3D has practically taken over video gaming. Lifelike, if not very pleasant, worlds exist aplenty–worlds that most users find easily navigable without any training whatsoever. Is the world of spreadsheets, word processors, and the like just unsuitable for 3D? Is it a case of “If it ain’t broke, don’t fix it”? Or is it that we’ve lacked imagination? John Littler recently talked to Hideya Kawahara about an open source 3D desktop project that he started and that Sun subsequently took under its wing.
Looking Glass doesn’t seem like “true” virtual 3d to me. It deals more with 2d planes in a 3d space. I wouldn’t even consider it 3d really. Not that there’s anything wrong with that. Something like 3DNA Desktop is more what I would consider a true 3d interface.
http://www.3dna.net/products/desktop.htm
don’t forget http://www.opencroquet.org/
thinking twice, the idea is deeper than only 3D looking
Sun is dropping the ball bigtime with Looking Glass. Here they have something that’s nothing less than revolutionary, jaw-droppingly advanced, yet smooth and fast as the same time, and they aren’t doing sh*t with it, treating it like it’s some kid’s project to work on in their spare time with no future whatsoever.
We will have 3D interfaces someday, but it won’t be Looking Glass from Sun. I remember when Looking Glass first came out in 2003, we were shocked at how amazing it was and couldn’t wait to see Sun include it in the Linux distro they were selling at the time. Since then? Nothing. Zip. They cancelled their Linux and are leaving Looking Glass in the ghetto. Sad. But typical of Sun. No vision. No innovation.
Sun is dropping the ball bigtime with Looking Glass. Here they have something that’s nothing less than revolutionary, jaw-droppingly advanced, yet smooth and fast as the same time, and they aren’t doing sh*t with it, treating it like it’s some kid’s project to work on in their spare time with no future whatsoever.
No it is not.Looking Glass is jaw dropping until you look at videos. Trying it is a completely different story. Anything but a usable project. At least my personal view on usability says that (after I was actualy playing with LG).
3D effects are cool until they are powering 2D views. Anything more is just confusing for average user. I used to love enlightenment, now I wouldn’t go back from Metacity, it just works and no tweaking.
Sun is just using LG to show Java possibilities, not to deploy desktop.
We will have 3D interfaces someday, but it won’t be Looking Glass from Sun. I remember when Looking Glass first came out in 2003, we were shocked at how amazing it was and couldn’t wait to see Sun include it in the Linux distro they were selling at the time. Since then? Nothing. Zip. They cancelled their Linux and are leaving Looking Glass in the ghetto. Sad.
[sarcasm] Yep, Schwartz and McNeally have yet to decide on which planet to resell their products. Yesterday they were mentioning Venus if I remember correctly.
[/sarcasm]
Thing that their clear vision on their bussiness direction is different every day is what it is bothering me too.
But typical of Sun. No vision. No innovation.
As much as I’m the first guy to bash Sun and Java I can’t agree here. If I like it or not. Java is a valuable contribution and inovation. Just as is OpenSolaris. And MS.Net (or better Mono which is really crossplatform) should be treated the same.
Personally, I have yet to see one single usable (covering all facts: I need it, consumes ram as any normal app, looks and works native on my desktop, speed is already satisfactory) application in Java. Although I suspect .Net already leads here (at least in my case). Mono and Gtk# on Linux, mono and Cocoa# on OSX and MWF on Windows (be it with mono or ms.net).
Personally, I have yet to see one single usable application in Java.
I can’t live without JEdit ( http://www.jedit.org/ ). I fits right in on my Mac OS X desktop. It can be a little slow when you don’t use it for a while (I presume this is to do with the OS putting it into the swap space) but it is great to use when you’re working with a few different languages.
Then there is Jext, then there is Azureus, and a whole bunch of others out there.
I fits right in on my Mac OS X desktop.
Wont speak about usability, that wouldn’t be fair of me since I don’t know the program.
But…
http://www.jedit.org/index.php?page=screenshot&image=22
Bad OSX theme. Differentiates everywhere. Spacings, alignments,…
Tollbars???
Vertical tabs???
Although I must say (until now I tried them on Windows and Linux only. Looks on those like piece of crap. So I tried as you use it and installed it on my OSX which I tend to avoid as much as possible) it fits preety good, probably first Java software that at least to some degree fits in some OS. Now if they would done the same elsewhere (other OSes) that would be nice. MWF on mono is in process of doing just what I talk about, but they are not even close yet
The interviewer constantly re-iterated the “Java is slow” myth. Modern JVMs are not much slower than other high-level languages, and his questions about how they “accessed the video card” are completely stupid. Does the interviewer even know ANYTHING about the material he is supposedly interviewing on?
The developer sounds quite nice… I think this project is massively underhyped, and could bring about some niceties. For one thing, I support any project that makes the Linux desktop crave better graphics drivers from proprietary vendors, so at least they’ll provide them.
Well, if it’s a myth, then please explain why every Java app on my machine is painfully slow while apps written in C and C++ are not? This is true for Windows and Linux alike. I have, in fact, yet to experience a java program on any machine/platform, that isn’t at least a little bit slower than native apps. Maybe I’ve just been having bad luck, but based on what other people have had to say, I doubt that I am in the minority.
I have been using the latest JVMs, by the way.
Yep, it’s a myth, at least partially.
I have been using Java for text-only applications (scientific work) and know of a lot of people who do. In that case it’s only slightly slower (maybe 15% slower, maybe sligtly more or less, I don’t remember the figure).
However I think the story changes when you include the graphics. But I am not 100% sure as I don’t use them.
That’s fair. I have actually written Java text programs to do calculations and simulations and they run decent (I couldn’t tell any difference between equivalent C/C++ programs). But in the graphics realm, it’s horrible.
this is my conclusion about that:
i think it’s a issue with any current JIT compiler: it indeed makes the execution fast but sucks a lot of memory, just read the psycho intro: http://psyco.sourceforge.net/introduction.html.
Besides the JIT constantly scanning code pages or the garbage collector moving objects between generations and compacting data, forces to keep pages on RAM. Testing the program running alone it performs fast and even faster than traditionally compiled ones, but running in a desktop, it ends on constant disk trashing.
I can’t see why you’d want your desktop to actually be 3d… While there are some very useful things you can do: Push a window way back so it’s small, or pull it closer to enlarge it, or clip it over and lay it on the “ground”, etc; none of them are feasibly useful with a mouse/keyboard/2d screen.
The project looks like it’s trying interesting things while sticking to reality: People don’t want their PC to look like a room; and that’s why every project that’s done it has attracted almost no users (eg, Microsoft Bob).
I didn’t notice this in the article. But, last I knew, looking glass was x11 compatible so you could run x11 toolkits on it; anybody notice a mention of this in the article?
This is what Sun’s best at.
They make amazing technologies that others will spend 5 years playing catchup to (or longer) then they move on.
Or, they make amazing technologies and want so much money for them that the only people who can afford them think they’re expensive and constantly look for an alternative.
Sun has some very smart people. Not many good artists, and doesn’t seem to have any crafty businessmen: The latter being why I like them, and probably also why their stock is under $10 (is this still true?).
Edited 2005-12-10 22:56
Looking glass is cool in a “hey that’s neat” kind of way, but I just can’t see it being actually useful. Every feature (aside from perhaps the file manager) is merely a gimmick. Sure, it looks kind of neat to pick a wallpaper from a spindle, but it would be far more efficient to use the same space to show a simple list of the wallpapers with a small thumbnail. Also, why would I want to take notes on applications? It’s completely useless. If I’m gonna take notes, the only viable place is on the desktop, or in a separate app.
Same goes for the configuration on the back of applications. Sure, that’s kind of cool, but the whole concept of “flipping” the window is ridiculous. You might as well just have a shortcut bound to bringing up the application’s config dialog. The physical concept of flipping a window brings exactly zero benefits to the table.
Tilting a window could be useful, but I can’t think of many situations where I would make use of it. How often do I really need to keep an eye on a window, but not work in it?
3D is just not very useful on a desktop. Shadows and transparency are, but I haven’t seen a good use of real 3D yet.
When I first got a mouse (1988) I couldn’t figure out what it was for, now I use it continuously. Windows 1.0 didn’t overlap windows so I think the initial 2D desktop wasn’t so great either. The problem is that you can’t imagine new things to do with the hardware.
Some examples of things the 3D hardware can do:
1) GPU generated scalable, antialiased glyphs
2) Mixing of true 3D and 2D apps on the same desktop
3) Use of a depth buffer instead computing all of the overlap in software
4) Hardware compositing
5) Hardware generated highlighting and shadows
The first version of everything new usually mimics the previous generation. The right answer is to get an initial 3D desktop running and then let it evolve for a few years. Only then will we discover the valuable features. My bet is that the final useful form doesn’t look like any of the exisiting systems.
You know he’s talking about Java, not Python….
I think the reason he’s seeing slowness in Java is mostly because: Swing is slow. Swing is also terrible as far as keyboard shortcuts, and other things; but hey it’s a sacrifice you make for easy portability and a nice language.
Someone once told me, on here, Java is slow, it’s just good at optimizing repetitive algorithms so it ends up running well in benchmarks and server code but it craps it up in callback driven desktop code. It’d make sense, but I’ve never seen proof of it.
And btw, for Python devs, Psyco is awesome…
yes, i was talking about JITs.
i know that formerly java was interpreted, just like python is right now, and it was slow, but at least it was OK using RAM, now it uses JIT, and is not so slow but tends to take all available RAM and sends other processes to swap.
i suppose a solution to this is to run all java applications as in one process, one example: the desktop environment is one process and applications are loaded like plugins.
Hmm…shortcuts in Swing? I thought it was up to the programmer to code in the shortcuts? I have written Swing apps where I have used keyboard shortcuts so I dont really understand what you are talking about.
Swing is not really slow. I mean there are glaring drawbacks in the 2d/3d space with Java and the annoying “gray rectangle” problem but that is actually possible to be handled quite well already. There are ways to use threads and so on to make gui’s written in Java run quite well. IT is not so much that Swing is slow…but rather I think that it is rendered through software. I think with the Opengl and DX bindings getting better and better we can see in Java 6 a massive improvement in the feel for Swing apps. Java has its faults but they are being worked upon I feel. It is definitely not suitable for everything though. Startup time is too long.
1, 2, 4, 5; all can be done on a CPU with less than radical designs; and it’s been done.
Everything can mix 2d and 3d apps, and there are already 3d apps around..
Software compositing is what OS X 10.0, 10.1 did. It was slow, of course.
5 is something I’ll never understand. Shadows are pretty, but functionally they’ve done little for me (I’ve used it with and without); and they’re also not very expensive to generate on the CPU (check out enlightenment 17 for an example). Of course, the trouble is when you hit overlapping your 2d cpu rendered shadows with a gl rendered app or video; can you say ugly blue?
They’re not saying 3d hardware is useless: They’re saying true 3 dimensional interfacing is. The nice things you can add to our current 2 dimensional interfaces is great, and we should have it. But your desktop shouldn’t be a room, there’s no solid metaphorical connection between a PC interface and an living room. Opening Word is not analogous to grabbing a notebook off the table….
I think what Sun is doing with Looking Glass is cool, but I’m guessing 90% of the cool things you see wouldn’t make it into a shipping product because they’re more obnoxious than helpful.
They’re not saying 3d hardware is useless: They’re saying true 3 dimensional interfacing is. The nice things you can add to our current 2 dimensional interfaces is great, and we should have it. But your desktop shouldn’t be a room, there’s no solid metaphorical connection between a PC interface and an living room. Opening Word is not analogous to grabbing a notebook off the table….
I think what Sun is doing with Looking Glass is cool, but I’m guessing 90% of the cool things you see wouldn’t make it into a shipping product because they’re more obnoxious than helpful.
>
>
Of course 3 dimensional interfacing is useless. What purpose does it serve? None that I can see (pun intended). It’s the ultimate example of a cheesy solution looking for a problem to solve. Much like the GUI was.
YET ANOTHER attempt at making something that’s been simplified over the past decade and making it overcompliated and confusing again – or worse just bloating it all to hell for eye candy.
When my 90 year old grandmother can sit down at a XP box and pull up her e-mail, browse the web, figure out how to IM people and type up letters, YOU ARE DONE DESIGNING THE @#$%ING UI.
These alleged genius programmers need to quit wasting their {censored} time with goofy {censored} like 3d effects, transparancies and redesigning things people already know how to use, putting their efforts towards fighting bloat, making their chosen OS more secure, have better hardware support, consume LESS resources and just plain run faster… and that goes for Linux, Windows, Solaris AND OSX.
Ugh, reading closely it sounds as though someone owes you something. For every person who crusades against change, for whatever reason, there are millions of people who are happy to welcome it, once they’re presented with it in a right way.
Why so angry?
Everybody is saying 3D will be the future.
Our brain only creates a 3D illusion – we are not able to see 3Dimensional.
Computer screens are flat.
Normal mice are only of use for 2D looking desktops.
It very much again seems like a “if it aint broke dont fix it” .
With current computer hardware 3D is IMO just eye-candy & will eat up rescources for fun.
If these very expensive 3D systems used for e.g. car-design were standard then maybe we can think about a 3D desktop – or else it will just be a cool “thing” of very little use in its’ own right.
Why are computer desktops still 2D organized ?
Because there is no need for the average computer user to be able to interact with what he/she sees as “real” objects (e.g. to be able to scratch off the window decoration of a window) instead of “just” flat rectangles.
It just is not neccessary when editing an 3Mail.
The content of the window is important.
When the content can be better made more “real” in a 3D environment than a 2D desktop – then I think it will be used – well is – for 3D design systems- else it is IMO a waste just like fancy drop shadows etc.
Sub-Pixel-Hinting or wahtever it’s called … actually improves the functionality – makes it more usable – of a desktop computer.
How many of U actually have alll the transperancy & fancy stuff available under Linux turned on ?
For me it just gets irritated really quickly – off topic – but that is what I cant stand about OSX & – I guess again – that is propably also something people love of many things about BeOS.
A clean slick fast UI is a more efficient way of getting things done.
Notha one – what is sexier – a fancy looking Vauxhall with a million spoilers & bits Or a really amazingly fast very expensive supercar ?
Okay might be a bad example AND/OR question … but the supercar is built to furfill one main purpose (be fast I guess) whereas the average car – no matter how many bits U glue onto it – will still just remain an average car.
….. sorry slowly getting really boring & selfabsorbed – overopinionated … & & off-topic.
A 3D desktop IMO makes sense if we are able to interact with it in a 3D space & there is a practical need for it.
Too many words – too much rubbish.
Just IMO
While I mostly agree, that for most people, these kinds of things are mere gimmicks. But its free, its being developed for those people who like mucking with new concepts.
Usually the effect is: early adopters try something, and rest of the population aventually comes to like it once all the work of early adopters (features requests, bug fixes, other improvements) is done.
I wonder how much this Looking Glass thang is just a PR project for Sun. People just like the readers of this site are supposed to go to see those demos and sscreenshots and (so hopes Sun) get impressed how cool it all looks… But is this project being seriously developed as a future desktop – besides of just producing supposedly cool looking screenshots and demos? Well, maybe it is but I ain’t sure about it at all yet.
After all the ooh and aah, how much of all those 3D gimmicks are actually enhancing anyone’s desktop productivity or usability? I’m not too sure about that either?
Some “true” virtual reality 3D technology is quite a different thing from Looking Glass, but also there are big technological hurdles to make that actually useful except in some restricted entertainment and demo setting. Or could you imagine working with heavy 3D glasses on your eyes all the time? Not me…
Not everything that is called revolutionary new technology is actually useful.