Virtualization and Linux can be a match made in heaven, which is why enterprise Linux heavyweights Red Hat and Novell are pushing so hard to make support for virtualization a highlight of their respective mainstream Linux operating systems. Also recognizing this potential – but pursuing a markedly different tack toward realizing it – is rPath, a company that offers IT organizations and the ISVs that serve them a clever new means of getting the most out of this technology tandem.
Linux is sucessful on many areas, but it has a longway to go on the desktop.
If the market share is less than 10%, it is just insignificant.
That’s a completely arbitrary definition. Why not 9%, or 11%, or 5%? Why not 3.1415%? By your definition, not only is Linux insignificant, but so is OS X.
There are much more sensible definitions of “significant”. Significant might mean when competitors start taking the product into account while deciding on future plans. Significant might mean when a product receives a certain amount of marketing success. Significant might mean when the product develops enough of a userbase to support an active and healthy platform for 3rd party software By all of these accounts Linux (and OS X) is “significant”. It might not necessarily be successful, or dominant, but its significant.
I think Linux has definitely had significant success in the server arena. Nearly every web hosting ISP offers a Linux option for web hosting.
IMO, Linux will be considered significant for the desktop when you can walk into Fry’s, Best Buy, CompUSA or other consumer electronics retailer and go to the “Linux Software” section of the store just like you currently do for Mac and Windows.
To get to that point, I think Linux needs to focus more on standardization for ease of 3rd party software distribution, configuration and support as well as backwards operability.
There are too many Linux variations out there, none of which have an appreciable controlling share of the Linux market that a software development company could reasonably focus their resources developing, distributing and supporting.
On top of that, the frequency at which Linux distributions are released coupled with no guarantee of existing applications working without re-compiling and possibly even re-coding puts a signficant drain on companies trying to keep their software working over the long-term.
Windows is dominant on the desktop because it’s easy to distribute and support software implemented for it and there’s a very reasonable probability that once a company spends the $$$ to get an application built and working for Windows, it will continue working over several years rather than being crippled within six months and requiring more work.
Edited 2006-08-30 01:15
Linux software isn’t distributed via retail, it’s distributed online via repositories. That’s the nature of the system. Setting “retail penetration of 3rd party software” as a goalpost is arbitrary and misleading. Linux could get double-digit desktop marketshare and never develop a substantial brick-and-morter retail presence of third party software.
It should be noted, also, that Linux itself has achieved substantial retail penetration. These days, you can walk into a CompUSA or Microcenter (and in some places even a Best Buy) and buy a copy of Linux. Once people have the distribution, though, there is really no reason for them to go to a store to get software.
There is Linux software out there that you actually have to pay for.
yolinux.com/TUTORIALS/LinuxCommercialApplications.html
& thinking that all these will somehow be replaced by supposedly superior OSS versions is – well – … inaccurate .
So commercial Linux software can be in a shop.
Maybe ALL software will be available online like (?) it is with all of Macromedia’s some time .
“There is Linux software out there that you actually have to pay for.
yolinux.com/TUTORIALS/LinuxCommercialApplications.html
& thinking that all these will somehow be replaced by supposedly superior OSS versions is – well – … inaccurate .”
That listing unfortunately is way out of date. A good chunk are dead links, and some that are listed such as Coffee Cup HTML editor, MusicMatch, and a fgew others were discontinued a couple years ago for Linux. Too many people want all the software on Linux to be free as in beer, so companies discontinue products that don’t make money. What they should do instead of canceling a product just develop a cross platform product..a shock I know.
And that was covered where in the fscking article? Have you ever even seen foresight desktop?
“No I’ve already got an opinion, screw it, here’s my less than 2 cent sound bite troll”.
Generally there is no need to have virtual machine and functional desktop on the same operating system. Just the opposite. Virtual machine should run on stable operating system with low frequency of updates/upgrades. From the desktop one expects that it incorporates all the toys which are available and that brings frequent update/upgrade process. So the conclusion is. Consider http://www.netbsd.org/ for virtual machine and Windows/Linux for desktop.
I am here in a project, where we virtualize about 600 servers. about 15% is Windows and the rest is Linux.
Everything runs on top of VMware ESX v2 (ahgrr! How would I like to run those VMs in ESX v3, but upgrading is not a topic right now).
Anyway… back to rPath:
Maintenance for such an amount of systems is not easy. Linux? Windows? It’s the same. The only big difference is, that for Windows you will find 10’000 tools for distribution-/patch management. For Linux you will not find that much.
I know, that alot of Linux admins and power user will jump up and down and tell me about rpm/yum/emerge/apt-get/whatever… But this is not the point.
I am talking about enterprise ready stuff. And there you will not find that much stuff. No mather on what OS.
The only big difference is, that for Windows you will find 10’000 tools for distribution-/patch management.
Such as ???
Good question.