Who wants to develop a Desktop with me?

Johan_Perjus

New Member
Joined
Sep 14, 2020
Messages
9
Reaction score
1
Credits
71
Hello!

I am a graphic designer and UI designer with some fresh ideas for a desktop environment, better than the ones that that are out there IMHO.

Who wants to help code?

Linux Forum Screen Desktop.png
 


Just wanted to say sounds like a fun project, but I'm not a programmer. Good luck finding someone! What do you dislike about the available DE's?
 
Thank you! I feel like most if not all existing DE’s are bad imitations of either Windows or MacOs. I’d like to create something that is better than both those two.
 
Please, concrete your project.

What part would you like to change?
* Window Manager
* Frame design
* Mouse interaction
* Desktop Icons
* File Manager
* Task Bar
* Applications menu
* Widgets
...
 
Please, concrete the design to a lot of draws. Drawing defining the following questions:

Tabbed windowsClick opens on a new tab
Bread cumb style navigatorClick changes the folder of the current window.
There are links to upper levels
Separated WindowsEach click opens a new window.
The window manager draws the frames of all the applications. All them have the same graphic aspect (theme).
The window manager closes, minimizes or gets back the windows. All the windows have the same behaviour (- minimize x close)

Two columnsFTP (file transfer, upload...)
Left column is the local hard disk source, and right column is the public folder on the web hosting.
Top menu each windowFile Edit View... inside the frame.
But outside the client drawable area
Common top menuClassical Apple interface or GNOME 3 draw the menu on top of the screen. Outside each window frames
ToolbarIconized menus under the menu
Tabs mixed with toolbarIn your picture tabs an toolbar are in the same line
Lateral PanelLeft margin reserved for links to parent and common folders.
In your picture Home, Movies, Documents, Planning
Adjustable lateral panelThe user can reduce or hide lateral panel
Icon viewA folder or a page is drawn for each document. The page draws different depending of type (text, picture, movie)
Icon view with special foldersSome folders have special icon: home with a house, movies with a film...
Icon view with previewPictures are shown in miniature, the first frame of the movie is shown, the text margins are visible on preview.
Detailed viewList of file names with size, date, type and owner.
Context menuClick on an icon opens a drop down specific to the type of document.
Folder: Open, move, rename, compress
Picture: View, Resize, Crop, Edit, Copy, Cut Convert to png
Sound: Play, Edit, Copy, Cut, Convert to oga
Movie: Play, Edit, Copy, Cut, Convert to ogg
Mouse buttonsThe action clicking with a mouse could mean open, copy, move, compress, change format, rename. Left mouse, right mouse, mouse wheel, double click, touch panel, multiple resize touch.... Each one should be defined.
Drag & DropMoving a document to/from other applications could execute different actions. Drag to movie player shows a film. Drag to gimp edits the image. Drag to printer prints a paper page, drop to a folder moves file, drop to an program, executes the program...
Each one should be defined.
Applications menuA name, an icon, an executable, and a folder should be defined on each application.
There is an standard package format.
An icon on several sizes is needed.
Translation to several languages also.
Control panel themeWindow manager could change the theme. At least, clear theme, obscure theme, and high contrast.
LanguageEach text should be translated to the language selected by the user on the desktop.


Order by priority, or add other ideas. Draw an scheme for each one. And write a user manual.

OPEN SOURCE
This site is linux.org. So only Open Source GNU license GPLv3.
 
Thank you, this is very valuable. I will get on it. By “draw a scheme for each one” do you mean provide graphics for each one and decribe how I envision the behaviour?
 
Draw a window whit the graphic controls. With text commenting each type.
For instance, a window with tabs, with annotation about behavior. Explain what a star means in your design. Then a copy with annotations on the toolbar. Another with folder types ....

An scheme is a first drawn, even by pen. For design the user inteface program we need general situations (top, aligned to the left...) Later you can draw graphics properly (test sizes, test selection colors, accesible themes...) Then we include the graphics on the program prototype.
 
academianini If you were to help and participate in this, How finished would you like the design to be before I present something? Final design and behavior or start with behavior then move on to design in a more iterative approach?
 
Have you ever used RISC OS? The ROX desktop and filer tried to bring its features to Linux, but sadly seems no longer to be extant. Many years ago Gnome had a GUI in which each directory could contain a hidden script in which the user could define the bindings of possible user-actions (clicks and drags) within that directory, with priorities cascading down through subdirectories, as well as the attributes of window furniture. In RISC OS, for example, drag moves one filer-window over another, using the left mouse-button, and under another using the right. Having the input focus and being top of the window stack are not conflated, so you can input into a window even while you are reading from another window that obscures it. Click the close-window icon with the left button and the window closes, but with the right button the parent filer-window is opened as well. Scrolling with the right button scrolls in the opposite direction, so that you don't need to move your hand. I would recommend having a long look at the range of graphical user-actions that RISC OS provides, which far exceeds anything currently available in Linux, Windows, or OSX, and see how the default actions hang together ergonomically. No need to re-invent the wheel.
 
First, this website doesn't determine what license must be used.

Yeah, I didn't say anything but I chuckled when I read it. We're posting this very conversation with a closed source forum script. They aren't even remotely strict about FOSS for this site. (XenForo isn't even remotely FOSS, or anything close to GPL.)
 
No need to re-invent the wheel.

I agree. One (I) should do a thorough research of existing Window UIs, but in the case of RISC OS, I want to be able to run Blender and DaVinci Resolve, and I doubt they will be ported to RISC OS anytime soon? But perhaps ARM-based OSs will see a resurrection now that Apple is switching to ARM architecture? Not immediately related but I watched the AlphaGo documentary and noticed EVERY SINGLE ONE on the AlphaGo team were using Apple Macbooks, although AlphaGo itself ran on non-Apple hardware for sure. So Apple hardware no longer only attracts designers, filmmakers, musicians and advertising agencies anymore. At least not in the laptop space.
 
academianini If you were to help and participate in this, How finished would you like the design to be before I present something? Final design and behavior or start with behavior then move on to design in a more iterative approach?
First, I would need the User Interface new ideas. I assume, you compared existent file managers and your ideas are quite different. Because you ask for a development, instead of a plugin or a theme.

There are libraries like GTK, QT that include pieces like tabs, toolbar, scroll, buttons, menu, drawable area. But if you want tabs mixed with toolbars, and window without a frame, full-screen for tablets, touch screen instead of multi-button mouse, ... Then I would test if customized widgets are possible, or a low level (Xlib) library is needed.

Please, start with a drawing by hand, with a ball-pen. Scan the paper, and post it. I know you are a designer, and you would like perfect graphics. But for a team, I would recommend quicker approaches:
  1. General idea scheme
  2. Select development tools
  3. Scheme for each window, toolbar, page, sequence, case...
  4. Prototype program of the critical widget
  5. Main graphics in low resolution
  6. Application simplified prototype
  7. Behavior redesign
  8. Tools redesign
  9. Application prototype 2
  10. Test main graphics
  11. Debug
  12. Finish graphics for all resolutions
  13. Public binaries
  14. User testing
  15. Translate to 3 languages
  16. Alternative themes
  17. Accesibility
  18. User testing
  19. Public sources
  20. Branch because of design or development
  21. Compile on other systems.
 
First, this website doesn't determine what license must be used. Second, there are many (better in my opinion) alternatives to the GPL like the BSD 2 clause.
Ok. I'm not a lawer.

BSD, MIT, Creative Commons, Apache ... are alternative licenses on Linux. I understand some are permissive license. Their code could be included in privative software for free. If linux.org will accept all of them, thanks.

This project is a little one. I don't expect to obtain money. There are a lot of file manager, and the top list is hard.

I've curiosity about new ideas. I would work like a sport.

But I only work on a copyleft teams. If I work for free, then I want my code to be public profit. For instance, Java was open source, but now Oracle an Google are arguing about money in a court. Microsoft was on Java team, but .NET becomes an incompatible branch....

Because of COVID-19, my previous job discontinued. And I can not use the code developed. But I had a legal contract; I shell it; and I received money for my work. Also most of the source is customized, and unusable on other hardware.

Anyway, I was reading your proposal. The BSD 2 clauses looks permissive license. Not a copy left.

Thanks
 
I appreciate and respect your position and your feedback so thank you! I myself currently does not have an income, largely due to the pandemic. So I am torn, I would love to contribute to the FOSS/copyleft space, but at the same time I am in the process of selling my car to pay rent so the chance of a source of income is quite appealing.

Would you be open to a joint commercial endeavor where any profit is shared according to a beforehand mutually agreed fair distribution? If not, what should one expect to budget for a task like this if I could find a backer?

Or is a for profit release of a file/window manager running on top of a FOSS Linux-system just plain wrong and imoral? I’d appreciate your honest opinion.
 
Or is a for profit release of a file/window manager running on top of a FOSS Linux-system just plain wrong and imoral? I’d appreciate your honest opinion.
Why not give an option for users to donate? I feel that for profit is not wrong, but it's unlikely to gain many users.
 
If you have client, or users crowdfunding, perfect.

In order to compute a budget, I would need more details on requirements.

While searching for a job income, I will practice with open source project. At home, with an old computer, on free time.
 
Last edited:
I'm not a programmer, so I can't help you but if you let me to throw in a suggestion: make the new DE non-dependent on flipping - an ancient and stupid feature that confuses the DE functioning in many cases. Cinnamon is such a DE - whose functioning depends on flipping and that causes many problems, especially with games.
 


Top