What Improvments Could be Made to Linux to Make it More Competitive with Microsoft and Apple?
I believe that open-source software, such as Java and Linux are the future of IT.
What improvements could be made to Linux to make it more competitive with Microsoft OS and Apple Mac OS X?
—————-
> Sounds like a homework question to me….
I have a degree in CS, but I want to be an IT instructor. I’ll keep this in mind for a homework question.
Tagged with: apple mac os • apple mac os x • homework question • improvements • Linux • mac os x • Microsoft • microsoft os • open source software • os x
Filed under: Microsoft
Like this post? Subscribe to my RSS feed and get loads more!
Linux is already very competitive in many areas and a very good tool for the skilled computer tech. What I believe your referring to the market share of the nontechnical end user.
Improving Linux will do little here. Significantly expanding market share can only be done by manufacturers, retail stores, and maybe Google’s marketing. My prediction: A hot new inexpensive linux netbook will come out within a couple years, targeted to the younger generation. It will take a nice chunk of market share, and near all will be completely oblivious to what OS is running on it. How many know that the iphone is running Unix?
Sounds like a homework question to me….
The problem is that Linux is free. Make one super awesome distro and sell it to people! Nobody wants free stuff (except open source nerds, myself included).
1. Brand new microkernel. Monolithic kernels are a poor choice for security.
2. Implementation of a useful system policy mechanism. This does not need to be Active Directory integration, but it would be nice to be able to truly integrate a Samba4 machine into a Windows domain.
3. Removal of the root account, replacing it with RBAC based administrator accounts. The name "root" can remain as a local account, subject to system restrictions. This will also mitigate the covert channel vulnerabilities that emerge with compartmented systems.
4. Better file permissions with cross platform enforcement.
5. Better effort put into "bells and whistles." Compiz Fusion and all the other add ons are nice, but overall, the desktop environments lack "flow." With Windows, everything looks the same, everything blends well.
6. It would be nice to see choices other than DAC when it comes to access control systems. Mandatory Access Controls and Role Based Access Controls can help overcome some of the limitations Linux has in terms of security. It’s main protection at the moment is lack of market share.
There are other issues that cannot be fixed on the Linux side, it is up to the manufacturers. Hardware support, available applications, and general availability.
Most of the issues I named are entirely dependent on the first one. A proper reference monitor is intrinsic when considering permission based security. SELinux and AppArmor are poor substitutes. This is not a failure in programming or design for either of them, but a design flaw inherent in monolithic architecture. OSX suffers from the same, as do all UNIXes. A microkernel is the only kernel architecture that can implement a tamper proof reference monitor. Windows is the only common general purpose OS with a microkernel, but there are no high assurance systems with monolithic kernels. Even Trusted Solaris cannot achieve a higher Evaluated Assurance Level than 4+. The best Linux/UNIX Enterprise releases have a 4+. Server 2003/XP have a rating of 4+. Server 2008/Vista have an EAL of 1, although that is a time issue. I expect it to go to 4+, if not 5 for Server 2008. Vista will not be completely evaluated. 7 will be evaluated at a 4+ when it goes through the process.
EDIT: Doug, binary compatability is not the issue. We already have that. Moving drivers cross platform DOES NOT WORK except in a limited number of cases such as NDIS network drivers. The framework of the two operating systems is too different. Heck, even moving drivers between WIndows versions does not work. Vista and 7 have an entirely different HAL and utilize different driver models.
I just installed Ubuntu at home. I love it.
However, I spent hours correcting problems with my sound card, video card, and issues with IPV6.
This type of thing is a deal breaker for the average user.
Someone mentioned the structure of Linux. And yes, it is pretty ridiculous to have to compile your video card drivers every time you update the kernel version (if you’re not using a driver that comes with the distro). Why should IPV6 support be embedded in the kernel, which is a very recent thing and has caused me problems? No way any of this type of stuff should be running at that low a level.
But the structure question, although important, is not what keeps people from adopting Linux. Although usability has greatly improved, there is still a ways to go. The file structure of Linux is a mess from a novice user’s point of view. There are way too many directories and it’s very difficult to know where to look for stuff. It’s not the least bit intuitive.
And yes, there is good reasons why things are the way they are. But that is hardly a good explanation to someone who expects their system to just work.
People will point out that Linux hardware support is terrific. But this is misleading. First of all, most OEM’s still do not make Linux drivers for their hardware. And if they do, they rarely release it under a FOSS license, which causes issues with the FOSS folks and the maintainers of different distros. In order to appease FSF and that bunch, the distros have to make a big deal in their repositories and package managers about proprietary drivers, etc. What typical user is ever going to care about that type of thing?
Also, even if an OEM makes a Linux driver available, if there are problems with it, the OEM’s typically don’t offer any type of real support to home users. They figure they’re doing Linux users a favor by making the driver available to begin with.
The open source drivers available sometimes work just as well, but not always. My printer prints much slower on Ubuntu than on XP, for example.
If Linux increases its market share on the desktop, OEM’s would support it more. So it’s kind of a catch 22.
If the folks at the WINE project and/or ReactOS (or someone else) can achieve binary compatibility with the Windows drivers, then this would bring all the windows hardware support to Linux. This should be able to move the Linux desktop share higher, hopefully high enough to make the OEM’s start supporting it more, and maybe eventually making binary compatibility with Windows unnecessary.
On the software front, there really is nothing holding back Linux IMO, except for of course big title games, and for business an acceptable replacement for Outlook and it’s backend (I forget the name at the moment).
As far as I’m concerned, the package managers and all the software available to Linux users is far superior to Windows. And of course, linux is famous for stability.
There are 1000 little minor usability issues also, a lot of which originate from the fact that Linux users still depend so heavily on the command line. I like the command line myself, but to a user market that has become accustomed to only using a GUI, it’s asking a bit much.
1. Java really isn’t a good choice. It isn’t a polish language or framework.We need an equivalent language for linux that as elegant as C#, with a framework as well thought out as .net. The language should have a Compiler for real time apps, run time environment for Desktop apps, And interpreter for Web apps. Although you can use either or. But those are the target groups. Open GL, and other graphics framework should work together to create a well thought out framework and work together. QT and other GUI frameworks should do the same. No need to waste time replicating effort. An IDE like VS only design in advance with software project management in mind. An IDE that Accommodates any methodology.
2. The Linux kernel should be developed in the OO language that comes about from the first. In Object Oriented paradigm.
3. The Distros should work together and design one official distro taking the best from all of them. And merging them together. That is currently what Sabayon LInux is doing. Implementing the best of all Linux. And integrating it nicely I might add. I was so freaking confuse about all the choice that I figure I would be wasting my time with any I try. Its to hard to choose, from so many negatives and positives. I finally decided to take a chance at linux, But it took me while to go through each one by one.
4. Conservative Linuxers are like a plague. They don’t want linux to advance at all. They are way out there. Take Debian, great OS,,but they still refuse to use propriatetary drivers from hardware manufactures. I mean, who cares if they are proprietary. You have to buy the hardware any way. Same thing with Flash videos, firefox etc..
Sabayon works of the bat, it is just set up from the get go to work, all that matters to them is that linux is free. And that sabayon continues to stay on the cutting edge.
It has been recently argued that OS X is the system to beat and I’ve read articles which have stepped over Windows completely setting the bar and sights on OS X.
As we speak, I’m typing on an OS X box which honestly is a great system which runs Apple software very well and casual usage. I’ve found much greater stress when I try to do business related work on this system as 3rd party software and limited functionality is common place.
However, Linux is really the most powerful system on the market allowing for so much functionality which wouldn’t even be made available within a closed source community.
Linux has already made the lives of it’s users much easier in many areas including installation and software management. I’ve found that Linux machines are becoming eaiser to install and run than Windows boxes. Plus the amount of software made available really allow for so much usability right out of the box.
Where Linux fails to deliver is matching Apples usability of it’s products. Linux needs some growth in these software areas and it will soon rival OS X. Right now though, the technical community already has it as it’s winner.