Archive for the 'Linux' Category

Published by Kenneth on 16 Aug 2009

Is Linux right for your netbook?

Article: “10 reasons Linux should be your netbook operating system

Linux is relatively fast compared to Windows, making it likely the better choice for netbooks than Windows. My mother’s netbook from Dell runs Ubuntu 8.10, if I remember correctly, and she can use it reasonably well. But in his most recent blog post, it is becoming clear that TechRepublic’s Jack Wallen has a few Linux sound bytes stuck in repeat mode in the back of his mind.

“1: Netbook hardware is the perfect match for Linux”

I will say that Jack is correct in that you won’t really be able to do much with a netbook as far as high-power applications and games are concerned, but he slips when he says that “Linux is the perfect networking operating system”.

This may or may not be the case, and it will depend on your netbook. If it came with Windows pre-installed and you are considering switching to a Linux distribution, check with the manufacturer to find out if all of the built-in hardware is supported, especially the wireless network adapter.

Also check to see what kind of support they will provide if you decide to change to Linux (some may cut off support if you do this, something Jack conveniently forgets to mention).

Jack, you need to seriously stop assuming that Linux “just works”, because as many have pointed out, including recently, that isn’t always the case. Just because you were able to successfully convert one person’s netbook to Linux does not mean everyone’s transition will be as smooth.

“2: Netbooks require a secure OS”

Ugh… Jack, why do you keep saying this, in one variation or another: “Your Linux-based netbook can travel anywhere you want and you won’t have to worry about picking up viruses or spyware like you would with a Windows-based netbook.”

As many security experts have pointed out, this again is not always the case, and it is actually only a matter of time before this may not be the case. Many have speculated that the only reason Linux isn’t a virus hotbed right now is because it is not popular.

As has been pointed out numerous times: No operating system is immune.

“3: It’s all about the interface”

In this section of his blog post, he’s definitely showing his “fanboy-ism”: “It’s obvious the interface was well thought out and aimed at the new PC user as well as the new netbook user.” Personally, just as I think I am far from qualified from making this kind of assessment, so too is Jack.

Any person who has been using a computer, any computer, for longer than a couple years is removed from qualification on assessing whether an interface is “aimed at the new PC user” for one simple reason: they are not new users. Sure, we can all remember our frustrations with trying to learn how to use a particular application or operating system, or can you?

The only way to know how friendly an interface is to “new PC users” is to observe new PC users, and how many has Jack observed recently? Probably far too few to make the broad assessment he’s made in his blog post. Which is also why you won’t see me making that kind of assessment regarding an operating system.

“4: Your netbook can be more than just a slow laptop”

There’s not much I can say about this section. He’s comparing two Linux distributions I’ve never used, so I can’t really say anything. However, he closes this section by saying that you can install a LAMP server if you wish. About the only individuals for whom such a capability is useful are those needing to demonstrate web applications (sales personnel, for example).

"5: Linux will keep your cost down”

There’s really only one thing to point out here: Jack seems to conveniently forget that there is commercial software for Linux. Sure, you likely won’t be running any of it on a netbook, unless you need it to keep in touch with the corporate office, but it’s out there. He also seems to conveniently forget that there are free security applications for Windows. I use one on all of my Windows machines at home: Avast AntiVirus.

My laptop runs Windows XP. I paid extra to get Microsoft Office Basic installed as well, because I prefer it to OpenOffice. For my laptop, I’ve only ever purchased one other piece of software: Visual C++.NET 2003. Most who buy a desktop or laptop computer, or a netbook, will not pay for any other software they run because what they need is available for free.

In fact, with my fiancée’s desktop computer, which we built a little over two years ago, the only software she runs that we actually purchased was the operating system. And everything on her machine is legal.

“6: Linux offers more flavors to choose from”

Some would say Linux offers “too many” flavors. It’s been pointed out numerous times as one reason why Linux has yet to “take over” in the desktop market. Sure there are only a few that are actually relevant, in my opinion, but there are still way too many out there and it needs to be scaled back.

This has also been demonstrated as one of the major flaws of the GPL model.

“7: You’ll gain speed”

While I will agree that Linux is fast, he points out one important caveat: there are differences between Linux distributions. And whether you see a gain in speed will depend on which one you choose.

Are all better than Windows? This depends on what you want to do and is highly subjective. I doubt Jack has tested every scenario, so he cannot say with absolute confidence that this is the case.

“8: Improvements will come faster and more often”

Again, Jack is continuing with the empty statements. First, he says “Just like any software in the open source community, the Linux netbook operating systems will continue to improve at a much faster rate than the Windows operating systems for netbooks”.

Jack, please, stop making blanket assumptions with Linux. Okay, you’re a huge fan, but stop with the blanket assumptions.

As one responder pointed out in the comments to the blog post: “Microsoft release security updates as frequently as required and in small "chunks". A Service Pack is merely a collection of already released security updates packaged with new OS features.”

How many people really want their operating system to change frequently? Sure you might like new features, but I prefer my applications to change and adapt, not my operating system. I just want my operating system to work, and I’m sure most everyone else out there feels the same.

“9: The next version will work”

Another statement that Jack is just a bastion of assumptions. With all applications you run the risk that an upgrade will break something. The only difference with an operating system is that the break tends to be a little more mission-critical – it is the software running the computer, after all.

Assuming that the next version will “just work” is preposterous. How will you know unless you actually try it? It is impossible to test an application under all possible hardware and software combinations, which is why no one software engineer will every risk his/her career and reputation by saying “it’ll just work”.

“10: Support is better (believe it or not)”

And I don’t believe it. Never have, never will.

The single biggest problem with community-based support models is the community. The community does not have any incentive to provide good support. Paid support techs at Microsoft, Novell, Oracle, and the like, do have one hell of an incentive: their paycheck. Now this doesn’t mean that the support you get will automatically be good, but it does mean that are repercussions for those who don’t provide good support.

If you get bad support from the community, who do you complain to? The community that gave the bad support.

The choice is yours

Linux is not for everyone. Just like with Windows, there is a learning curve with Linux, though that curve tends to be larger with Linux than with Windows. For me, I prefer Windows, not only because I’ve been using it in its various versions for at least a decade, but also because it works with what I use most. I’m also a Windows software developer, though I’m trying to find ways to break into cross-platform application development.

On my media server, however, I use openSUSE 11.1. Why? MediaTomb. That’s the only reason. I’ve found MediaTomb to be better than the Windows UPnP software I previously used: TVersity. It’s also running MySQL as MediaTomb’s database backend, which is also used as the database for GnuCash, as the latest unstable version can talk to MySQL. I wonder how well GnuCash runs on netbooks?

But the choice is yours as to whether you change over to Linux. Just always make sure that in whatever change you make you have a way to undo what you’ve done.

Published by Kenneth on 14 May 2009

MediaTomb and MySQL

Note: This information may be specific to SuSE and openSuSE

I mentioned in a previous post that I am setting up a media server with openSuse Linux as the OS. The media server software I’ve selected is MediaTomb, which is a great piece of software and very flexible. An interesting issue comes up though if you enable the init.d script.

In my setup, I am using MySQL as the data back-end instead of SQLite. This means that MediaTomb is now dependent upon MySQL at startup, however the default init.d script for MediaTomb does not have MySQL as a dependency. The init.d script must be manually edited in this instance to add the mysql dependency to the “Required-Start” line. After doing this and running insserv, everything is set up.

Just a quick heads up on this for anyone else who may have grown frustrated by this.

Published by Kenneth on 08 May 2009

Assumptions of a Linux geek

TechRepublic: “10 ways to sell corporate on Linux

This article was written almost a year ago, but that doesn’t matter much, in my opinion, as its presented as arguments that can be made, regardless of when, and regardless of what may have changed. Many of his arguments have been ripped at in the comments section, especially in the case of this comment, but I’ll give my 2-cents here and voice my opinion. I should also point out that many of his arguments are nothing more than repetitions of countless other articles trying to also “sell” Linux in the corporate world.

Let’s start with this jewel, of which I’ve heard many variations over the years:

But the best thing about Linux applications is that they’re open source. If there’s something about the application that doesn’t suit your needs, you can change it. If you have the developers in house, more than likely they can adapt an application to do something perfectly suited for your company. Imagine taking a free application and reworking it so that it’s exactly the application you need (down to the look and feel).

This argument provides several implications, whether intended by the author or not. First is the implication that there are no open source applications for Windows, and any Windows user who’s browsed around the Internet knows this is not the case. OpenOffice is available for Windows, and so is GIMP. And of course there’s Mozilla as well. But Windows actually provides more freedom for software distribution over Linux because you’re not forced to make your application open source, unless you use a library with that kind of licensing model, but I digress.

Along with this is the implication that all applications for Linux are open source, another item that he’s pointed out in other blog posts is not true. There are plenty of closed-source and proprietary applications for Linux. I wonder what RMS says about that – seeing proprietary software, which he makes no secret about despising, running on what was supposed to be the OS of his open source movement.

Jack also makes the implication that software is easy to change. Yeah right. If so, I’d be making modifications to OpenOffice to make up for what it doesn’t provide. My biggest frustration with OpenOffice is the lack of postal bar code printing for envelopes. That is one reason, among others, I stick with Word. And why do I care about postal bar codes? Because postal bar codes allow for faster processing and sorting of mail, allowing the check for my car payment (I can’t pay it online right now without a hefty fee) to get where it needs to go quicker. If you want a way to save the USPS some money this year, start printing your envelopes with postal bar codes. Oh wait, except if you’re using OpenOffice, you can’t, not without fidgeting with barcode fonts and other pieces of information! But I digress again.

Anyway, changing software requires many things. I am a software engineer, and I’ve written and developed my own software, along with modifying open source software that is available. To change software, you first have to understand its source code. For some projects, this may not be very time consuming, for others (like OpenOffice or anything else large-scale), it’ll be very time consuming. You will require time to study it, and while documentation can help with some details, without engineering diagrams to go with it, time will be required on the part of the person or team that will be modifying the code. This task only becomes more difficult with the larger projects. How many people or companies are going to do this?

I remember one person in a ZDNet comment area say something similar to “Well if you want a feature that’s not in OpenOffice, then add it yourself.” I remember thinking to myself, “He obviously doesn’t write software”. There is not only modification to the code, but possibly modifications to the GUI. It’s a complicated situation that can be saved by finding software with the features you’re looking for.

“Imagine taking a free application and reworking it so it’s exactly the application you need.” Yeah, then imagine trying to justify the time spent to corporate when you sell them on a piece of software that falls short in many ways. This is where cost of ownership can really be offset by buying a package like Microsoft Office. How so?

Let’s say you have a team of 10 engineers who each make $50K per year. Imagine it takes them approximately 6 months to add a feature to your in-house copy of Open Office, going through the entire engineering process. This means you’ve essentially spent $250K on this one feature. If the engineers already have some understanding of the Open Office source code, then some money is saved, and if they don’t, then more money is probably spent. What if you need more than one feature added? See how it really starts adding up?

Looking at Jack’s profile on TechRepublic, I don’t see any mention of him writing software or being involved with software development, so his fumbling on this idea is not all too surprising, but it’s something he could’ve researched or at least pondered a little more before saying “you can just modify the software if it doesn’t suit your needs.” I’m afraid not, Jack.

And this is true even with other open-source packages. I use Windows Live Writer to write posts for this blog. I could’ve gone with an open-source tool like Zoundry Raven, but I decided on Live Writer because, after trying out a lot of different tools, I felt this one suited me the best. Sure, it’s not open source, but it’s a situation where I bypassed an open-source tool in favor of another freeware application. Sure, I could’ve modified Zoundry to add what I needed, but why when I found something else?

So going back to his first statement, “TCO is bunk”, I think I’ve pretty much shown that it’s more than bunk, depending on what you’re going to do. He is right in that most companies aren’t going to spend money training an IT staff — not when ITT Tech is putting out reasonably-qualified individuals who don’t need training, except in specific procedures relevant to the company and their environment.

Jack also says, “once Linux is up and running, you won’t be wasting precious time fixing, patching, or solving security issues.” Except he seems to forget that the applications running on the system might compromise the security of the system. Open source does not mean “secure by default”, otherwise Mozilla wouldn’t have to patch exploits and release security updates. Same with Apache.

Speaking of security, he also says “Administration is world wide”. Not if the firewall is configured to not allow it — something he doesn’t mention in his article. And why did he use an anecdote of the “Love Bug” virus (which was 7 years old at the time of his posting) to talk about the security of Linux?

First, any server should be properly configured for remote administration, whether it’s Linux, Windows, or whatever the case may be. You never know when you’re going to need to remote log-in to the server to correct a situation. And of course, make sure the remote access is properly secured at all points.

He also says, “Linux is constantly gaining traction.” I believe in politics and marketing this is referred to as “getting on the bandwagon before it rolls away and you’re left behind.” And it’s an argument as atrocious as much of what else he says.

Companies spending money on IT should evaluate their options, and allow time for all reasonable options to be evaluated. Linux might be one of them. Just because “70% of 420 polled business-technology professionals are using Linux” doesn’t mean you should as well. Evaluate your options — you’ll be happier if you did.

Oh and another thing: don’t trust polls, as they’re not scientific.

He also posits WINE as a way to run Windows applications on Linux… Okay he needs to redo his research. WINE is an emulation layer for Windows over the Linux kernel. Is it perfect? Far from, otherwise they wouldn’t maintain a compatibility list on their web site – they’d instead just say “everything works out of the box”. If you’re thinking of implementing WINE, be sure to check out the compatibility list. And the fact the WINE developers appear to be targeting only popular applications means you could be left in the dust for a long time, depending on what you need to run.

Hmm… I wonder how well digestIT 2004 would run under WINE? I’m sure the KCCI project wouldn’t have any issues, but digestIT 2004 might be a much different story.

Well I think that wraps up this post. I’ve used Linux myself, and it has its perks and it has its downfalls, in my opinion. Is it better than Windows? Not for what I do, and that’s the other thing as well. He presents this as “Linux can replace Windows.” Not in all situations, something he also admits in other blog posts.

Plus one place where Windows has always had the edge on Linux is with hardware. Sure Linux can resurrect older hardware, but Windows has always had the edge on the latest and greatest. Why? Because Windows developers can release closed-source drivers, helping to protect their R&D just a while longer.

Now to give Linux some good press in this article, I will say this: I am planning to put up a Linux server running a UPnP media server package on my home network. Onto it will also go Apache, MySQL, and other things. It’ll be no more capable than a Windows server would, but I don’t have any available Windows licenses to put onto the computer, so Linux, I know, will do the job just fine.

What distro will I use? Likely SuSE.