The Pro: Linux Realities - Gavin Sherry, editor, LinuxWorldIn the last few years, Linux has really exploded into mainstream computing. Whereas once I got a double take from people I spoke to about it, I now hear kids talking - bragging! - about it on the train. It is, however, much to the confusion of disgruntled Windows users, very much a UNIX-style operating system.
But, let's think about this for a second. UNIX technology is not old and obsolete but rather proven. Thirty years in the coming, we find UNIX-style operating systems today more solid and functional than ever. They power the most fundamental aspects of the Internet: indeed, the Internet grew out of a UNIX culture. But, how is Linux any different?
Most UNIX-style operating systems are remarkably inaccessible. Traditionally, computing enthusiasts needed to wait until university and/or extensive training had been reached before they could really delve into UNIX and powerful computing. Linux has changed all this. You can get it on a magazine cover CD, off the Internet at conferences, at uni, off your neighbour -- hell! I bet you can find a CD with Linux on it in a corn-chip packet.
Computer users are given immediate - and free - access to production-level computing software and introduced to mission-critical administration, internetworking, programming and UNIX in general in their own home. This drastically changes the way in which computer administration can be learned: instead of expensive courses introducing users, one can get Linux running and see what it is actually like.
You can fiddle around, break it, teach yourself: all the documentation is there. Using a computer, after all, has always been about personal intuition. But, at the same time, this means that it is not for everybody and it is not everything.
This, of course, does not mean that it is just a geek-tool. Being a UNIX-style operating system, it taps into an existing and skillful market of UNIX professionals: if you can use Solaris or HP/UX, you can use Linux. It comes with the same standard tools, file system and networking model. This is where the fact that it is free and highly accessible becomes most useful: Linux has found a strong role in business and enterprise computing in peripheral but nonetheless necessary roles. Web servers, ftp servers, gateways, cache engines, Internet routers and load-balancers, mail servers: technical roles, but ones fundamental to today's computing. No Linux professional worth their money believes or recommends that Linux should replace all computing roles; nor that it should be used throughout the business environment in general, immediately. It belongs, after all, to no one. Linux does not, in itself, impose: it offers an alternative. It will function within an NT environment, it will run quite happily next to a Sun Ultra Sparc, without a hint of jealousy.
So if Linux does not push itself, why is it being employed in mission-critical and other roles? Namely, because it is useful there. It can utilise old hardware (it will still run on a 386, and now runs on IBM S 3/90 mainframes); it will integrate seamlessly into a Windows environment as a file server or print server; it does a beautiful job in UNIX environments as a development server, backup server, or a UNIX server in general. You choose.
It also has a strong community grounding. If you cannot get something to work, there is bound to be a plethora of other people who could not get it to work either. Solutions to even the most abstract problems are documented by the community of Linux users: solutions are disseminated with a rapidity suitable to the growth of computing knowledge in general.
Linux has also brought something else to the world of computing: a large scale and relevant open-source movement. SGI, Sun, SCO, HP, IBM, Netscape... all the big names are now opening UNIX and other technologies. This open-source movement is commonly misunderstood. It does not exist to encourage every single person to contribute work to remarkably complex computer code but rather so that the attention of all those people that are capable can be sought out. Professional and enthusiastic programmers alike can develop solutions which can be streamlined by one another.
This parallels the very history of the development of computing and clearly signifies its future. Pragmatically, it ensures that mediocre applications can be made more efficient; broken programs can be fixed immediately by those in the know, as opposed to tedious work-arounds being the only solution (think ILOVEYOU virus and MELISSA); and, developers can better integrate applications into computing environments, because nothing is hidden. This is found in the reality of programs like BIND (which powers an effective 100 per cent of the Domain Name System), Apache (the world's most popular Web server) and the dramatic increase in the use of Linux in general.
What Linux offers, then, is really something very simple. Proprietary software is provided as is' -- you cannot change it, no one but the company that wrote it, or an owner of a source licence can change it. Not so under Linux: if you think there is a Linux feature which is lacking, someone is bound to have developed a solution. If not, kernel developers and Linux applications developers are always looking for new ideas. One need not sit back and hope that 'the next release' will have the desired functionality only to find that faceless software development companies have determined that a different feature would be more 'useful'. Open source does not promise to deliver on requests, but it is much more probable than in any other area of software development.
Truly, Linux is an operating system which, from the fringes, has entered the computer industry and changed its direction for ever. It offers alternative and choice; it teaches, but is at the same time a field of new technology; it can be used in minor and major roles, with an unending list of functionality; it comes out of a history of proven and trusted UNIX computing but is, at the same time, deeply grounded in the social setting of computing, in the technology of today's reality.
Agree? Disagree? Respond to email@example.comAnd the Con: YAU! - Ian Yates, editor, The Industry StandardNot since the debate between Macintosh and Windows has there been a topic guaranteed to generate such a stream of hate e-mail. Yes, I'm going to talk about Linux again. I've quadrupled the size of my e-mail database and installed profanity filters, so do your worst.
I don't have a problem with Linux per se, except that it is YAU. Yet Another Unix. It looks like Unix, smells like Unix and feels like Unix. And it is free. And you can have the source code and play with it until nothing works properly. Oh yes, and Microsoft is going to go broke because everyone will switch to Linux. Not.
How about a reality check guys? For starters, it's Unix. So if you expect Windows people to go out and switch to it, you'd have to ask why they didn't choose FreeBSD Unix which has been around for yonks. It works fine too. Trust me. I'm certain a lot of people will grab the CD off the cover of nearly every PC magazine and have a look at Linux. I did. It's a perfectly useable operating system. Easy to install. And current flavours have plenty of GUI add-ons so you don't see much command line confusion.
Then what? Well I know I'm weird, but I use a bunch of applications on Windows that either don't exist on Linux or require me to hunt them down and learn them. Couldn't be bothered. I got past the stage where the computer was the reason to have a computer about 10 years ago. Nowadays I use my computer to do things that have nothing to do with which operating system it runs. Stopped using my Macintosh for the same reason. Lack of applications. If you want to get my attention, come out with Loffice - free and totally compatible MS Office clone. Let me know when you get out of litigation with Microsoft and I'll try your product.
Yes, I know, I've missed the point. Linux is going to replace department and workgroup servers. Oh, really? Because it's free? I've just ordered some new Intel-based file servers, and the price was about $20K each. These are not serious kick-ass servers but good solid machines for a medium-sized website. You could spend a lot more. And Windows NT cost $1K out of that price. So I should use Linux to save myself five per cent? I suspect it will cost me a lot more than that five per cent to get someone to support my new free operating system. Darn. Missed the point again! Linux gives me the source-code! Let's take a serious reality check this time. Exactly how many people do you know that have the skills to actually change the code of a multi-user multi-tasking re-entrant virtual-addressing operating system - and guarantee it still works tomorrow?
I didn't think so. And I'm not about to let you dick about with my web server just because you can. The sort of IT shop that actually has the skills to do this sort of work is probably already using another Unix. So if anyone should feel threatened by Linux is probably the dudes who charge serious money for their proprietary Unix. Mr. Gates is not worried. Mr. McNeally probably is worried. Who knows, with all the vendors rushing to support Linux the major win situation here might well be achieving the goal of interoperable Unix across various hardware platforms. That alone would be worth all the hype we've been putting up with to date.
The very best thing about Linux is that hundreds of spotty geeks are staring at some very clever code and probably learning a lot more than the previous alumni about the way computers and systems actually work. That is not a bad thing. But why should we all run around like lunatics, screaming that the sky is about to fall, because a great knew teaching tool has arrived? I know Windows NT is not perfect. But then, my car isn't a Rolls Royce either. And I couldn't afford to maintain one, even if I got it for free -- with the workshop manuals.