The Mockup project published screenshots of the recent released source code. The development of the GUI is following these mockups, using Qt 4.0 beta1 and key technologies like HAL and DBUS. The announcement, with instructions about how to compile source code, of the latest release is here.
Mockup screenshots and new release
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSAlert.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
In the mockup pics, where did they get the background (the one with the buildings) from? That looks cool.
Let’s see how the replacement of /etc to Elektra (a registry system) will work. I bet it will feel big different for Linux/BSD/UNIX users.
Will it be limit to Linux or possible support more platforms (BSD?) in future?
Where did this one come from? Or is this a renamed project from one of the earlier BeOS clones.
This may be picky, but I don’t think projects like this should be under the BeOS heading. This is linux, not BeOS.
Yellowtab, Haiku and true BeOS, are BeOS. To put this there is a stretch.
It’s a photo of Bilbao made by our graphics designer.
The photo will be available next year in a package with wallpapers.
I believe the building is the Bilbao Guggenheim museum.
http://www.guggenheim-bilbao.es/ingles/edificio/el_edificio.htm
> Let’s see how the replacement of /etc to Elektra (a registry system) will work. I bet it will feel big different for Linux/BSD/UNIX users.
I like very much the idea of Elektra, it’s registry done better. It uses text files and a directory hieararchy, well 0.4.9 have a plugin system so the file system is supported by a plugin and you can write your own (for example you can store your configuration to Postgres).
> Will it be limit to Linux or possible support more platforms (BSD?) in future?
I started reading the source code, I would also want to patch some small Linux packages (like module-init-tools and sysctl) and I don’t think Elektra is Linux specific.
It’s covered by BSD license so it should pretty cool for BSD core developers.
Mockup is an operating system that uses BeFree as its graphical user interface, BeFree was one of many BeOS projects.
Right now we are not trying to port the BeOS API but anyway the spirit and goals are the same: create a good desktop operating system suitable for home and multimedia usage.
What advantages will this have to KDE on Linux?
Is it a desktop modeled after BSD, or what?
The Docs are pretty light on the Mockup Site.
Did anyone notice the strange date format in the properties dialog? It is “month day time year”.
Also the arrow in the tree list view with the gray stripes has a poor contrast.
But all in all the shots look nice and clean. Plus the icons are pretty Be ish. I kind of like the rubics cube for settings.
I’m confused what is it?
People mention BSD, it says it’s Linux based on the website, it uses QT, yet has a BeOS icon on the osnews report? Eh??? WTH is it?
I take it is deliberately obscure so that when it doesn’t materialise as a finished something, we won’t miss it?
First of all:Congrats! I believe that it is an extremely good idea to start from a good foundation(like fd.o projects, a good toolkit like Qt and linux kernel(yeah, yeah, I know linux is just the kernel)), rather than trying to write everything on your own from scratch.
I also see that you are already using the bleeding edge stuff(like qt4, d-bus) which also is very good.
Second, would you consider hosting this project on fd.o?
Your goals seem to collide with fd.o people and both KDE and GNOME has benefited _massively_ from fd.o. (I really think that fd.o is the best thing to ever happen to the linux desktop:))
Last, what is the advantage of elektra when compared to gconf? In your website, you say that gconf is too heavy. Is gconf dependent on bonobo? Is that what you are saying?
Congrats and good luck:)
Thanks for your feedback.
Well, the date format is automagically localized by Qt.
The format here in italy is day month year so for example today is 31/12/2004 or venerdì 31 dicembre YYYY where venerdì means Friday and dicembre means December.
My locale is it_IT.UTF8 so I think Qt is doing a great job
The arrow in the tree list should be black, but I just started working on the GUI style.
About the icons… They will be done using SVG but I will continue to provide PNGs with many sizes (16×16, 32×32, 48×48, 64×64 and 96×96).
Unfortunately in the Qt 4 model/view architecture I cannot know from the model if the view wants large (48×48 by default) or small (16×16 by default) icons, so I always use large ones. This is because in the treeview they look so bad.
This should be fixed by SVG icons.
How is it different from BlueEyedOS ? A BEOS-like window system running on top of a linux kernel…
First of all:Congrats! I believe that it is an extremely good idea to start from a good foundation(like fd.o projects, a good toolkit like Qt and linux kernel(yeah, yeah, I know linux is just the kernel)), rather than trying to write everything on your own from scratch.
Thanks I’m very happy!
I spent a lot of time trying to do what I wanted and I just saw other projects failing too.
Using Qt makes you productive, half of the work is already done and you can start working on high level pieces of the puzzle.
For example I work on the API when something is needed by Tracker.
I also see that you are already using the bleeding edge stuff(like qt4, d-bus) which also is very good.
Yep it’s cool to take the latest technologies but sometimes when something doesn’t work you don’t know where the mistake is (in your code or in third party’s?
Second, would you consider hosting this project on fd.o?
Your goals seem to collide with fd.o people and both KDE and GNOME has benefited _massively_ from fd.o. (I really think that fd.o is the best thing to ever happen to the linux desktop:))
Mmmm I don’t know if they will host Mockup, KDE and GNOME aren’t hosted on fd.o (probably it’s because they have their hosting).
Usually fd.o hostes only projects that can be shared by KDE, XFCE, GNOME and Mockup is another thing; Mockup uses fd.o standards (I implemented only icon theme and base directory standards right now).
We’re using a good hosting, just need more money to get a full solution with apache2 and ssh support for Subversion (otherwise we’ll continue to use cvs) or someone that provides us svn, it could be fd.o but I should ask them.
Last, what is the advantage of elektra when compared to gconf? In your website, you say that gconf is too heavy. Is gconf dependent on bonobo? Is that what you are saying?
Bonobo is dying as far as I know.
I didn’t say much about Elektra because its Web site explain the problem very well.
Anyway /etc is a hell, every program has its own configuration file format.
Gconf tries to fix the problem but in a bad way. You don’t have a full directory hierarchy just many XML files that you have to parse with their pair of library and daemon.
Gconf just like many GNOME and KDE stuff requires a lot of libraries while Elektra is a simple and small library w/o daemons.
Elektra stores system configuration in /etc/kdb with a real hirarchy.
For example the font name is stored in system/sw/befree/interface/font_name and you have /etc/kdb/system/sw/befree/interface/font_name.
Congrats and good luck:)
Thanks!
This might sound a bit harsh but where is the individuality of this Operating System. What benefits does it have over Linux or any other Operating System outside ? The Mockup uses QT 4.x and now HAL and DBUS and as it looks to me every other POSIX compliant system starts to adopt these things and I ask myself why ?
I was always believing that the reason behind writing a new Operating System and Desktop Environment was to be ‘different’ because there are things on the current systems that people don’t like. What benefits does it have to start yet another Operating System if at the end it uses the same components as found everywhere else (speaking of duplication now).
In 10 years all the Open Source operating systems have adopted so much from each other (due to publicly available sourcecode) that it starts raising the question ‘for what ?’. If all the systems start to go the globalisation way then we have 100 different operating systems in 10 years but yet all of them share the same code, do the same, has the same tools, same toolkit, same bottom libraries etc. This totally breaks the concept of writing a different OS. From my own philosophy a different OS’ purpose is to be ‘different’ and not ‘become the same’.
… like XFCE is a “Gnome (GTK-based) light”?
I ask this from this point of view: speed of the GUI
Yes, that Mockup will be more, but hopefully Mockup becomes the XFCE only with QT-libs
Your screenshots looking very good! I think if there is a complete (qt4,qtdbus,elektra,mockup) binary tarball for eg. Fedora3, then many many users can test Mockup in realtime (hint hint hint
Ok ok, but it’s a long time to compile even qt4beta1…
And again people complaining about why yet another OS… Let’s just ignore those people.
Anyway, Pierre, great work! It was getting a bit quiet on MockUp.org (and I think I subscribed to the wrong mailing list), but I did catch the latest few updates and posted them on eXp Zone as well.
You also mentioned the availablilty of an iso release. Any timespan on that? I don’t want to mess up my perfectly fine GNU/Linux system at the moment
.
but would someone explain to us dafties (or just me), what it is, what it does, what it’s for???
Is it just a window manager? Is it a windowing system, kernel thingy? Complete OS??? Linux, BSD, BeOS? Relationship?
Anyone?
Seems like BeOS’ GUI just cannot leave many developers’ minds in peace!
It seems it would be a linux based operating system plus a beos inspired gui based on QT4, plus beos api’s (always created on a qt4 base) plus Elektra config system plus eventually other things. If I didn’t get it wrong
(Pier Luigi, correggimi se sbaglio ehehehe)
Hehhe in fact I don’t reply.
But I written http://www.mockup.org/documentation/faq/difference to let people understand.
I think Qt API is bettern than BeOS’ although the first time you see it you probably have some problems to understandand something.
For example it doesn’t use message passing like BeOS, Qt C++ classes expose signals (for example clicked() for a push button or triggered() for an action) that you connect to your handler.
Mockup inherits some concept of BeOS because it’s coherent and modular.
>Mmmm I don’t know if they will host Mockup, KDE and GNOME >aren’t hosted on fd.o (probably it’s because they have their >hosting).
>Usually fd.o hostes only projects that can be shared by >KDE,
>XFCE, GNOME and Mockup is another thing; Mockup uses fd.o >standards (I implemented only icon theme and base directory >standards right now).
>We’re using a good hosting, just need more money to get a >full solution with apache2 and ssh support for Subversion >(otherwise we’ll continue to use cvs) or someone that >provides us svn, it could be fd.o but I should ask them
Sorry for the misunderstanding, I didn’t mean that it should be hosted within fd.o. I was asking if you were willing to follow fd.o specifications, but it seems you already are, so I am happy:)
If you’re interested in this kind of feedback:
I think the gray stripes in the tree list should have less
contrast and would look better with full width.
There also should be some more space after the icons.
But that’s just nitpicking.
What are the plans for mockup regarding Qt4?
Nice piece of writing, and I mostly agree, but, I asked if there’s a timespan on an iso release
.
Don’t wanna rush you but I’d like to know.
Thanks for your work Pier, thoses screenshots are really great. I changed my mind and i now believe that a ‘Linux’ could make a decent desktop OS
your approach is professional, i praise the Mockup project. Maybe you could ally your project with B.E.OS ?
Does Mockup utilize Reiser4? Since Elektra uses an extensive hierarchy of individual files, it seems to me that Reiser4 would be suited well as a filesystem for it. Since Mockup goes bleeding edge with other components too, I suppose that using it as some sort of testbed for Reiser4 would make sense too.
is mockup a DE like kde/gnome?
and is it free or commercial?
how can i test it? is there RPMs/DEBs or any repository?
I like this idea, I like BeOS, and I like Qt!
<<is mockup a DE like kde/gnome? >>
From the FAQ:
First, Mockup is not a desktop environment. It’s an operating system designed for creative users. It uses a modern and lightweight desktop environment based on Qt because it’s an object oriented toolkit that makes us more productive, it let us work on higher level components like the file manager.
Mockup is different from Linux distributions because it includes only a small amount of the Open Source software. Each program is coherent and integrated with the system, you won’t find a program written with a toolkit and another written with a completely different toolkit. GNOME and KDE are general purpose desktop environments, they work on a lot of different operating system using X11. Mockup’s desktop environment is born for the Mockup operating system, it’s lightweight and doesn’t necessarily need X11. Right now we’re using Xorg for development, when Mockup R1 is reached we can concentrate on Qt/Emebedded to let our graphical user interface run on the framebuffer, possibly with DirectFB.
The desktop environment is done to be both usable and accessible. KDE is not a good example of usability, it has too many options while GNOME is going to kill choice by forcing users with GNOME’s defaults that sometimes are not sane for all the people.
Unfortunately GNOME is not coherent. Some things under GNOME can be configured too much. There’s a GConf key to use another window manager instead of metacity, this is bad in our humble opinion because metacity is coherent with the whole desktop.
We think that choosing good defaults is hard but with your feedback we can do it the right way. When doing GUIs, people should balance the availability of configuration options. Mockup’s desktop environment doesn’t let people choose the GUI style and icons can be changed just because people with disabilities have some needs. You won’t find a dialog box that let’s you choose between a sound system or another (like the Gstreamer configuration dialog box in GNOME) and you won’t find big window with a lot of options of the sound system (like the Sound system configuration in KDE). Mockup users should use Mockup while its developers must choose the best sound system to let you play music and enjoy yourself.
We use also new and bleeding edge technlogies and in some case we break compatibility. For example ALSA is the sound system used by Mockup and each program uses it directly instead of using the OSS compatibility.
As already said only a small amount of software that you can find on a Linux distro is included. You won’t find neither a lot of command line editors nor stuff that you don’t need. Choosing what is suitable for desktop users and concentrate the effort in the integration is the key for a good desktop operating system for creative users!
We would also create applications for musicians and artists to let people express their own fantasy.
yeah, quit bashing these people, if you don’t like it then dont use it, Personally i prefer Slackware for my purposes, but i still welcome Mockup to the Linux community. it is the very nature of OpenSource & GNU/Linux that encourages people to build thier own ideas in to another distro…
Sounds good to me, but why limit yourself to a BeOS clone GUI?
Why not design your own based on concepts of BeOS and implement them in a better way?
I know QT is a well-designed library, but I think it is a mistake for you to choose it. Now it will be impossible for people to do commercial development on your OS without paying a hefty fee to Trolltech. Maybe you should try a free toolkit like VCF (http://vcf.sourceforge.net/) instead.
I was always believing that the reason behind writing a new Operating System and Desktop Environment was to be ‘different’ because there are things on the current systems that people don’t like.
Well shucks. And here I thought the pointof writing a new operating system was to design something you and others like using. Who cares if they reuse existing technology? There is no point in fixing things that aren’t broken just to “be different”.
It seems that these guys and the BlueEyedOS guys have similiar ideas. But where they part is where the B.E.OS guys want to write a toolkit directly on top of Xlib and these guys chose the QT route. QT will definitely speed up development.
But what else does this have besides an /etc/ replacement (which I agree is a mess)? Are they replicating the BeOS API on top of QT and Posix?
In any case, the screenshots look nice. Good work.
Each program is coherent and integrated with the system, you won’t find a program written with a toolkit and another written with a completely different toolkit.
That sounds nice and all, and is a good goal, but in order to do that you have to provide a nice framework for developers. The whole OS needs to be based on components. D-Bus is a good start.
I would suggest having a high-level scripting language (invent one or reuse one) as *the* official desktop scripting language and obviously hooking it up with D-Bus.
The key is to keep everything locked down or you’ll have another linux distro.
Hey, you’ve got a head start on KDE already by using D-Bus
Desckops are one of those things that nobody ever seems to get right- theres always somethin we dont like about them.
Written in the wrong language, not enough config, too much options therefor no co-herence, too relyent on X11- too abstract therefor not enough speed. relys on hardware accel, not as portable. is/isnt posix. etc.
At a point I get tired of all this and use what best suits me, if I had enough time on my hands I might gripe about them and try code somthing better but as it stands Im happy to use icewm, if that didnt exist Id probably use gnome, the point is that there are lots of good DE’s and I dont realy care that much which I use, since most of the time Im using applications which is where the real productivity stuff is- not how I go about activating/removing an icon or what shade of grey my button is.
can’t this be done without the use of Trolltech?
If you want to develop proprietary software you will earn more money than a Free Software project, so you won’t have problems buying a copy of Qt.
it is still application-centric desktop environment.
What we really need is to go beyond the desktop paradigm and leave it forever. In Mockup you still have to close, minimise, maximise, rezise etc. applications. What a waste of resources. If Mockup removes this last obstacle then it could be a walid alternative not only to KDE & Gnome, but also to MacOSX & Windows.
It’s not application-centeric but window-centric.
If you use windows you have close, minimize, resize, actions is obvious.
But, how to replace windows?
I don’t like much application-centric desktops, for example instead of an Applications menu I’d like much a Tasks menu with a list of tasks like Play movies, Listen to music, Write a document.
I would just like to say that this project excites the hell out of me – it’s no longer a beos clone, but something that’ll take from teh best of all worlds, and build on one of the bigger strengths of beos – the UI. Good work, and keep it up.
Sorry Pier Luigi, I actually meant windows-centric
. I agree that task oriented desktop is much better then aplication-centric desktop.
“But, how to replace windows?”
OK, here is my idea:
1. Think www.
Web designers have solved the problem already. If you take look at Apple’s web site you will find a tabbed panel which can be applied (of course as task oriented) to “desktop” UI.
2. No desktop at all.
On your computer screen you can only see a task oriented panel and a application.
3. Applications are embedded.
There are no window managers, no minimising, no maximising, no resizing, etc. because application is always maximizied and opened.
4. All applications are always opened.
Task oriented panel is tabbed, and applications are always opened in the background. Something like a browser with tabs. You can have default tasks as tabs, and user can always remove or add task’s tabs. User can also assign specific applications to specific tasks.
wygiwyg,
How do you propose, using your description of a tabbed application manager to allow the user to see multiple applications at one time? If all programs are always open on tabs, how many tabs would you have? How do you propose to manage so many tabs? System resources? If all applications are always running, then the system resources would be used up. One good thing though… I usually do work with a single application completely maximized, and if I need two I use two monitors… I’ve always wondered about the strange fact that the non computer literate that I’ve worked with always seem to work with their applications using about 80% of the screen, and they always seem to have smaller monitors as well!
Tabs within a single application are great, though I’m not sure that I’m sold on the idea of all applications being tabs –> maybe within a subset (office suite), but not all.
Have you tried fluxbox? It seem to be a hybrid between what you’ve described and window based systems. It’s primarily window based, but allows you to group applications in tabs.
Yep, I agree it’s pretty hard to switch between documents easily.
I think SDI instead of MDI should be enough.
In the 5th mockup (it will be published soon on the Web site) there’s a better panel. It seems MacOS X dock but it’s different, it has icons for all minimized windows.
Applications are SDI, so for every document open you have a window, its title contains also document file name.
In the panel you have an icon for every application but the user don’t think about applications, he associate an icon with a task.
Example:
There’s a menu with tasks that may be done.
You click on Play movie and you can see the related icon on to panel. You open a .mpg file in your hard drive and another window of the player is open.
When the mouse pointer go on the movie player icon in the panel you see a list of open windows (with a close button) and some actions like “show all” and “close all”.
This way people don’t think about applications but instead they think about documents and actions accociated with documents (tasks).
“How do you propose, using your description of a tabbed application manager to allow the user to see multiple applications at one time?”
Right click on the tab and choose: add to screen (or something else). In Konqueror you have possibility to split views already. You can already embed applications in Konqueror.
So, I am thinking: we can have a “super-application” which is a tabbed panel and all other applications embedded in it.
I think I could better explain that with some mockups which I am going to upload somewhere, heh
Of course, you have to be able to work with two or more applications at the same time, but in a more intelligent manner. Now users waste a lot of time just resizing applications windows.
“If all programs are always open on tabs, how many tabs would you have?”
As many as necessary. Once again, if you take a look at Apple’s web side you will see what I mean. Look at the tabs on the top of the web site.
“System resources? If all applications are always running, then the system resources would be used up.”
Hm, on KDE desktop you have many virtual desktops, and many of Linux users have many applications opened and running in the background. Think also that many resources can be freed by not running a window manager and all other unnecessary things, especially moving, resizing, minimising and maximizing windows etc. etc.
“Tabs within a single application are great, though I’m not sure that I’m sold on the idea of all applications being tabs –> maybe within a subset (office suite), but not all.
”
Hm, I think you misunderstood me
I am not talking about applications being tabs, I am talking about a tabbed task-oriented panel, and applications are embedded. Applications don’t have window manager, because there are no windows anymore, you don’t need to resize them, you don’t need to minimize or maximise them, you don’t need to close or launch them: they are always there, opened and ready to use. Actually there is no desktop anymore, because you don’t need it. Why should an user need to minimize a window to find an icon on the desktop to click on it and to launch an application and to resize or move that application etc. etc. ? How many billions of times computer users do these stupid things like resizing, closing, moving, minimising and maximising windows? We are terrorized (sorry for the word used her 
by windows management.
IMHO there is no progress in computer UI without abandoning windows paradigm.
What we need in free software (open source, whatever) is to offer something totally new, and I would really like some brainstorming here, because I think that we have technology but we don’t use it. Why? Many people talk about usability, where to put this pixel, and where to put that pixel, which bauttons are better from the usability point of view, but none or a few are questioning the core problem: windows-centric UI.
Mmmh, now it’s explained better.
It would be cool to merge your tabbed proposal with windows.
When I bought my first Apple computer (in September) I saw how Exposé works.
It’s good and it’s possible to balance windows with tabs.
What do you think about a key that shows windows thumbnails organized by task.
If you don’t want to search the window in the panel you press a key and you can pick your window from the thumbnails of the current workspace.
With another key you can show a window with thumbnails of the workspaces, this window can be sticky.
With another key you can align windows automatically ala Konqueror.
It would be cool to merge your tabbed proposal with windows.
But then you would nullify the concept with a tab/task-based environment, since you would still have to rely on the windows/application paradigm. I like wygiwyg’s ideas, and I think he/she’s on to something here! This might be a nice easy-to-use environment!
I don’t like much this tab environment.
Maybe some mockup can change my mind
If someone would like to send a mockup or something else to let me understand better this idea, he/she can subscribe desktop development mailing list.
https://lists.berlios.de/pipermail/befree-devel
The background image I beleive is from the Walt Disney Concert Hall in Los Angeles. I say this because I work down the street from it and on occation I walk past it. The shot looks like the front entrence. For more pics:
http://wdch.laphil.com/wdch/building/exterior.html
The once title “Main Entrance” is the best angle to the background pic.
Before bushing the desktop-windows paradigm, what is wrong with it?
Is it uncomfortable to you to work with it?
If yes, then when, why and how is it uncomfortable?
If it is ok now, then why to invent something new?
There were some nice atempts to invent 3D desktops: from Microsoft (http://research.microsoft.com/adapt/taskgallery/), later by Sun, the most early was probably 3DWM. But no one of them succeed. Let’s try to learn from these attempts. If we want to make our desktops better then let’s do don do it blind.
Btw, imho there are many other ways how to make OSes better.
XML-based configuration system is great idea.
What about filesystem?
I personally think that OS/2 and Win filesystems are more clear to users then UNIX-like. In OS/2 and Win we can consentrate all OS programs and libbaries in one place, all third-part programs in other place. And never mix them together. That is an advantage, because just OS2 or WINNT catalog contains the whole system, looks more clear then /bin + /usr/bin + /usr/sbin + /X11/bin.