Hats off for trying this in Linux. I gave up on that path a long time ago and am running nine monitors with a Windows 8.1 base OS, utilizing DisplayLink USB docs in combination with the DVI out on my main doc to drive the 2560x1600 30" display (the yellow one) with all others running at 1920x1200 from DisplayLink. I then run Linux in a VM, all from a laptop whith 32GB of ram and three built in SSD's giving me a bit more than 2TB of storage.
Interestingly there appears to be an eight monitor limit total when using DisplayLink for the OS so the ninth monitor is driven from its own DisplayLink doc which I've directly hardware associated with a VM. Of course because the host OS doesn't know about that monitor I also need a second mouse so I can access it. Here's the latest photo of my setup:
Honest question, do you feel a little ridiculous with that setup? I'm a coder and have 2x24" monitors, and found that to be borderline ridiculous with neck strain, thus the reason I'm going 27". But unless you're running security at a place of business or something with 50 security cams, I can't imagine what you're doing that requires a wall of monitors?
Our philosophy in the office is that alt-tab is for suckers. While my setup is the most over the top we have three other people with five monitors or more. We are all security researchers and find the configuration saves a bunch of time in reviews. As an example, I can associate a MitM proxy like the Burpsuite or Fiddler 2 with the server side application which might communicate to web clients as well as to additional web services behind the scenes. That takes up one monitor, typically the one at the far top left. Under that monitor I can then associate another MitM proxy with the client. I can then run the client from my laptop display. If I'm working on a fat client, on the 30" I'll then run Wireshark which will effectively be watching the client. On another monitor I can run sysinternal tools. What remains I use for writing code necessary for the review, running additional tools like Metasploit, e-mail, chat, and research. I arrange my workspace for the task at hand. On a daily basis do I use all nine? No.
Interestingly enough, with this sort of setup it's pretty easy to visually see what's happening when going after an application. Before going to the monitor extreme I'd constantly alt-tab between my monitoring & exploit tools with every action. Now I can run an action and see the results within one workspace. Of course there is a massive downside. It makes competing in CTFs a pain in the butt as I can't drag that setup with me to physical events.
The moment I use more than 1 screen, I find alt-tab to be cumbersome. I find that in a GUI environment, having your mouse autofocus on the window it's hovering over is far easier than using alt-tab. Especially when you have multiple windows open, cylcing to the correct window using alt-tab is usually slower than moving the mouse over the window.
> Our philosophy in the office is that alt-tab is for suckers.
Which is why I've partly switched from Unity to xmonad (a tiling window manager). Dual monitors is nice to have for me, but the real boon was to have 9 (or 10?) easily accessible workspaces, which allows me to only have to worry about a handful or less windows in each workspace. Unity also have workspaces, but they suck (at least out of the box).
I can't go back from three monitors (currently running 24" displays):
Left Portrait, Terminator with two horizontal splits
Centre Landscape, Vim with a vertical split, usually have NERDTree open
Right Portrait or landscape, browser and/or documentation
I find if I don't have all of those constantly open I can miss stuff, and having them all squished onto a single monitor means I'm not able to display enough information to cover my needs.
I just got my fourth monitor. Two of them are old shitty 900p screens though. My main displays are my new 144hz 24" 1080p panel ($160 black friday sale) a 21.5" 1080p panel over hdmi, and the two shitty old panels I got from clients throwing out their old screens.
Usually the layout is IDE on the main window, docs on the second screen, IRC, git, and github on the third / fourth.
Back in the day, I was much happier with two 4:3/5:4 monitors than I am with two 16:9 monitors, as you say the neck strain... or wasted space is a bit ridiculous. I prefer two squarish monitors to 2x widescreens _or_ giant screens.
If you're getting neck strain from using 2 monitors try moving them further away. Ideally you want both screens in your front facing FOV, with only your eyes doing the moving.
Aaaand this is why I switched to first a tiling WM, and then to OSX for its intuitive 4 finger swipe workspaces.
Till that switch I was a multi monitor junkie, but after 3 monitors it starts getting bad for your neck, bad for your eyes (with all those monitor backlights burning your retinas) and you also need humongous desk setups.
Nowadays it's just one big monitor and 4 finger swipes.
I work both on triple monitor and single monitor+workspace setups and still prefer the triple monitor for a single reason: you cannot put workspaces next to each other and get an overview of multiple windows at the same time..
Just saying it might work for you, but it's by no means a proper replacement for all of us
> you cannot put workspaces next to each other and get an overview of multiple windows at the same time
While I personally prefer a single-monitor no-workspace setup, there do exist several window managers that let you independently choose workspaces for different monitors rather than having each workspace specify a layout for all monitors. For instance, the window manager "awesome" uses tags (primarily 1-9) for windows, and lets you say "this monitor should display 1, and this one should show 2-3".
that's what I was thinking, I often throw a frame onto the left or right monitor.....having a single monitor would mean I'm adjusting positions all the time?
Synergy is amazing. It's hard to explain to people, but as soon as they get somebody to set it up you can watch their faces the first time they move their mouse over to another machine and they go "oh"..."ooOOOHHH" and the light clicks on.
They've recently moved to a paid model, but some older free versions are still around.
Video is great! Still it's weird, I can even show somebody a setup with Synergy running, and until the moment they move their own mouse cursor over that hard boundary at the end of their monitor, it never seems to click...and weirdly people are kind of resistant to it.
I almost think the next Tron movie needs a character named Synergy that bashes down arbitrary walls between systems to really get the idea out there ;)
I think the video looks great. It conveys the purpose very clearly for those who don't even know that they need something like that. Although it's not my cup of tea, I think the music is very current in this context. Nice work!
Hehehe. I once worked in an agency, where I commandeered three iMacs, and used Synergy (it may have been a similar application mind you) to use them as "monitors". Each iMac was for a different purpose :P
xdmx does slightly different thing, it allows one to build X display that uses another X displays as it's screens (everything goes through the xmdx server and last time I tried it it was incredibly slow, as in significantly slower than using normal X11 over the same network).
Wow. That is a pretty crazy setup. I would worry about neck problems with the position of the top row. How long have you worked with this setup? Also what make/model/spec is the laptop you run it all from?
It's taken a few years to build up to this. To mitigate neck pain I:
1) Have set my chair so that it rocks as far back as possible with little force.
2) Sit directly across from the gym and try to visit it twice a day for at least two 15 minute HIIT cardio / body weight workout. I do lots of pull-ups, archers, and have been working up to the iron cross.
3) Only use the upper deck monitors for security reviews where I need "eyes on" resulting actions. I don't use them on a daily basis.
Truly, if this works for you, then all the best. But we have lots of people at my organization programming, analyzing big data, and troubleshooting operations, and never have I ever heard or seen anyone wish for more than four monitors (two is the standard). I feel the physical issues, the problems with space (if everyone wanted that) and the capital investment all point to the use of workspaces.
This setup makes me think of this, which has been a joke in our circle for some time.
Not when you're trying to monitor a lot of data. Even on my dual monitor setup, One monitor has 4 terminal windows for informational purposes. It would take longer to move to another workspace than to just simply take a quick glance to the left.
Exactly! Over the years more than a few people have called me crazy for having a setup like this because at the end of the day they try to picture themselves using this sort of solution. The fact is, if you aren't trying to keep track of a large number of things as near real time as you can, nine monitors is absolutely absurd. In the past you'd only find this type of setup on a stock brokers desk; however, when you consider what's required from even a moderately scoped pen-test, going to this point makes a lot of sense from a time perspective. I've greatly reduced the amount of duplicate work because I can see the entire result of an action in real time while sparing myself from switching between tasks to work on code & exploits.
I'd go further and say don't berate someone's tools full stop.
They're not making you use it. They're not making you pay for it. Why do you care at all? What does it have to do with you? Why do you think they want your opinion at all?
It is utterly pointless and expensive to have an entire monitor dedicated to an application.
A basic monitor is a couple of hundred dollars. That's only a few hours work for a $100k employee - if that setup makes them more efficient, then it's easily worth the money.
Yeah, because you never, ever have to reference multiple documents/apps/consoles at the same time. Why haven't I thought of the solution of simply constantly switching a monitor to many different views 10-20 times a minute?
It's the same, except instead of turning your head to see the other monitor, have your fingers, which are presumably already on the keyboard, hit ctrl-alt-direction.
Just needs one really good monitor. Requires no special monitor mounts. Has worked this well for 15 solid years. 15 years! You could have been doing what you do now, but slightly better, 15 years ago.
So at the moment I am working on one screen while on the other I have my perforce instance syncing the whole repo - I want to start something as soon as that sync is done. Right now I can tell when it's done because the sync window is in my peripheral vision, on the other monitor. In your setup I would need to switch between workspaces every 30 seconds to see if it's done - how is that not giving you a headache?
Also - one monitor runs my application, while another feeds me the console output - switching workspaces to see both? Looks like a lot of hassle. And monitors are cheap - if I had more space on my desk I would be super happy to have 3+ monitors,but I have to work with only 3 at once.
I switch back and forth pretty quickly - the screens just flick left or right. Every 10-ish seconds, for just a second, if I care about something going on in the other workspace.
Switching is faster on linux than on other systems I expect.
This is a typical response from someone who doesn't understand what's necessary to manage workflows requiring the real-time monitoring and interpretation of complex data across system components. I'm more than familiar with multiple workspaces from my HP-UX CDE days in the early 90's to more than a decades use of LiteStep. You'll also note that next to the laptop driving all those monitors is a Retina MacBook pro in which I utilize multiple workspaces. Honestly, it's the only thing that allows me to do my job when not sitting in front of a multi-monitor environment, but it's much, much slower because of the need to swipe across workspaces and requires a fair amount of processing overhead in my mind to keep everything in context.
Alt-tabbery and workspace swapping are more likely to generate mental confusion and delay than having things simultaneously visible side by side, surely.
Interestingly there appears to be an eight monitor limit total when using DisplayLink for the OS so the ninth monitor is driven from its own DisplayLink doc which I've directly hardware associated with a VM. Of course because the host OS doesn't know about that monitor I also need a second mouse so I can access it. Here's the latest photo of my setup:
http://defaultstore.com/mydesk.jpg