, I'm quite happy using the programs.

Get a couple of decades experience and come back and argue.

C'mon gramps, no... "get off my lawn?" Maybe the reason why you don't like java is because it wasn't spoonfed to you by a teacher? Can't handle learning any of these "new fangled" technologies coming out?

But by then, it'll be irrelevant anyway - java is dying, [snip fanboy whining].

...confirmed, you're an irrational, blabbering zealot.

Comment Re:I doubt this is Oracle's motive (Score 1) 341

Your post is wrong in every paragraph; the lack of pointers in Java is a feature; it's designed that way, and is one of the advantages of the platform. Remember that java is not competing in the kernel / driver space; it's designed for networked / business / enterprise applications. In that space, developers playing with pointers means that I can crash an app by uncovering a bug in the toolbar (happened to me the other day in a kde app; click on a button, entire application goes down with an incredibly helpful "segfault"). Again, I have no idea what you were trying to do in your application that you benchmarked, but trust me, if it's 50 times slower then the c version (like you claim), then you better think about hiring a competent java developer, because you're most certainly doing it wrong.

Python and Ruby are perfectly valid comparisons; at the very least, it shows that the jvm is not "slow / bloated" in comparison to other technologies, and in the second place, you're still running an application in the jvm vs running an application in native code (the application being an interpreter running a script), and in these cases the jvm version performs better. It's really not that hard to understand.

Regarding firefox, the compressed image only holds true for the first run; subsequent runs better be almost instantaneous (because, hey, it was written in c, right?), which it's not... unless firefox is already running and open in another window, it's always in the order of several seconds to launch it. And besides, you're trying to dodge the argument on a technicality; I just gave you a bunch of examples, proving that using C and friends does not mean your application will be blazing fast; that is the point here.

I'm sure it would be fun to argue with you all day, but the bottom line is that Java, when used where it is designed to be used (so don't try and write a device driver), is a very competitive platform by any metric; performance, stability (whole classes of bugs that you find in c just don't happen in java, like segfaults), maintainability, scalability (all the way from cellphones (blackberry) to the enterprise (j2ee)), flexibility (run it on any os worth running on), etc. etc. Some clueless dolts here on /. need to get over themselves and stop repeating garbage from 1995. Java stopped being slow at about version 1.2, 1.3.

Final hint: There is a reason java currently owns the enterprise space.

Comment Re:I doubt this is Oracle's motive (Score 1) 341

Nice that you've ignored my comments about all those "blazing fast" c apps I listed; I however, am not going to argue with you over your benchmarks that I don't have in front of me.

I'm not denying that C will do better in raw number crunching, video encoding etc. It probably will, but there is not as much difference as you people like to think there is. And if you take your own advice and research it, you will find cases where the jvm outperforms the c equivalent (python and ruby apps run faster in the jvm then they do in their native c implementations).

And I'm not denying there are slow java apps (netbeans, eclipse), but if you want to go down that route then maybe you can explain to me how it is that firefox / ice-weasel can take a solid 5 seconds to appear on my screen, when the live-cd I was using (on my dual core with 2GB ram) was already completely loaded into ram? Have fun with that.

Comment Re:JavaFX (Score 1) 379

You go ahead and create your gui's by hand then, have fun with that. I'll use a tool that generates the exact same code for me in about 1/1000th of the time.

I wasn't talking about web tookkits in the firstplace, but gui design is difficult because gui design is difficult. You simply can't get around that unless you're making a simple popup dialog or something. Use tools to make your life easier, that's what they're there for.

Comment Re:I doubt this is Oracle's motive (Score 1) 341

Oh, you "looked at it", how nice... that makes you quite the expert.

I always get a kick out of "developers" like you who know nothing about the java ecosystem, but feel quite the need to bash it and exclaim how slow it is (my personal favourite rant). I've been programming java professionaly, for the last 8 years, and I've used it for at least 3 years before that (in school etc), and I'm telling you, you're wrong... in your words, "get over it."

Neither the jvm, nor libraries like swing are slow (both of these myths are repeated by "developers in the know"); but of course, it takes a somewhat competent developer, and I will admit that... it's quite easy to make a slow gui app in Java (for example), but looking at applications such as OpenOffice, or even the Mozilla Firefox / Thunderbird apps, (all programmed in C / C++, btw), simply saying "Learn c" is quite stupid on your part.

How I wish C / C++ and friends were the magic bullet you think it is; I wouldn't have to put up with the crashing and slow redraws in Kde / Compiz / Gnome / Xorg etc.

Slashdot Top Deals

"Someone's been mean to you! Tell me who it is, so I can punch him tastefully." -- Ralph Bakshi's Mighty Mouse

Working...