Submitted by “On Tuesday, January 25th Mandriva introduced a new project: Metisse LiveCD. In this article we are going to investigate the features offered by this promising project and see if Metisse can compete with the popular desktops in terms of ergonomics and ease of use.”
I just hope this project can make FVWM become “mainstream” (I know it is alive, but let’s face it, it’s not in the news for quite some time). Also I think this will be a great push for developers writing configuration applets for FVWM integrated with a Desktop Environment (That is, probably GNOME for Mandriva, I guess)
And is surely a good thing to have different approaches to the 3D Desktop metaphor. Competition between Compiz/Beryl and Metisse may prove a great thing for the OSS community, much like the competition between Gnome and KDE did.
Let’s just see where will this go.
I can finally rotate my windows! I’m so happy! Now all I have to do find a reason to rotate them. I fail to find any use for this, beryl/compiz gives me much more in terms of usability.
Anyway good thing about open source is that people come and go with different ideas, the good ones stays, about bad ones You hear only when mandriva or someone else makes a hype about.
Edited 2007-03-08 17:40
A window manager is not a game. I don’t care about neat tricks, I care about usability and features to make me more productive. So what are the features of Metisse according to this article?
1. Virtual desktops and a pager, nothing new here.
2. “Bird view”, aka, Expose for virtual desktops. Neat trick I guess.
3. Window transformations. The article claims that “this is what makes Metisse much better than any 3D desktop”. In ~12 years of computer use I’ve never once had the desire to rotate a window in 3 dimensions or map it to a sphere. Yet somehow this seems to be a major feature of Metisse.
Duplicating a window also doesn’t seem to have a lot of utility, not to mention being confusing. I remember seeing a demo showing the ability to make a mirror image duplicate of a window, which makes even less sense.
The copy-paste trick is neat, but a very niche scenario. This is only useful if you want to copy from a partially obscured window where at least the beginning or end of your desired text is visible. Can’t say that has happened a lot that I can recall.
4. Activate on hover. Nothing new here.
5. Keyboard shortcuts, yay.
I don’t want to be this pessimistic, but I can’t even think of a single scenario where Metisse would actually help me be more productive. Its one or two interesting features would better be integrated into existing window managers.
I can see an instant use for the copy and paste tricks, and like the “type behind the window with focus” trick. I think some of it is bizarre, though, I can agree with you there – I can understand that wobbly windows and drop-shadows actually have a purpose – “passing through the looking glass?” Not so much.
Having something popping up in front of what you are doing is one of the effective way to get the users attention. It tells him that a shift in context has happened. Drop shadows strengthens the impression of being in front. In that sense drop shadows could provide a slight improvement of the usability.
Finding a logical explanations for wobbly, or even worse mirrored windows is much harder. It is nice to see how new technology gets invented but its even better when it actually have some sort of use. In this respect compiz and beryl seam to do much better.
It’s a great way for my monitor to get a flying geek-punch to the front of it.
One of the most impressively annoying features of the Windows paradigm is the continual focus-stealing if you try to use the machine, particularly multitask.
According to the article, a 3D accelerator is not required. That’s a plus.
As you have noted though, the stuff demonstrated in the article isn’t too exciting. I don’t care much about window tilting and rotating either. Scaling would be nice though. Going over to the actual Metisse site you can see that there is a bit more to it.
The User Interface Façades page looks quite interesting. The ability to customize your UI in this manner could be very useful.
Is Metisse only good for playing with windows?
Of course not! However in the current version (that comes with Mandriva One) not much more can be done.
So currently it is good for them who curious about what is drawn in the backside of a window?
No one had the ‘desire’ to use expose either until someone invented it. That’s kind of the _point_ of innovation: coming up with ideas people haven’t thought of before, but which they find useful once they try them. Maybe you should try Metisse and see whether you discover new ways of doing things with the abilities it provides, before dismissing it offhand?
“No one had the ‘desire’ to use expose either until someone invented it.”
I did.
The ideas behind Expose have been around since the ’60. People like Alan Kay and Jef Raskin have been presenting interface ideas like this for decades, and I for one have been eagerly awaiting/developing them. If anything, Expose doesn’t go far enough, as those windows are not live, and mess with spacial searching, but I digress.
Expose was also DESIGNED to address a PROBLEM. Good design begins with a purpose and an understanding of the problems inherent in pursuing it. What problem does Metisse attempt to solve? How does it solve this problem more effectively than other existing solutions?
It’s nice to have cool toys on Linux, but right now I’d rather have cool tools. YMMV.
Copy-paste thing it’s just an example – you don’t have to copy anything – you can simply click on the window that is covered by another one and move you mouse to make the covering windows bend – when you release the mouse key windows will back as they were on the beggining – it’s useful (belive me, I know it because I used it) even if you don’t see it now.
Moreover the ‘bird view’ thing is really great in work, when you have a lot of things to do (you just have to get used to it)
Metisse is made to be integrated with other graphical environements (like GNOME now or KDE later), it’s not going to be a “another-one-preety-but-not-usefull-environement”
But I think it looks really good, and the features do actually seem useful to me. Why can’t using an OS be fun, anyway? Not everybody wants to be stuck using Windows or something similar. I like how different this is.
Metisse…that’s French for lame, right?
> Metisse…that’s French for lame, right?
No.
“Métisse” means something like “mongrel”
Brewmastre> Metisse…that’s French for lame, right?
superman>No.
“Métisse” means something like “mongrel”
Thanks for the translation, but it was just a joke
But in all seriousness, it doesn’t seem as though it would be all that useful in it’s current state.
It sounds neat. I like the bit about setting window transparency; that could be very useful for seeing information in other windows. The rest of it looks very un-new, however. Advertising this with the question, “You Thought You Knew What 3D Was?” is overselling it.
I like the bit about setting window transparency; that could be very useful for seeing information in other windows.
I could never appreciate window transparency. I would rather shell out money for a second monitor and view multiple windows simultaneously. I can see how it’ll be useful if you have a lot of blank space in the app window so that details don’t clash with one another, such as large amounts of text being overlapped.
Someone else mentioned drop shadows as being useless (I’m paraphrasing). Consider Mac OSX and how it seems to encourage floating windows. Focused windows have a drop shadow to give the appearance that they’re standing out above other windows and those out of focus windows appear to be faded back. So when you’re looking at a bunch of windows from multiple apps and none of them are overlapping, the drop shadow is a nice visual cue to tell which window is active. If/when 3D displays become more mainstream this effect would look even better because the windows will litterally look like they’re popping out from the screen.
Desktops shaped like a mobius strip … think of the possibilities.
Desktops shaped like a mobius strip … think of the possibilities.
I assume they are endless!
Just lots of desktops on a loop. Kinda boring
But the desktop switching could be fun 
Oh, come on! Imagine how weird would it be having only one Desktop in a moebius tape. You get a window slightly off-screen and it shows on the other side of the display.
Gee… That would be weird.
Well yeah, with a translucent background, it would be kind cool
Finally we have a reason for having built high performance 3D accelerator graphics cards in all the new “treat me like a typewriter” PCs. Along with “Vista”, Compiz and Beryl, Metisse is nice if this is what you need. Sure, using an OS (or an application) can be fun. As it has been mentioned before, some of the concepts Metisse “introduces” are not new. Some of them exist since 1980. But hey, we’ve got enough ressources! We need a CPU and GPU intensive window manager / desktop environment. And if the PC is to slow, just let’s buy a new one. Buying is good for industry.
What did I want to say? 3D desktopping offers new concepts of what a GUI is. Window contents as textures, mapped on a collection of vertices that represents a task, polygons representing projects. What about shading? Could the way of lighting be a representation of a programs priority? System messages are lighted with specular light, while “usual” work is lighted ambient? Translation and rotation matrices representing an application’s status? And could a raytracer soon be an essential part of a GUI OS? But finally we need to project R3 onto R2 because the screen has only X and Y, but no Z. Wow, what a playing field for a mathematician!
That’s what we need 3D hardware for, in case we are no gamers. 🙂
It’s not that I don’t honour innovation, but Metisse is surely not intended to be used by myself. I even like some of the concepts, because they are not new to me (focus follows mouse, copy + paste with only one middle mouse click). But, as it has been mentioned before – nothing new in general. But nice and useful.
So I’ll stay tuned. 🙂
Not to disagree or anything but allow me to quote from the article.
“Metisse does not require a 3d-accellerated desktop to work. It is not necessary to have a supercomputer to use it. Metisse performs reasonably even on elderly computers.”
Not to disagree or anything but allow me to quote from the article.
“Metisse does not require a 3d-accellerated desktop to work. It is not necessary to have a supercomputer to use it. Metisse performs reasonably even on elderly computers.”
Ah… elderly computers, such as Intel Celeron with 3.2 GHz and 1024 MB RAM and ATI Radeon GPU. 🙂
I’ll give it a try on some systems (from 300 to 2000 MHz with various ATI and nVidia GPUs) and see how it performs.
Metisse can get a good promoter for Linux on the desktop (such as KDE 4 surely will be), because people like eye candy, spinning windows, flipping, clipping and puppies and dancing elephants. 🙂
Metisse is layered ontop of the root window much like Xgl is an x server layered ontop of an X server. By design, this is a *BAD* idea. I went through the hoops and compiled a copy of metisse on my Ubuntu box.
Not too bad and it has some nifty features, but it seems slow. The idea is cool, but (in my own personal opinion) this isn’t all that impressive as it doesn’t use things like composite which fundamentally limit it.
With the exception of the windows transformation feature all of the listed perks are standard FVWM2 capabilities. The amount of customizations you can do with FVWM is staggering (though it takes quite a bit of work in the conf files.).
While I have not used FVWM for a while there were a couple of extremely long threads (well over 1000 posts each) on the gentoo forums discussing and showing off different configurations.
The big question has got to be why they’re trying to push this as something powerful, exciting and impressive. It’s obviously just a fun experimental project and should be treated as such. Not “the next steps in Linux Desktop evolution”.
Yes most of the features are something we have seen before. And yes many of them are more eye candy than useful.
There is one feature that looks quite innovative and quite useful. For some reason they buried mention of it in the section about Windows transformation, so I assume most people looked at it and though “Yaa, useless angled windows” and skipped on the the next point.
I am talking about the bending of windows when dealing with obscured windows. To me this functionality is a godsend for copy and paste and drag and drop. All of the quick operations that require you to briefly go to another window would require so much less fumbling the arranging your windows, or pressing the correct hotkeys.
I really hope the environment I use copies this feature, or Metisse gets to a point where I can use it fulltime.
So I spent a few hours downloading the LiveCD via torrent last night only to have it kernel panic on my laptop. Great livecd you’ve got there, Mandriva.
Adam
I must say that this “folding windows” thing is disturbing for me (in the meaning that, it doesn’t help me, only is confusing to sort windows).
Beside this one feature, I’m using beryl and compiz which are rather fast(er then metisse) (40FPS) on a intel 855GM GPU (64megs agp aperture), 1.1Ghz CPU (running at 600mhz usually) and it provide similar features.
I’m sure mandriva wanted to have “their own thing” for marketing reasons, but I fear they’re lacking ideas to make it truly innovative. Of course, finding the right ideas ain’t an easy task and takes a lot of tries…
Instead of using the existing technology they’ve forked their own X-server. That’s just duplicate effort. I also wonder what’s so special about metisse. By looking at those screenshots and videos I’d say there is nothing metisse does that couldn’t be done by writing plugins for compiz. Or is their technology somehow superior?
The article also says that metisse doesn’t need a 3D accelerated desktop to work. What does that mean? That it doesn’t need XGL? Or that it doesn’t need any 3D acceleration at all? If it’s the later metisse would be really slow and use a lot of CPU for sure.
Why do people even waste time making thing like this.
Because they can.
Why do people even waste time making thing like this.
I’m still wondering that about GUIs
you can also use beryl without 3D hardware. I used it with mesa software renderer and it was slow on my 1.1ghz cpu. like, 0.2fps (=pain)
I’m sure on a faster cpu you can get half decent framerate like 20fps