Of course, they suck for completely different reasons, which is part of the frustration. And each have places where they shine. Why can’t there be a nice combination of all three? That would be perfect.

Why Mac OS X sucks:

• Apple is becoming a controlling, closed system, and with the advent of the Apple App store, developers have to go through an approval process to get their apps on the store, there are specific things you can’t include in an app in the store, and there will come a time when most people get their software through the store, so there will be less and less incentive to maintain non-app store versions of software apps
• These days, you can find most kinds of software for the Mac, but there still is a relative paucity of apps in comparison to Windows.

Why Windows sucks:

• Viruses, Trojans and Worms, Oh My!
• Although I have only seen the Blue Screen of Death once in my year of Windows 7 use, there are still inexplicable slow-downs, crashes, and weird problems. And it takes FOREVER to boot, even with Soluto.
• Internet Explorer

Why Linux (in my case Ubuntu) sucks:

• I have to go through arcane (and luckily for me, fairly painless) procedures to get simple things to work (like plugging a headset with a mic into my jack!)
• Hardware manufacturers ignore Linux for the most part
• Most software developers don’t make Linux versions

The only good news I can see is that the operating system is getting less and less relevant. And, on balance, for me, Linux is winning. Now that dropbox and scrivener work on Linux, and I’m moving from Quicken to some online cloudish thing (suggestions?), I can pretty much leave Windows behind. (Oh, there is still Netflix. Sigh.)

There are a number of ways to do this. We started using S3fs as a way to mount an S3 bucket in the filesystem, then just copy the files to S3. One of the scripts we’ve use is here. (We also use rsync.) However, S3fs isn’t very actively supported or in development. So we’re thinking of moving to use S3cmd, which works really well, and is still under active development.

I got involved in Linux a long time ago. I was a professor at the time, and a nonprofit organization wanted to get on the web, and give some of their staff email, and at the time, colleges and universities were the only organizations that had easy access to the internet, and virtual hosting companies cost a fortune, way beyond what a nonprofit could afford. The date was sometime in 1995. We set up a little box in the corner of my office, and loaded several piles of floppies containing the Slackware distribution onto this box. After a few hours (as opposed to the few minutes it would take now) we configured that server to hold a website and serve email. The old site is still up on the Wayback Machine. I co-administered that box for a few years. Eventually, they got a T1, and moved the server in-house. I left academia to do that sort of thing with nonprofits full time. In fact, that experience, and the work I did around it with that organization, was the first step into this whole nonprofit technology field.

What I learned about Linux back then was that it was a way (along with the help of a college) for a nonprofit organization to get on the web easily and relatively inexpensively. It leveled the playing field, so that an organization without many resources could do what at that time, required a lot of resources. In many ways for me, the most important aspect of free and open source software is that it does just that – it levels the playing field so that people and organizations with few resources can have access to quality tools to do what it is they need and want to do in this software-driven world.

I’ve learned a lot about FOSS since then, of course, and the other aspects of FOSS have also come to be very important to me. I do agree, fundamentally, with the four freedoms laid out by the Free Software Foundation:

• The freedom to run the program, for any purpose (freedom 0).
• The freedom to study how the program works, and adapt it to your needs (freedom 1). Access to the source code is a precondition for this.
• The freedom to redistribute copies so you can help your neighbor (freedom 2).
• The freedom to improve the program, and release your improvements to the public, so that the whole community benefits (freedom 3). Access to the source code is a precondition for this.

And ultimately, in Michelle’s perfect world, all software, all content, all hardware, etc. would all be free (libre). But we don’t live in Michelle’s perfect world, we live in this world. This broken, very imperfect, very problematic, and quite capitalist world.  And in that world, I am a realist.

I am ecstatic, and do many happy dances a day, that there are people who write and support free software. I think of myself as one of them (Besides working with NOSI, I’ve been involved in several projects over the years in varied capacities.) The number of situations where one can still argue on a functional and cost level that proprietary software is a better bet get fewer and fewer. I could easily argue that for the overwhelming majority of places you need an operating system, the free and open source alternatives are better (if you count the BSD core of the Mac OS. If you don’t, it’s still the majority.) There are innumerable really great free and open source desktop applications that can run on any OS, and there are more every day.

And, surprisingly to me, I’m quite happy that lots of big corporations are now really getting into free and open source software support. I think, ultimately, it’s when big corporations want to ditch Windows on the desktop that the biggest strides will be made in Linux desktop usability and support. That’s a tide that will really lift all the boats.

Bottom line for me: free and open source software is about leveling the playing field, free access, community benefit, and community control allows this. That’s why I got into it in the beginning, and that’s why I’m sticking around, and doing what I can, even though I’ll be using my Mac (with, of course, a lot of free software applications on top.)

And, guess what? I’m giving up. At some point, when I’ve saved up enough pennies, I’m going to buy a Mac laptop again. I’ve basically switched to using my Mac mini for just about everything except the bit of systems admin and coding I do, because it’s just so much easier to set up things on Linux for that type of work.

What happened was I felt like I was wasting too much time on things that should be easy. It should be easy to plug in a new monitor. It should be easy to get wireless, it should be easy to add a new printer. It should be easy to play a DVD. These things are far from the fault of Linux. On the Windows side the hardware manufacturers make proprietary drivers for Windows, and very few make drivers for Linux, or open source their drivers so that Linux developers can use them. On the Mac side, Apple controls the hardware, so there never is a problem with it. (And, of course, there are plenty of peripherals that don’t work with Macs.) And then there are the proprietary codecs and DRM, that are all tied to an OS.

So what does this mean for Linux on the desktop? This experience has made me think a lot about where organizations should think about using Linux, and where they should steer clear.

My theory is that where Linux is going really work is in dedicated devices built from the ground up to run Linux, and used for relatively limited purposes. The eePC is a great example, as are cell phones, media players, etc. This is where Linux will shine. And, of course, there are some other situations where Linux also shines: kiosks, internet cafes, computer labs and email/web workstations. There isn’t good reason, at this point, *not* to use Linux there. I think it can also work for the folks who perhaps use laptops as their primary machines, and don’t do anything except email and web. And, of course, always, for developers. I probably will always have a Linux desktop around, even if it ends up being a virtual machine, for the varied (small, at this point) development projects I have going. It’s just dead easy to use Linux for development – easier than using the Mac, even though it has BSD as it’s basis.

Where is Linux not a good idea? Most creatives and knowledge workers who are not developers. There are just too many things we need – too many cool new peripherals, we want to manipulate too many kinds of data and media, etc. And we don’t (at least I don’t) want to spend to much time getting all of that to work. And there still are lots of newer software and services that aren’t coming out in Linux versions (for instance there does not exist a decent usable twitter client for Linux – gwitter is not very usable at all, and the others just don’t work.) And, of course, designers need software that just don’t have high enough quality open source alternatives yet. I think that for a while, at least, Linux won’t be a good desktop option this broad group of people. Which is unfortunate, and I hope it changes. Linux has made huge usability strides in the past few years, as has open source software in general, so I think the future is still bright.

Today, someone on the progressive exchange list asked about a tool called Rapleaf. A story about Rapleaf in Clickz (a newsletter for online marketers) says this:

Rapleaf allows you to quickly and inexpensively find out the social networking footprint of those you’re marketing to. Just send the company your e-mail list and tell it what social networking sites those on your list are using, their demographics, the numbers of friends they have, how many widgets they’re using, even their interests. Rapleaf digs into the usual social networking sites (Facebook, MySpace, etc.), as well as newsgroups, commerce sites (like Amazon), review sites, forums, and news groups, and even searches the general Web to find out where your people are and what they’re doing online.

An interesting conversation ensued on the list – with some arguing that this was a problematic thing. I actually thought this could be quite useful for organizations to figure out how to allocate sparse resources in the Web 2.0 space. But that’s not the point of this post.

I realized that one of the most important things that we can do is educate the organizations we work with (as well as individuals) about privacy issues and data. When is data public? When is it private? How do we know? How can we assure privacy?

It is important to understand that Rapleaf is just gathering public information on people, based on their email addresses. It is an inevitable result of our desire for social networks, as well as our desire for information to be portable (like in RSS feeds.) What’s important is that we understand what is actually public, and what isn’t, and how to keep what we want to be private, really private.

Linux on the desktop has come quite far, in just a few years. And recently, there is increasing evidence to suggest that it can, indeed in large part, replace Windows on the desktop. Why should it? Linux is more secure, more stable, and can be used on older hardware.

Walmart was selling $200 PCs running gOS (no, that doesn’t stand for googleOS, but greenOS, based on Ubuntu 7.10,) and they sold out. If you read the reviews (most of which were quite positive,) the people who liked it were looking at the real functionality (it could edit their documents, it could surf the web, read email, etc.) and those who didn’t, seemed not to like it mostly because it doesn’t run Windows (although one could install Windows on it – but it’s going to be pokey – it’s not a well powered machine, but more than enough for Linux.)

So, if students and Grandma can use Linux, can nonprofits?

There is a good whitepaper that was released this fall from Novell, which has a section which talks about what to think about with enterprise migration to Linux on the desktop. It basically echos what I would suggest when thinking about a mass migration:

• Planning is key
• Do a software inventory – figure out:
  • What has a version that runs on Linux
  • What can be replaced by software that runs on Linux
  • What can be run in an emulator such as VMWare
• Identifying types of users (by what they need to do)
• Choose a distribution that makes sense (I wrote up a review of Linux distros recently.)
• Figuring out a clear migration strategy that takes all of this into consideration.

It wasn’t so long after that that virtual hosting companies became ubiquitous, and affordable for nonprofits. But it’s only been in the last few years that mega storage, and mega processing power were available to organizations to power big web applications and the like.

Amazon seems to be leading in the next wave of evolution of hosting – pay only for what you need, when you need it. They started out with their S3 – simple storage service. And now, there is E2 – Elastic Compute Cloud – use only the storage, processing power and bandwidth you use. I did a quick calculation of what my own usage might be, and actually, my Dreamhost account is a better deal. But for much larger/high-traffic sites, or sites that fluctuate a lot, it might be a great idea, especially if you want dedicated hosting.

The news today, and why I’m bothering to talk about this, is that Red Hat announced that it will offer RHEL – their enterprise distribution, on Amazon E2. If a nonprofit organization has a server, it’s actually not so unlikely that it is running RHEL. A lot of organizations of all types want support, and are willing to pay for it, and Red Hat is, at this point, built the best business model around this than any other distro (Canonical, with Ubuntu, is sneaking up behind, but I’m not sure it has the “enterprise” style some people look for.)

So running RHEL on Amazon E2 is a potentially low-cost, low-pain way for nonprofits (with appropriate levels of tech staff, of course) to dip their toes into hosting complex applications on Linux, without having that noisy box in the corner.

Gutsy Gibbon is certainly easier to install and set up than Windows Vista, and it’s very close to matching Mac OS X when it comes to making things “just work” out of the box. Wi-Fi, printing, my digital camera and even my iPod all worked immediately after installation — no drivers or other software required.

I’m in the middle of moving, otherwise, I’d be checking it out immediately. Once I get settled in a couple of weeks, I’ll be giving it a spin, for sure. It seems that with Ubuntu, Linux is getting closer and closer to being a completely viable and usable desktop for everyone.

Well, all of that is true. But in this newer era, when we think that cost is more than just dollars, but we also have to think about the hidden cost of all of those toxic chemicals, fossil fuels, and water that goes into manufacturing computers, as well as the bulging landfills all over the country.

So I’ve been thinking a lot about the role of free and open source software in environmentally sustainable computing. Some of the biggest reasons to replace computers is not as much the hardware failure issues, but software bloat and cruft, and planned obsolescence. You can’t run Microsoft’s Vista on a computer that is more than a year or so old. Computers that now run Windows 2000 (there are plenty of them in nonprofit offices, I’m sure) probably can’t even run XP, let alone Vista. But computers of that vintage can pretty happily run Ubuntu Feisty (the current Ubuntu version). And older computers running Linux make very handy single (or even multiple) purpose servers – file servers, backup servers, dhcp servers, routers and firewalls, print servers, etc.

The great thing is that converting to Linux won’t just help environmentally – Linux has lower IT costs – lower support costs, and no software acquisition costs. If an office converted from Windows to Linux, they could keep their hardware much longer (five years easily), and have much lower IT costs, thus, in the end, creating a more environmentally and economically sustainable office.

Of course, there are caviats. There are software options that don’t exist yet, there are hardware incompatibilities, but these decrease every year. Sometime in September, NOSI will be releasing the new version of the primer on open source software for nonprofits. I’ll be announcing it, for sure. There will be some much more detailed information that will help make it easier to figure out if it is right for you.

At least in Massachusetts, the majority of nonprofits work with network support people who don’t focus totally on nonprofit organizations (there are some wonderful exceptions, however, of companies that focus on the sector.) The good thing is that since the business world seems to be moving ahead much more quickly on Linux and FOSS adoption, companies that work in both the for-profit and nonprofit sectors are gaining Linux expertise – expertise that nonprofit organizations can benefit from. But I think more is needed.

One of the interesting dynamics of any technology provider of any stripe is the way they choose the technology they will focus on and support. I think this is something that many nonprofits, especially those without in-house technology expertise aren’t that aware of. No technology vendor, even the largest ones, can support everything. Most support only a subset (sometimes a very small subset) of the available options. This is because all providers start out with some personal experiences or biases, and most are too busy taking care of clients to spend lots and lots of time gaining new expertise in a broad range of topics – they need to focus. And even if they hire expertise, there has to be enough overlap for synergy to happen. Most of the time, for clients, this doesn’t matter. Sometimes it does matter, both to an organization that might use that vendor, and to the vendor themselves in terms of viability of their business model.

I came across this discussion of Linux distributions and their strengths and weaknesses in terms of vendors who might resell Linux. It was interesting not as much for its focus on the presence or absence of “channel programs” but for the way it characterized the qualities of the different distributions in terms of the business model of reselling Linux. In a sense, of course, if you are a technology provider and you install Linux on some servers or desktops in an organization, you are “reselling” Linux. But since most Linux distributions are free (as in beer), that’s not really quite the way to look at it.

So I thought I would take from their model, and instead, talk about distributions from the perspective of the needs of a small to medium-sized technology support organization (for profit or nonprofit) that serve nonprofits. I’m really interested in helping technology providers get up to speed, so that the amount of support available for nonprofits using Linux (and open source in general) increases. If you are a provider, please feel free to email me if you want more info or help and support in moving forward.

Ubuntu

As this article states, Ubuntu is a very popular distribution right now, and Canonical is working hard to get Ubuntu in as wide a range of hands as possible. The basic philosophy of Ubuntu “Linux for human beings” is certainly one that makes sense for a lot of nonprofits, and it also has made Ubuntu the easiest distribution to set up and use, especially on the desktop.

Ubuntu also has also focused a lot of effort on building community, and has, hands down, the most vibrant, helpful and deep community of any Linux distribution. They have mentorship programs, they are building regional networks, they have almost unparalleled bulletin board, email and IRC community support. This community is one of Ubuntu’s great strengths, in terms of the ability to find helpful and sometimes instant support (via an incredibly active set of IRC channels.) And, Canonical also provides professional support. Canonical has a lot of connections in the nonprofit sector.

In my opinion, it’s a good distribution to start with if you are just beginning to learn Linux, and thinking about adding it to your business because of the vibrant and deep community that is there to provide support. But, as the article linked above says, they don’t have an official affiliate program, so it will take some shoe leather on your part to build the business aspects.

Red Hat

Red Hat is the old standby, and is in the server rooms of many nonprofits all over the country. The old adage “no one ever got fired for buying IBM” which, of course morphed into “no one ever got fired for buying Microsoft” in the Linux world might be “no one ever got fired for implementing Red Hat.”

Red Hat has a well-built business of providing enterprise level support for its distribution. It was the first Linux distribution to make it big in the business world. Its focus is on servers, and Fedora, it’s “community version” does certainly benefit from Red Hat’s development expertise and resources, but it doesn’t have as vibrant a community as Ubuntu, for instance.

However, because it is so common, and there is reasonable community support, and because of the strength of Red Hat, it might be a good choice, especially if a technology support organization works with larger organizations.

Debian

I’m a real fan of Debian, and have been using it on the server-side for a very long time. It is a rock-solid distribution with what is arguably the best package management system. (Ubuntu is based on Debian). It has a vibrant user community. Debian is the favorite of most serious Linux geeks. The Debian community is dogmatic in their approach to licensing – nothing in Debian depends on software that is not truly free (as in ‘libre’.)

Debian used to be one of the most difficult distributions to install, but that is no longer the case, so it is definitely a distribution that you could try as a Linux beginner (although beware that the community isn’t going to be as friendly to newcomers as the Ubuntu community.)

I think it’s certainly a possible choice, especially if you’ve got a philosophical approach that’s resonant with Debian (a lot of activist technology organizations use Debian) and have (or can find) the expertise needed. But it is a bit less known and popular, especially with the presence of Ubuntu, which has most of the strengths (in a technical sense,) and none of the weaknesses (in a business sense) of Debian.

CentOS

I’m only including this in my review because I have come across a surprising number of technology providers that have chosen to focus on CentOS. I had not heard of it until then. CentOS is a bit of an odd beast. It is Red Hat, with branding removed. In their words: “CentOS is an Enterprise-class Linux Distribution derived from sources freely provided to the public by a prominent North American Enterprise Linux vendor. CentOS conforms fully with the upstream vendors redistribution policy and aims to be 100% binary compatible.”

So they take RHEL (Red Hat Enterprise Linux) and repackage it. So it has everything RHEL has except … Red Hat support. Which, of course, is why most people buy Red Hat, instead of installing Fedora or Debian, etc. I’m not sure what I think of this. I guess it makes sense at one level – here is a way for a vendor to provide basically Red Hat without having to either provide a markup in reselling it, or charge a nonprofit for it.

What makes RHEL “Enterprise” is basically the support. That’s about it. Debian or Ubuntu are just as “Enterprise-class” as RHEL without Red Hat behind it. So CentOS really isn’t any different a choice – unless as an organization you are very familiar with Red Hat, and want to stick with it, but don’t want to pay (or have organizations pay) for it. If you are new to the Linux biz, there isn’t any reason I could see to adopt CentOS.

Novell

I think the article says pretty much everything that needs to be said about Novell, and SUSE Linux. They take directly from the Red Hat playbook, and have a community version called openSUSE. It’s certainly a good option, although in the US, SUSE isn’t very common.

Other Distros

There are, of course five bajillion Linux distributions of varied popularities. Any of which could make a reasonable choice for you as a technology provider (There is a good review in Distrowatch of the top ten distributions.) Only you really know what makes the most sense, given what you want to do with Linux, and what expertise you have on hand. And, luckily, once you’ve learned some Linux, supporting other distributions isn’t such a big deal.