Slashdot Mirror
Open Source: Facts and Figures
Eloquence writes "Much of the debate about GNU/Linux and open source is dominated by rhetoric rather than facts. David Wheeler has just released a new version of his "paper" (which, at 440,000 characters, is more of an e-book now) 'Why Open Source Software / Free Software (OSS/FS)? Look at the Numbers!'. According to David, this paper 'examines market share, reliability, performance, scalability, security, and total cost of ownership. It also has sections on non-quantitative issues, unnecessary fears, OSS/FS on the desktop, usage reports, other sites providing related information, and ends with some conclusions.' May come in handy when talking to your boss about Linux."
The article referenced does a fair job of displaying the info used. References are linked to, explanations are provided (I.e. the difference between "all sites polled" and "inactive vs active" sites when talking about market share). All in all, an article that raises many good points. Useful, from my perspective.
"Work is the curse of the drinking class" Oscar Wilde
The problem that a paper like this might help solve is convincing others in a corporate/government environment that there are viable alternatives. Only a couple years ago a co-op student was considering writing his paper on Linux vs. Microsoft and wanted to know if he could ask me some questions (being the only Linux guy in the office). I gave him a lot of information and some links. When he presented the idea of the paper and his initial research, the dean laughed at him. The student was told that any competent 3rd or 4th year CS student should be able to crack a Linux box.
If this kind of attitude and mis conceptions exist in University CS departments, how do you expect our managers and directors to have a realistic view of Linux and OSS in general?
XP was workable; it didn't break any speed records, but it was OK to work with.
Debian was workable only after I kicked KDE off the hard drive and went for an slim X setup with Ion as window manager. With Ion it was working OK, as long as I refrained from using Qt and GTK applications at the same time. But then, I wouldn't want to force Ion on an inexperienced user.
As a state gets corrupt, its laws multiply; the most corrupt states have the most numerous laws. (Tacitus, Annales 3:27)
Best guess I can make is that "usability" and "features" are too subjective to comment on. At least with the other categories, you can generate some hard numbers based on records and tests instead of opinions.
=Smidge=
As do I, although most of what I use is on Linux. That may be hard for Windows people to believe, but it is true. I fire up the Windows box when I want to render videos (tmpgenc) or burn DVDs. That is where the DVD burner is installed. (Although k3b is AWESOME). I wish that Irfanview was available on Linux, it just rocks. I haven't found anything I like as much on Linux.
As a wise man once said, can't we all just get along?
Well, that is the kicker. We should all just be able to get along, but then you get proprietary software companies (no names mentioned) that have a heavily vested interest in NOT getting along. The OSS community is more than willing to just get along, but all parties have to be willing.
My beliefs do not require that you agree with them.
Depends on what you install. I feel sorry for you if you install Linux on that and try to use KDE or Gnome. We couldn't even get Mandrake 9.2 to install on a machine with 64M because the installer crashed repeatedly. I've found that I need a machine with 256M to run KDE or Gnome at a non-frustrating speed for the GUI. I had a P3-450 w/ 384M for a while running Mandrake 9.2 w/ KDE and it was usable but not very fast. The same machine running Windows ran fine. I suspect that a bit of it was video card support though.
In any case, I've found that lately, most Linux distros (with GUI) to require the same or greater resources to have a similar experience to Windows (not waiting for redraws and such). This didn't used to be the case. 100% of my job is developing on Linux and has been for a year now and my Linux roots go back to pre-1.0 kernel days so I've played around with it a bit.
i've done that. xp is heavy out the box but if you switch to classic ui theme and remove some of the cruft (indexing, sys restore) it is pretty snappy.
it seems that to get a linux desktop as fast i have to dump the whole advanced uis and switch to really barebones stuff like fluxbox, and then run dillo or something. that says a lot. why cant a linux desktop be as powerful as windows without being a lot slower and memory hogging??
oh and try out windows server 2003, it runs fine as a desktop and is faster than xp/2k. lighter on memory too.
hey i like linux but i need efficientcy too, and every distro i try seems to be so weighted down, disk thrashing, swap usage, slow app starting. same with the apps - compare ie to firefox (tho ff is much better), outlook to evolution, ms office to openoffice, the open source ones all take quite a bit longer to start
i'm seeing more and more posts like this and nothing seems to be being done. linux on the desktop is gonna have a hard time taking off if it remains considerably slower, esp in companies that have loads of 64meg machines. if they have to upgrade to run proper linux desktops then theres one less incentive to switch from windows...
neh.
I run KDE on a laptop with a 800mhz duron, 128MB ram and a crappy 8MB "shared memory" vid card. It works fine. Never slow, sluggish or frustrating. Slackware, btw...
Even when going from my P4 WinXP work machine back to the said laptop, it's ok.
I'll openly admit I didn't RTEFA. Still, through "critical skim" (a frequent management trick), this doesn't seem to be very persuasive.
GNU/Linux is the #1 server OS on the public Internet (counting by domain name), according to a 1999 survey of primarily European and educational sites
Interesting -- using a survey prior to the release of Windows 2000, XP, or 2003 server as the basis for trends today. Reading the article critically (as the hypothetical "boss" would), those numbers aren't as significant as the state of the world today. I may be completely ignorant to research turnaround, but doesn't it seem more recent data would be more relevant?
Consider this one as well:
GNU/Linux is more reliable than Windows NT, according to a 10-month ZDnet experiment
How many companies today are deciding between Linux and Windows NT?
Clearly there are reasons today that companies / governments / users are seriously considering OSS. However, to try to convince through comparison with 5 year old OS is probably not very effective.
Well, I don't know about you, but in my experience, Linux has better fonts than Windows, and I have no fewer hardware problems with Windows than I do with Linux. So I'd be paying $10 - $12 for things that work worse.
I can't use KDE on Windows.
I can't use k3b on windows, I have to spend $100 on Nero.
I can't use TV Time on Windows, and the WinTV drivers on windows don't deinterlace for my card, so TV time is unquestionably better.
If you're going to use open source software, Linux simply has more of the best available. Porting all applications would be non-trivial.
And if you're going to switch all of the main applications of some nebulous naive user, then what's the difference of going a little further and switching to Linux as well? You're still pulling the rug out from everyone, so why not do it all at once? Or are you suggesting having people gradually switch everything to open source and then finally switching to Linux?
From my perspective, Linux almost unilaterally better than Windows. If it weren't for Windows' ability to run more games, then your mere $10 - $12 a year would be equivalent to my flushing that money down the toilet (or worse, because I'd actually be using inferior products). Sure, ripping up a $10 bill once a year isn't much, but you still don't do it. You might as well take your $12 and get something you enjoy with it.
I've come for the woman, and your head.
"Who cares if it runs on 1997 hardware? I want it to run well on todays hardware."
If it could run well on 1997 hardware, imagine how much better it could run on today's hardware. Also remember when all those Linux zealots were comparing their "lightning fast" Linux running 486's to Windows NT running pentiums? Linux then was all about getting the most out of your hardware.
Congratulations, your post totally sums-up why Linux is making VERY slow corporate desktop uptake.
"Oh, just go buy a new box."
You think businesses want to hear that? You think they want to spend millions to run this OS that's supposedly "free" and "lighter on hardware"?
Keep that attitude up -- it's no better than Microsoft. And then keep wondering why Linux has negligible desktop market share.
Those are the same reasons uttered by drunk drivers and drug addicts.
There's actually a LOT of FOSS software out there for other OS's, you're correct. And in many cases FOSS is supplying a need not met even by shareware on those platforms. From a user's perspective, that's pretty much it, right?
Well not from a developer's perspective. Every commercial OS requires you to pay to play; it protects their API and incidentally makes good money. A FOSS OS is better for developers because that barrier to entry is gone, and they benefit from FOSS sourcecode, which means the user SHOULD benefit from better applications, since the motivation for writing software is different.
And the argument from the level of re-education costs is moot. You can't argue for the benefits of whatever's built into an existing OS and on the same hand claim that users don't interact with those features (hint: there's a commercial reason for those features, it cuts out a developer).
So, for the sake of educating users, and giving them choice, there is a place for FOSS on commercial OS's, but it's never more than a stopgap solution.
insecurity asks the wrong question irritation gives the wrong answer
I have no fewer hardware problems with Windows than I do with Linux.
Well, you've been lucky. I've frequently had to go a long way out of my way, spend hours in research, and order more expensive equipment in order to get hardware that will work with Linux. My main problems were with ISDN hardware a few years back and wireless networking adapters right now. In both of these cases, everything I have available from my usual local suppliers was/is based on unsupported chipsets, meaning I had to purchase via mail order.
I can't use KDE on Windows.
Yes you can. It works fine under the cygwin environment.
I can't use k3b on windows, I have to spend $100 on Nero.
There is plenty of free CD burning software available for Windows. I, personally, use the command line tools mkisofs and cdrdao, although I know that there are GUI front ends available for them.
And if you're going to switch all of the main applications of some nebulous naive user, then what's the difference of going a little further and switching to Linux as well?
The point is that they don't have to switch everything, all at once. They can learn to use one OSS application at a time, and then move on to the next when they're happy with it. Then, once most of the apps they're using are open source (and presumably equivalent to Linux versions), they can switch to Linux painlessly.
Client demands Linux
No commercial OS alternative
Even if we have the chance to use it, the requirement is that we do not actually deliver Linux with our code (the customer must obtain the Linux OS themselves and provide a clause to protect the project should actual IP issues arise). Again, this is a legal department requirement. Guess this will provide some additional level of protection.
No one seriously belives that there are "legal issues surrounding IP claims in Linux"
As much as users would like to believe no IP issues exist, in the legal world this issue hasn't been completely resolved. We all await a final decision on this.
Just an FYI, the company I work for does not have a problem with Open Source or Open Standards. In fact we support both (my group has some level of involvement in the Open Geospatial Consortium actually).
As a small business owner these types of studies are interesting, but a write up on the use of open source (like OpenOffice) really drives home the strength of open source.