It's always fun to see companies reference other companies with which they work. VirtualBox is a software package that allows you run several different operating systems in their own virtual machines (like VMWare, Xen, etc). Their about box is great. They have the Suse Iguana (wearing a Red Hat), the Linux Penguin playing a tambourine emblazoned with the Ubuntu logo, the Microsoft "Butterfly," stars (alluding to Mandrake/Mandriva) rising from the fire, and the BSD Daemon eating a multi-colored Apple. No, VirtualBox can't boot OS X, but the apple is probably an allusion to Darwin, which I assume VirtualBox could boot, and upon which OS X is built. Did I miss any logos?
In the world of system Unix/Linux system administration, one of the statistics often quoted by system admins is their systems' "uptime." That is, how long has the system been up and operational since its last reboot. This is a conversation in which Windows system admins never join due to the fact that even the smallest system or software update often requires a reboot. My personal best? 500-plus days. Of course, I probably should have rebooted it in that time for some security updates, but for some reason it didn't get done.
It seems Linux admins will soon have even more bragging rights when it comes to uptime. A new project from MIT (where else?) has developed a way to apply security patches directly to a running kernel, without having to reboot. Unless an update makes large changes to the kernel (in this case, modifies data structures), a kernel update will not require taking the system down. Considering that a kernel update is the only reason to ever have to reboot a Linux system -- even hardware drivers can be updated without a reboot -- it might be possible to never reboot a Linux system, and still keep it up-to-date with all security patches.
More information at the Ksplice site. Very interesting stuff, indeed.
In response to Tristan's post which is in response to Aaron's post, the logo for Canonical is probably not accidental. Merriam-Webster defines canonical as:
of, relating to, or forming a canon
And a canon is:
a : an accepted principle or rule
b : a criterion or standard of judgment
c : a body of principles, rules, standards, or norms
But it also says canonical means:
conforming to a general rule or acceptable procedure
Which Canonical most certainly doesn't. Well, it may conform to free software, but certainly not to the widely accepted practices of proprietary software.
So, Canonical is a cannon designed to change/improve/destroy the currently canonical definition of how software is supposed to be created, sold and supported, thus forming the new software development canon? Sounds good to me.
It's nice to see Dell promoting Linux now that they are selling machines with Linux pre-installed. For a quick overview on what this Linux and Open Source thing is all about, take a look at this short video.
(Via: Joey Stanford)
Melissa Draper points out that every has their reasons for not adopting Linux. Whether it be the software they have to use, the hardware, or other reasons.
But she raises an interesting point: if your Linux-adoption hurdle is hardware compatibility, it is very likely you could buy Linux-compatible replacements for less than the cost of a Windows Vista upgrade. So if your hardware is keeping you from giving Linux a try, check out alternatives...it might be cheaper in the long-run.
It's happened! Dell has officially said it will support Linux! And what distribution will they officially support? Why, Ubuntu, of course. Details are still filtering out. Keep watching.
For details released before the official word, see this Desktop Linux article.
Woohoo!
UPDATE: More information on the Ubuntu site as well as the Dell site.
Mike Kruckenberg writes about a talk given at the 2007 MySQL User Conference. It seems the United States Navy is using MySQL running on four-node Linux cluster to handle operations aboard one of their aircraft carriers. Great to see Open Source getting its sea legs. Or is that improving its sea legs? Anyone know of other "nautical" Open Source applications or deployments?
The Firefox web browser has an extention called IETab which enables you to open a tab and browse websites via Firefox, but using the Internet Explorer rendering engine. This can help, for instance, if a site staunchly proclaims itself "IE only" and will not make changes for alternate browsers. Over at Hacking for Christ, Gervase Markham has a post entitled "IETab Considered Harmful?". He references the Slashdot article about MovieLink suggesting that users use IETab so MovieLink doesn't have to go to the "trouble" of making their web site Firefox compatible. Gerv observes:
The harm is that this 'solution' still excludes everyone on a Mac or on Linux, and its availability also makes the site far less likely to change to support Firefox properly. In other words, whereas before Mac and Linux users could add Windows Firefox users to their numbers when petitioning sites to upgrade to support web standards, the existence of IETab divides those two groups and gives those of us using non-Windows operating systems, and those who want to see sites supporting standards properly, far less clout.
Very true. The problem I have with IETab, however, is one of security. If you are using the IE rendering engine, you are opening yourself up to all the problems, standards non-compliance, and security holes you'd have if you used Internet Explorer directly. I keep telling friends, family, clients, and colleagues to use Firefox for better security. If there are sites out there encouraging users to "just install IETab to use our site," most of that security advantage will be negated.
At the conservative end, I think IETab should carry a big, fat warning along the lines of "This plugin is for development and testing only! Using it may compromise your system's security!" On the extreme end, I would have it done away with altogether. If you need multiple versions of IE for testing, on the Linux side of things, there is IEs4Linux, which allows you to install and run multiple versions of Internet Explorer on your Linux machine. I'm sure there is something similar for Windows. For Mac, you're still stuck, as IE no longer is available for Mac (at least last I checked). I suppose you could install Linux under Parallels and then install IEs4Linux. 
So, yes Gerv, IETab is harmful, but not only for the reasons you bring up.
Izzy has a good post about security, viruses, and myths. In it, he explores the "age old" question as to why there are no viruses for Macs. While, in the past, the answer may have been market share, the main reason now is the same reason there are no viruses for Linux, Solaris, or any operating system with a strong security model: You. Simply. Can't. Write. One. At least not one that will have any measureable affect.
Why? Two words: security model. In the non-Windows world, users run as normal users, and not as administrators. Any application or script compromised can only modify files owned by that user. Any attempts to modify system files or system binaries will be denied. Now, Windows Vista is supposed to solve some of this by making a user run as a normal user, and prompting for additional privileges when needed, but I've read it's so annoying, people are turning the feature off and running as administrator.
Another reason: bad software design. Example: scripts in Word documents and in Lookout, er, Outlook e-mails that run without user intervention, and e-mail themselves to everyone in your address list. Yes, KDE4 will have DBus-in-everything-even-your-toaster ®, but to my knowledge, scripts attached to e-mails will not run without user intervention. And if you run a script or binary attached to an e-mail, you had better know what it is (there are also the holes in MS products that can be used to lie about the type of a file: calling an executable file a jpeg, for instance).
Another choice that makes Linux a little bit more secure (not sure about Mac) is that any file on a web site or in an e-mail is either opened by the application that can view it ( not execute it) or it must be saved and have its execute permission set. No accidentally executing a program attached to an e-mail here. Granted, you could say, "Open this EXE attachment with Wine" (a windows emulator) or, "Open this Python script attachment with the Python interpreter," but again, you must take an explicit action, and are warned that opening an attachment can compromise your system's security. [Disclaimer: I've not checked the behavior of Kommander scripts for KDE. Clicking on one may offer to open them with the Kommander script interpreter.]
But even if you open an attachment with malicious code, it is running as your user, and no files can be modified other than your own. Conclusion: the virus can't spread on the system, and it can't infect system files. The worst it can do is replicate itself (poorly) to the user's files and maybe other people in the user's address book. It can't install itself as a system service, install a key-stroke logger, or other such malicious behavior.
Note: If you have scripts in your home directory which you run as administrator, make sure they're owned by administrator and not by you. Hmm...need to check my ~/bin. But then again, if you have scripts in your home directory that you run as administrator, it probably means they are custom scripts that would be very hard to write viruses for anyway.
So, it all comes down to security and how it is enforced. If Microsoft wants a secure system, they should write an emulator for backwards compatibility, throw away Windows, and start from the ground up to design and write a secure operating system.
Paul Graham says so. I'll not belabor the point, other than to agree, but when "up and comers" have no sense of "fearing Microsoft," and when those in the tech know-how are all using Mac and Linux, Microsoft is dead. The leading edge eventually becomes standard operating procedure, and Mac and Linux will continue to be adopted on a wider scale.
My fellow blogger and former CDE colleague put up a few random thoughts mostly inspired by, it seems, by his trip to O’Reilly Emerging Technology Conference. I'll respond to a couple.
The trend seems to be: managers and accountants and investors have PCs, the hackers have Macs, the Ubergeeks have Linux.
Well, I guess that makes me an Ubergeek. Or not. But I do enjoy using Linux. However, with trends the way they are, Mac's may soon be losing their "elite" status and become much more common. Microsoft Vista is doing a great job of generating Mac converts.
A nice analogy to flawed project management practices regarding innovation: it’s like assuming that since it takes a woman 9 months to have a baby, all you need is 9 women to have a baby in a month.
A good point to remember from the (almost mythical itself) book "The Mythical Man Month." Just because you have more people doesn't mean you'll be able to program any faster, or generate ideas that are of any higher quality, or even generate more ideas. It takes quality people, quality time to come up with winners.
A couple more that don't need any comment:
Another nice project management thought: schedules are a shared hallucination.
Shouldn’t a medication that helps men stop having to go the bathroom dozens of times a day be called Flo-Min rather than Flo-Max?
Over at Roughly Drafted Magazine, Daniel Dilger has an excellent article on why Apple is set to overtake Microsoft in the desktop market. It is interesting to read how Microsoft is basically repeating the very same mistakes Apple made back late 80's and early 90's: no innovation, high prices, and writing off competition as not being a threat. As I have written in my previous posts about OS X, Apple has a viable (and actually better) competitor to Microsoft's Vista.
Daniel mentions that Apple may overtake the middle to upper tier market, with Windows machines (and their makers: Dell, HP, Gateway, etc) fighting for the lower-tier-deep-discount machines. You know, the ones filled with all sorts of pop-ups and trial-ware begging you to buy the full versions? And for which you must pay an additional $200 to $300 for a decent warranty. If that becomes the case, you'll have pretty much three markets: 1) the low-tier Windows market, 2) the mid-to-upper tier Mac market, and 3) the Linux dominated (in my opinion) server market.
IBM was once seen as unstoppable and un-topple-able too...but arrogance and marketing mistakes changed that. I will be watching to see if Microsoft wakes up and changes their business plan before that happens.
As most of my readership probably knows, I'm not particularly fond of Microsoft. While some of their products are of high quality, for the most part (e.g. Office 2003 or before), the operating system with which they have held dominance in the computer world is a constant source of irritation and frustation for me. Personally, I run Linux and home and at work. However, I support users that run Windows, mainly my parents. I decided long ago that XP would be last Microsoft operating system my parents used. And due to various frustations my father had with system, he agreed. Now, it seems, that decision has even more merit. The negative (and neutral-you-don't-really-need-to-upgrade type) of reviews just keep coming. I highlight three more today.
For the impatient: If you're thinking about a new computer, get a Mac. Upgrading your system to try to run Vista won't be worth the pain, so you'll soon be thinking about getting a new computer, which should be a Mac. If you are techincally inclined, investigate Linux. Kubuntu Linux runs on most hardware out there, and should be a relatively pain-free experience for someone with a modicum of technical knowledge.
First, an article from the BBC about trials and travails in trying to upgrade a computer to run Vista. First a new video card, then the network wouldn't work, then his webcam wouldn't work (even though the Vista upgrade advisor said it would). And why was his three year old PDA no longer supported?
Next, we have an article about the United States Department of Transportation. It seems, for the time being, no computers are to be upgraded to Vista or Internet Explorer 7. The imposition of the ban is interesting enough, but the most interesting part of the article is a quote by DOT CTO Tim Schmidt:
"We're analyzing different client software options and also integration issues," says Schmidt. Among the options the Transportation Department is weighing as a possible alternative or complement to Windows Vista are Novell's Suse Linux and, for a limited group of users, Apple's Macintosh hardware and software, he says.
Oddly enough, moving to Linux will probably cause less compatiblity issues, and will enable them to use some of their older hardware longer.
And, for our final link, this one is another in a long line of "Mac-OS-already-has-what-Vista-has-and-more" articles. Information week says that Max OS X shines in comparison to Windows Vista They say, despite OS X's age, it matches or surpasses Vista in many areas.
Now, I'll be honest, I've read good reviews of Vista, I've even read glowing reviews, but never in my paying attention to computer reviews of Microsoft operating systems (which goes back to about Windows 95 or so) have I read so many reviews that said either "You don't really need this, you can get what you need elsewhere," or outright "Don't buy it." I suppose there are a couple reasons for this. One, there simply are a lot of things wrong, or done poorly in Vista. Two, for the first time in Microsoft's history, there are viable alternatives to their product. Almost all software out there has Macintosh versions available (and if not, you can run them in Windows XP under Parallels or VMWare). If you are more technically inclined, you can try Linux.
Explore the alternatives. You'll be glad you did.
My fellow good blogger Rohan Jayasekera comments on why Macs are more popular now. He's spot on: in industries that have no particular tie to a platform (mainly web and other tech), one is free to use the best platform available. That is why you'll find most tech-oriented people using either a Macintosh or Linux machine to do their day-to-day work. As a system administrator and programmer, I've exclusively used a Linux box at work since 2000, and at home since some time in 2003 (I think).
There are other reasons for the surge in popularity, though. For one, it's just a really good platform. The stability, usability, and security are top notch. I remember using OS 8 and OS 9 systems a long time ago, and while fun, they never really impressed me enough to even think about switching to the Mac. If I had not already been on Linux when OS X started getting good traction, I probably would have gone that direction.
Mac's just work. Their support in the third-party component world has historically been good, but now it's fantastic. You can buy pretty much any piece of hardware and expect it to work on the Mac. And work well, with minimal muss and fuss. Support on the software side is quite good as well.
And for those occasions when you really need Windows Parallels Virtual Machine provides a virtual environment in which Windows runs at nearly native speed.
All this adds up to traction. When the only feasible alternative out there is Windows CloudedVista*, Mac is set to reap some real rewards. It's a system that is fun to use, and easy to learn.
No wonder they are growing in market share.
*Linux is nice, and works well. But it's not always the best choice for a home user.
Michael Urlocker writes that Vista is "overshot." That may be a generous assessment.
Before vista even came out, it was already being called " XP warmed over." A pretty negative review for a company that has spent the last five years working on the replacement. After five years of no updates (save patches to XP's horrendous security, but that's a subject for another post altogether, and no SP1 and SP2 don't count), one would expect an amazing new system that would blow the socks off of reviewers, impress the users, and function in a way never before conceived, giving Vista a unique place in the OS world as a system that was a pleasure to use and generated more productivity than ever thought possible. Instead, what users are getting is a system with relatively minor eye-candy upgrades, and a restrictive digital rights management system that can actually degrade the playback quality of certain media. Wow...that's really compelling.
Something really sad about Vista, especially considering how little of an upgrade it really is, is the considerable hardware requirements that come with it. Vista will mostly likely only come with new computers, and systems sold even a year ago may have trouble running it (consider, for example, low-end systems sold with "only" 512MB of RAM and a built in sufficient--but not spectacular--video card). While I am not a greenie by any stretch of the imagination, I do somewhat agree with them that Vista will unnecessarily obsolete thousands of computers. I highly doubt upgraded systems will have a pleasent user experience.
Another theme among the chatter regarding Vista was the fact that its "new" features were features that were already in Apple's Macintosh operating system, and had been for some time. When Mac OS X was released in 2001 it already had much of the polish and eye candy that Vista is bragging about now. And over the nearly six years since, every time Apple has released an update, most users and reviewers have reported that the operation of the system has actually sped up. Something that stands in stark contrast to the slower operation and new hardware requirements of many of Microsoft's past "upgrades." Another interesting point to note is that the most recent release of Max OS X [at least the most recent version that could run on PPC (G4/G5) hardware] is that it could run on systems that were released when OS X was initially released, and in some cases, even systems released before OS X's initial release.
Let's look for a moment at the "other operating system:" Linux. Built on Open Source principles, and developed by hundreds of developers around the world, it has grown from a "hobbyist's operating system" to an operating system upon which fortune 100 companies depend. While large companies are supporting it now, it attained its status and proved its worth without the corporate backing which MS-DOS and MS Windows enjoyed since the beginning. While OS X and Linux still can appeal to widely different markets, Linux has gained popularity for the same reason Apple's OS X has: it delivers an experience which enables the user to be productive, efficient, and "just get things done." In short, delivering what the user wants.
So why is Vista a miss and not simply and overshot? It's because Microsoft is aiming at the wrong thing. In his article today, Scott Finnie points out that Microsoft's two main priorities have become
1. Avoiding negative publicity (especially about security and software quality)
2. Making sure the largest enterprise customers are happy
If you have the time, go read Scott's article. No need for me to reiterate all the excellent points he makes.
There are many reasons for this, but it comes down to retaining market share. Apple and Linux are biting at Microsoft's heals like never before. OS X and Linux are serious business competitors. And once a user uses a system at work (possibly because they're forced to) they will then be comfortable using that system at home. If you lose the business market, you're well on your way to losing the home market. [Merrill Chapman, in his book "In Search of Stupidity" shows that this is how the PC took over the home market some 20-25 years ago: businesses used PC's first, so people wanted them at home.] Thus, Windows cannot appeal overwhelmingly to the consumer market, or else businesses will no longer want to use it, and will happily chose something else. On the other hand, Vista cannot appeal overwhelmingly to the business market, or it will be dead dull at home (witness Apple's I'm a Mac/I'm a PC ads). No really, go witness them...they're worth it...I'll wait. Apple and Linux, who have nowhere to go but up, don't really care about appealing to one market or the other, they just care about making a secure, easy-to-use operating system. And that philosophy is winning converts.
It seems Microsoft's aim was all (or at least mostly) wrong in this release. Instead of focusing on eye candy and crippling digital rights management (something done to keep media companies happy), there should have been more efforts invested in stability, usable security [Vista's User Access Controls is already becoming a joke. Linux and Mac OS have prompts for additional privileges (when needed, like at software install) down pat], and a better user experience (read Joel Spolsky's rant about the Vista shutdown menu for a good example of features designed by committee).
No...I don't think Vista overshot the market. I don't even think it was aiming at it.
UPDATE (2007-02-02 11:00): Zack Urlocker (Michael's brother) has more observations and humor regarding the Windows Vista launch at his blog.
|
