Poor man’s KVM can be done much better than that, in fact can be done so well that it’s arguably better than a hardware KVM in a number of ways.

—Aristotle Pagaltzis

First, to respond to Aristotle (this is getting to be a habit), Synergy was the first thing that popped into my head after reading Chris' post. Of course, it admits to being a KM (instead of KVM) solution. Another issue is that in some setups, you may need to login to a client machine before the Synergy connection is established, which would require a separate keyboard, if briefly. I think it would be good if Synergy had some way of allowing the server to authenticate and initiate the client remotely. (I suppose I could do this with SSH; I wonder if anything more is really needed.) Those are probably some of the reasons why you used the "arguably" qualifier, though. When you don't have an extra monitor, Chris' "poor-man's KVM" is a reasonable solution. I only really want one monitor for myself, so I use one of those nice Dells with multiple inputs and use it with multiple computers as well as console gaming systems, and it serves as the only standalone display device in my apartment, where space is at a premium. I then combine it with Synergy to hook in my laptop. Fun fun.

After reading Chris' previous post on this general topic, I spent some time reflecting on why I use the various interface organization paradigms. There is certainly a benefit to being able to simultaneously obtain information from multiple interfaces if work done in one is dependent on the others (copy and paste, drag and drop being WIMP examples). This is sometimes the case, but I find myself more often focusing on one interface to the exclusion of others. (Or do I?) In those cases, why should the others be available (at that time)? For example, if I'm reading one web page, I don't need to see the other web pages I have open at the same time, so they are usefully organized into tabs (or virtual desktops, or both). How can we best organize our visual (and other?) input streams to accomplish particular goals? Some other tasks in this space are finding and transitioning from one interface to another; being notified of updates in non-focused interfaces; and providing structure, organization, and context for work, but I'm also curious to know if there is some sort of systematic characterization of all the issues involved here.

   - John L. Clark

For me, the 'poor man's KVM' solution is for situations where you rarely use your secondary machines, so you want your primary machine to have all of the monitors almost all of the time. I think that this is a different usage scenario from Synergy; Synergy seems more aimed at when you're regularly using multiple machines but don't want to use multiple keyboards and mice.

(The poor man's KVM also works best if you don't try to use your primary machine and a secondary machine at the same time, because you have to remember that half your primary machine's display is invisible.)

Over the last few years my company has embraced the idea that giving people the tools they need to do their work is a good thing! Faster desktops, more screens, more disk space, more bandwidth.

Currently I now have dual 20" displays on my work desktop, some developers have triple headed systems and there are now plans to give some of our NOC type people 8x30" monitors to stare at. They currently have 8x20" displays as a minimum.

While I do find KVM type virtual desktops useful for remote operations, nothing beats good old screen real-estate.

Chris Gregors

I too have been in love with multiple monitors for many years. This should be empirically tested, but the productivity does seem to increase with spatial place. (Classical example for me would be a side-monitor with either some .pdf to read while coding or .dvi to use with LaTeX.) Yet, I have also noticed that too much can be exhausting: human mind has very limited capability for multitasking. Once you have, say, three monitors and various things blinking, it becomes hard to concentrate on any particular window.

Cheers,

j.

I tend not to multitask very much even with multiple monitors; instead I use them to spread out a lot of stuff for a single task, and sometimes to preserve a clear space to put windows needed for interruptions.