I was reading @tbernard 's post about Rethinking Window Management with great interest, but I kept having this feeling that there was something missing in the picture…
Wide and ultrawide monitors, and multi-monitor setups, are quite common nowadays, and users can work comfortably with two or more windows tiled side by side. Then, we have several workspaces, also sitting side by side, and the user can jump from one workspace to another.
The metaphor of a workspace, as an space where the user works using a set of tools, works well when the workspace only a single maximized window. However, when the user has several tiled windows, perhaps spanning across multiple workspaces, then the metaphor does not work so well; I think we have a clash between the tiles and the workspaces.
- Instead of several fixed-size workspaces (all the same size of the viewport), we could have one single extensible workspace. New windows will be placed to the right or left of the existing windows, and the (single) workspace will expand to accommodate.
- Instead of jumping one workspace at a time, the viewport will jump according to the current contents: a full viewport, if there is a maximized widow, or half a viewport, if there are two tiled windows, etc.
The metaphor of a workspace, as an space where the user works using a set of tools
That metaphor doesn’t hold with the rethinking. At least to me this part seems to say that:
As you open more windows, the existing windows move aside to make room for the new ones. If a new window doesn’t fit […] it moves to its own workspace. If the window layout comes close to filling the screen, the windows are automatically tiled.
Typically I have 2 to 3 maximized windows and often additional 2 half tiled windows open on a workspace, to work on a task there. I work on multiple tasks throughout the day. Those windows wouldn’t fit tiled; I only have a 24" display. So with the rethinking this would spread a task out across 3 workspaces — breaking the metaphor.
The metaphor does not hold right now, and that is why I am proposing to rethink it.
No, my proposal would not change the number windows or it’s placement, only how you move from one window to another.
Perhaps this before/after image illustrates better what I mean:
That image does help No windows overlapping may work well if only a few windows are open or the display is bigger.
For me it would have the same problem: I only have a 24" display and typically work on 3-5 tasks throughout the day. With no windows overlapping this would give me a workspace 10-16 times the width of my display. I think that’s more unwieldy to navigate than stacked windows.
This takes me back. I can’t remember which environment it was but at least one of them let you have multiple workspaces larger than the screen. (FVWM, I think.) You could move your screen-sized viewport around a workspace and you could switch to another viewport on a different workspace.
If I’m not mistaken, the control room at my last job has this set-up. There are two 3x3 workspaces so the effective screen space is 18 times larger than the monitor. It wasn’t my system, so I didn’t dare touch it.
Kudos on the illustration. That seems a lot clearer than I remember mine being. I tried to do it in 3d perspective to show all the layers on the screen, from the desktop at the bottom to OSD “in the glass”.
I am not advocating for non-over lapping windows (even if I did not show them on my mock-up); I am just advocating for a large non-partitioned workspace (and a window manager that can glide along it).
Found it. It was a long time ago. I only had 800x600 on a 10.4" screen. I don’t know why it’s a JPEG either.
And I didn’t remember correctly. It shows a “sticky” panel, stuck under the glass.
This topic was automatically closed 45 days after the last reply. New replies are no longer allowed.