Multi-Monitor Support Forces More Use from Windows Partition

I've been struggling with multi-monitor support for a while now. I'm still not 100% satisfied, but I think I've got the best solution possible for my circumstances. I should probably share my system specs before I go any further..

CPU: Intel Core 2 Duo E8400 3 GHz
2 Nvidia GeForce 9800 512MB video cards
4 GB DDR2 RAM

This was what I wanted: SLI configuration, compiz effects, both monitors spanning 1 desktop (to drag windows from one monitor to the other), separate virtual workspaces for each monitor including different wallpaper.

The end result of much effort and several attempts is:

SLI simply does not work with a multi-monitor configuration at all. At least in any configuration I tried the 2nd monitor simply would not start in SLI. So if I wanted to use both cards, Twin View is not an option.

Separate X screen without Xinerama works but it was unusable for me as it caused some really strange things to happen. Nautilus windows just started popping up in the taskmanager and would not stop launching them until the system crashed. Until the crash, the two monitors worked however and compiz was enabled as well. With the separate X screens though, dragging windows from one screen to another was of course not possible.

I tried enabling Xinerama with the separate X screens, and this worked fine but for two things:

1. No compiz effects. This was disappointing, but I could live with it.
2. No KDE apps. Indeed, trying to start up any KDE application would cause the entire desktop environment to crash to the login screen. If I tried to log into the KDE desktop environment rather than Gnome, it would load and crash, again leaving me at the login screen. This I cannot live with as there are a few KDE programs that I just cannot live without.

That left with me with Twin View, which does work very well, except for 2 things:

1. Separate virtual workspaces for each monitor are not possible. Xorg sees only 1 desktop , which means that each monitor must use the same wallpaper and switching virtual workspaces will cause both monitors to make the switch. It cannot be done independently.
2. My second video card is now not being used. Now I would like to point out that SLI never worked very well anyway. Or I should say that it never worked very well when compiz was enabled. I actually switched to KDE briefly for this reason (I could have the plasma desktop effects enabled and SLI worked fine). I'm now really regretting going the multi-GPU route when I had this machine built. I wish I had just purchased one 1024MB card, but c'est la vie.

This current setup suits me fine for the most part. For graphically-intensive gaming however, I will have to rely on my Windows partition a little more than I do currently which is a real shame. I would like to point out that I believe blame can be laid squarely on Nvidia for lousy SLI support in Linux. In Windows I can have a multi-monitor setup with SLI enabled with no problems whatsoever. I know this has to be possible in Linux as well, but Nvidia just can't be bothered working it into the drivers. And they refuse to release the source code so that others may do the job for them.

I'll admit that compiz is not innocent here. But even with compiz disabled there is some discernible stuttering in the display when SLI is enabled.

This brings me to another point. when something quirky happens in Linux, it doesn't mean that Linux is quirky. It means that some software running is having some undesired results. To put this in perspective, lets look at some software that runs on Windows that has less than desired effects. Perhaps the most common that I see is umpteen-thousand toolbars that people for whatever reason install in Internet Explorer. It has the result of not only slowing down the browser, but with all the spyware that inevitably comes with it, the whole system slows to a crawl (and why the hell are you using IE anyway?)

I could bring up more examples obviously, but you get the point. In either of these cases, the OS is not to blame. It is either the software being run, or a configuration problem (well it's both in each case. The Windows example would include the configuration issue because if a guest account was used instead of the default administrator account, the spyware wouldn't be permitted to install).

But I digress. I didn't really mean to turn this into a Linux/Windows comparison. This was only meant to be a more personal post of my own experience.

If you have any experience with the issue I've described, please comment and if you have a solution I may have overlooked, please share it.

The Year of Linux (Take 19)

The general consensus (according to w3counter.com, which gets its data from web-usage only) the current Linux market share for the Linux OS is approximately 1.5% (as of October 2010). Yes there are other stats that give different numbers, but this seems like an approximate average. Some have Linux's share at below 1% and more than a couple have it at above 5%. The specific numbers don't even matter to me at this point. I only think that they should be better.

Now of course I've also heard the arguments that globally the percentage could be as high as 40%, taking into consideration 3 points:

1. Countries outside of Europe and North America are not considered in most if not all data collecting websites.
2. The population of China and India together roughly doubles the population of Canada, the United States and all of Europe combined.
3. The governments of both China and India actively promote the use of GNU/Linux over Microsoft and Apple OSes.

I only point to this as a way to discourage the mindless Linux hate coming (mostly) from Microsoft users. I fail to understand where this hate comes from, especially considering that it mostly comes from people who haven't given Linux a chance at all, or from those who tried it once over a weekend. But I digress.

My real question here doesn't have anything to do with Linux popularity in various Asian countries. My question is what would it take for the market share in Linux to rise above, say, Mac OSX's desktop market share?

I think I have an answer, but it will take a bit of explanation.

Linux is the No-Name Brand OS

At least to Windows and Mac users it is. Most people consider Linux to be the cheaper and less-usable alternative to the "name-brand" Windows and Mac. This, of course, couldn't be farther from the truth but the only way to change people's minds is to actually get them to use it for more than a day. And I don't think that this can happen without a cultural change.

Consider the time-frame that Microsoft came to dominate the market.

In the early '90s, Microsoft was still the young, upstart software company that was starting to do battle with the big, evil corporate Apple. Windows 3.0 was released in May of 1990. (3.11, the big one before Windows 95, was released in 1993).

Anyone remember what else was going on at the time? I mean outside of the computer world. Culturally speaking. Anyone?

There was Seattle, the whole "Grunge Movement", and everyone in any advertising agency freaking out because no one was listening to them. Fashion magazines had no idea what to do and were clothing models in plaid and knee high combat boots, selling outfits for $50 and less. No-name brand products, for the first time, were outselling name-brand products. People were about function, not style. And people loved to support the underdog. Western culture, in these three or four years of the early '90s, was saturated with these ideologies.

People also started buying computers a lot more than they had ever before. Technology started to be more in the front of people's minds in the mid '90's. I remember "multimedia" as the buzzword, but that was quickly replaced as the media started to become aware of this mysterious thing that hackers were now doing: "surfing the web". And the Internet began to be seen as a re-emergence of the Wild West, at least that is how it was portrayed.

Enter Windows 95, the latest offering from that young upstart software company that ran perfectly well on an IBM clone -- the no-name brand computer. The timing was really perfect. Microsoft was well on its way to dominance, but already the function-not-style mindset was dissipating. Grunge was dead. And musical tastes were changing too. "Electronica" was now becoming a movement. Raves became popular again. And this edgy sounding music was all technology-based. Kids could make this sort of music in their basements if their parents had bought them a computer at some point.

Windows 98 was released in June of 1998. Advertising agencies were breathing a sigh of relief, and starting to relax a bit. The recession was over, and the US budget was balanced. People had disposable incomes again. The depressing Music of the early '90s was gone. The Spice Girls, The Backstreet Boys and Brittney Spears were at the top of the charts. The Smashing Pumpkins disbanded in disgust. And Microsoft was king. By the time XP was released in 2001, America was thoroughly back in the name-brand, style before function mindset. As long as something looked pretty, it didn't matter if the core was rotten. If it was expensive, it must be good.

Microsoft got really lucky. Windows had matured in a way that completely fit into the rest of America's culture. Don't get me wrong, Gate's business sense had much to do with Microsoft's success too, I'm just saying that the timing of it all expedited its popularity.

Considering this, where are we left? Well Microsoft is still undeniably king. No longer are they "the young, upstart software company that was starting to do battle with the big, evil corporate Apple". They are now the big evil corporation that people are going to Apple to escape from. (Fill this space with any ironic comment you wish).

But we are in a recession now. Why are people paying even more for a Mac, rather than switching to Linux which is cost-effective and ultimately better than both? I would say the reason is that the culture didn't change with the economic situation like it usually does. The reasons for that are many and varied. Not the least of which is the anti-sharing media campaign put forth by the RIAA, MPAA and other groups. (And hey, telling people not to share is telling people not to use free or open-source software).

So to finally bring this around and include the reason for the title, will the year of Linux ever come? Probably. But not any time soon. If 9/11, two wars, the worst recession in decades and Michael Moore can't seem to make a significant cultural change, then I just fail to see how it's going to happen. And without it, I can't see people abandoning their OS of choice en masse.

Unless of course it's Linux that actually causes this cultural shift I'm waiting for... hmmm...

Richard Stallman is Right

I've been "dabbling" in Linux for a number of years now, but I've really only given it a serious try in the last few of months. During that time I've become very interested in the Open Source and Free Software philosophies. I've watched documentaries on the history of Linux and the FSF, and videos of Richard Stallman's lectures.

Let me say first that in the history videos I've seen, Stallman really does come off as the freedom fighter that he seems to see himself. I really admire what he's done, particularly in running with his ideas and finally coming up with the GNU GPL. His more recent lectures, however, leave a sour taste in many people's mouths, including mine. He seems more like a stubborn zealot than a freedom fighter. To demonstrate this, let's have a look at the Free Software Foundation's website, specifically the list of approved free OSes, and why some are not included.

My current OS of choice, Ubuntu, would definitely not make it in the list. Ubuntu offers me the choice right at the point of install to use proprietary codecs and drivers. Some included repos have proprietary software that I could install using the included package manager. Indeed, on my personal system I am very conscious of this. I am using the Adobe Flash plugin for my web browser, and I am using the proprietary Nvidia drivers. I have Skype installed. All of these things would exclude me from being a part of the FSF community.

So what if I wanted my system to be 100% free? I've some experience with Debian, I have a server here running LMDE. So I could switch everything to stock Debian, a distro that many, many other distros start from. Well even Debian is not on that list. And the reason is that there is actually proprietary code within the stock Linux kernel. Proprietary kernel drivers to be more precise. It also offers proprietary software through its repos.

This in of itself, that the FSF condemns Debian as a non-free OS, is enough to turn most people away. But there is another side to this coin. There are real, inherent dangers to proprietary software, and the FSF is really only trying to avoid those dangers, and warn others of those dangers at the same time. The more proprietary software that the kernel depends upon, the bigger the danger that it will at some point have to go backwards to replace those proprietary blobs in order to remain free (as in beer). There is always the chance that those blobs will some day come only with a price tag. Or cease and desist letters. If that day eventually comes, those blobs may very well have to be replaced using code from the FSF (accompanied by a very smug [and justified] "I told you so").

Some may say that this is overstating things a bit, and in the kernel's present state that may be. But the list of proprietary blobs occupying space in the kernel seems to grow with every release. Will the day come that Linux in its official state will have to be considered proprietary? Its present growth certainly indicates that as a possibility. Or at least that the proprietary code will become predominant over the free code. And all that proprietary code is owned by someone. As Linux grows in popularity (it does experience exponential growth every year), do you really think that at no point in time any of the companies that own that code are going to start withholding it pending licensing and payment? It's simply naive to believe that. That cost will then have to filter down to the user in order for Linux to continue in development. And the biggest side-effect would be (is?) that as Linux goes down the commercial path, it becomes more and more like the restrictive, closed environments that all of us, developers and users alike, wanted to avoid in the first place.

It's either that or Linux as a desktop will have to take giant leaps backward in functionality (and all the bad press that comes with it).

Thinking in this direction, I begin to see Stallman's point. This could all be avoided if the "narrow winding path" was chosen and all proprietary code was rejected in the first place. It would be a slower process, sure. Many things that can be done on a Linux desktop now wouldn't be possible--yet. All that proprietary code would have to be rewritten from the ground-up, reinventing the wheel as it were. But there would never be the worry of anyone owning the system. It's owned by us all.

This makes me thankful for the Free Software Foundation. In avoiding the "wide and easy path" they are ensuring the future of the Free Software philosophy. They are ensuring that computers will in some fashion always be affordable to virtually everyone.

So will I be switching to an FSF approved OS? At some point I think I'll have to, but that will not be anytime soon. I'm still dual-booting Windows. But when that day comes, boy will I be grateful that Richard Stallman stuck to his guns. He may be a thorn in the side of the Open Source movement right now, but he certainly has his place.

I think the Open Source movement needs to be reminded of its roots now and again, it needs people like Stallman to point out that this direction goes against the ideals that prompted the creation of GNU/Linux in the first place. It may not be nice to hear at times, but sometimes the truth hurts.