1. 82

  2. 16

    The obvious solution is to mount the monitor on a servo that continuously rotates it in a full arc each day

    1. 4

      and attach an IMU that continuously updates the transformation matrix to make sure it is level at all times.

    2. 6

      I hope that is not your password on that sticky note

      1. 41

        I doubt it is. It’s hunter2 base64 encoded.

        ➜ echo "aHVudGVyMgo=" | base64 --decode
        1. 28

          I only see *******, does lobste.rs hide passwords if you put them in posts? That’s a neat feature. Here’s mine to test: *******.

          1. 8

            <DavidDiamond> Here’s mine to test: *******.
            thats what I see

            1. 3


              1. 3

                OH NO!

      2. 5

        45° is a good average monitor rotation: Just have one at 0° and one at 90° :)

        e.g. like this image of my setup, which allows me to have three times A4 in 1:1 next to each other, or two IDEs and a browser, etc.

        1. 5

          People often call me silly for sticking with my old 4:3 17” monitor. But then when I see the hoops people jump through (and yes I know, this link is being deliberately silly, but there’s a grain of truth in it) to try to make their huge extra-wide monitors actually usable for daily tasks, I’m thinking they’re the silly ones.

          1. 1

            Which one do you use?

            5 years ago when I moved from Boston to the Bay Area I left behind my 3x2 array of Dell 2007fp monitors (20.1”, 4:3, 1600x1200) because they were way too power consuming, falling apart, and heat generating, but I miss them every day.

            I would go back to that setup in a heartbeat but you just can’t buy anything with even roughly the same specs but modernized.

            1. 2

              I have three, all about the same. One is an acer v173 (that’s my first… my mother bought it for me for my birthday in like 2007. it recently refused to turn on so I had to swap it out but more on that later. I think it is a capacitor failure but I can’t figure out how to get the cover off to test/replace it! it feels like the plastic is fused together). One is a HP 1755 that I actually like kinda a lot too, it has both VGA and DVI inputs and can switch between them, so I use it as a secondary monitor.

              Then the third, which is replacing my Acer right now, is a Dell… E 173 FP. Interesting, both it and the Acer have 173 in the model. I wonder if that’s just coincidence or if it has more meaning. The Dell I got off a used trash heap and it does have some bright and dark spots in its screen but it is very minor and I still find it quite usable. (I’d love to fix my Acer some day though! Its screen is still working excellently and it is a bit darker than the Dell’s.)

              Anyway, they’re all 17” across and run at 1280x1024 which I find works very, very well for a desktop computer. It is detailed enough to have the space but not so big that I’m turning my head or breaking my wrist trying to fling the mouse across the screen. I do have that HP hooked up too, so if I want to watch a video or remote desktop to a work computer, I can do that over there while keeping the main screen ready for my own work. (And I have a custom xrandr setup to ease this - instead of putting them side by side, I put the two screens touching by corners on the virtual framebuffer. So the one screen is upper left and the other is lower right. This means the edges of the screens remain barriers keeping the mouse cursor corralled so it doesn’t get lost on the wrong monitor (I leave the second one turned off if not actively using it), yet it is also easy to deliberately move the mouse between the two by just aiming for the corner. Fitts’ law in action! Then my custom window manager always redirects full-screen windows to occupy only the second monitor, so the video player, etc. naturally lands over there keeping my main screen ready for using. I’m very happy with this setup and I’ve never seen anyone else do a thing similar.

              Anyway, my monitors all pull about 20 W when operating… the internet tells me your things eat 75 W… yikes. I can see how that’d add up fast and get hot. I can’t complain too much about my 20 W, though I am sometimes tempted to buy a new one to see if they are any more energy efficient…. just all the new ones come in these sizes and shapes I simply don’t care for. So I’ll keep recycling and repairing other people’s trash as long as I can.

            2. 1

              You just don’t fill the screen with one thing. Like a physical desktop, you put multiple things on it so you can look at them at the same time.

              The only thing wrong with those ultra-wide monitors is the resolution — the ones I’ve seen are all still stuck at around 100ppi. I refuse to stare at blurry text or graphics anymore.

            3. 3

              My favorite rotations are still 360 and 920. by far.

              1. 2

                Cute :)

                From a practical perspective, I’ve found the best use of lots of monitor space to be having more than one thing on the screen at a time; the long line length seems mostly irrelevant since (1) readability drops off way before you’d run out of screen space anyway, and (2) it’s far more useful to have docs or a second editor window/terminal or such open. You don’t need much screen real-estate for a single editor window; when I’ve found myself wanting tons of screen space it’s because I’m debugging something involving N services and M log files, and I want to be able to glance at info from each of these without having to dig for the window (and more importantly, for the motion to jump out at me if it’s from one that I’m not expecting).

                1. 1

                  I’m wondering how feasible it is to implement such a rotation with an FPGA. The advantage is that the hardware HDMI adapter will work for every OS and monitor, without posing any computational burden on the host :P

                  1. [Comment removed by author]

                    1. 1

                      I don’t think any of these are serious concerns. To address them in reverse:

                      The rotation is something that is trivial to parallelise and most of the arithmetic would work on fixed-function units on an FPGA. You can build very wide vector processors on an FPGA to the extent that DDR bandwidth becomes a bottleneck. At 3840 × 2160, you have 8294400 pixels. An FPGA doing this kind of thing can probably clock at around 200MHz, so you’d need to do 2.5 pixels per clock to handle a 60Hz display, maybe closer to 3 if you were rendering a larger area and clipping. I’d probably build this kind of thing to do blocks of 4×4 pixels at a time (fully pipelined, so even if it’s taking multiple cycles per block, it’s dispatching one block per cycle), at a minimum, so that’s quite plausible.

                      Adding a single-frame delay in output is unlikely to be noticeable. I’ve been playing games with Xbox cloud gaming this week and the RTT from me to my nearest Azure datacenter is 12ms (I don’t know if I’m actually using the nearest one for the cloud gaming, but I hope so) and that’s fast enough that I can play Halo 5 without noticing any lag. HDCP is probably adding more latency if it’s enabled.

                      Rendering a large area isn’t that much of a problem for most GPUs. It would be a problem for gaming, where you’re close to the maximum GPU throughput, but for normal desktop workloads your GPU is massively idle most of the time. The one in my laptop can drive two external monitors in addition to the panel and I only have one connected.

                      The biggest problem would be that sub-pixel AA would look awful. Font rendering makes a lot of assumptions about the topology of the underlying pixels. To get the same quality, you’d have to turn off sub-pixel and massively oversample (as in, render at least four, probably eight, times the number of pixels in each axis), then do the downsampling and AA on the FPGA. That might stress the GPU and would definitely stress the FPGA. If you do it with cooperation of the source renderer, then you can communicate the topology of the pixels to the font engine and start off with something useful.

                      1. [Comment removed by author]

                        1. 1

                          You don’t need to push all 5148x5148 pixels out of the HDMI port, just the ones you want to put on the screen.