Slashdot Mirror
Open Source on Windows - Boon or Bane for Linux?
A not-so anonymous Anonymous Coward puts this tough issue up for discussion: "There seem some more
determined
efforts underway currently
in some corners of the KDE project to port substantial parts of the software
stack to the MS Windows platform. These efforts are now met by fierce resistance on the part of some of their core developers. Aaron J. Seigo summarizes his reasoning in his blog:
'If the applications people want are available on Windows, they will tend to stick with Windows...by porting software to Windows, we eliminate the
majority of the competitive advantage of Free Software desktops in the eyes of the overwhelming majority of consumers while Microsoft has all the rope they need to shut the door once again on us ... Free Software desktop applications on Windows represent a no-win situation for Open Source, but
Open Source desktops on Free Software operating systems do.'" (Read more below.)
"Does it hurt the 'Linux to the Desktops!' battle fanfare, if Linux apps and other OSS are ported on a large-scale to MS Windows,
or will it rather have a 'pave the way' effect? Does it help to migrate enterprises and public sector units if users to Linux if users are already familiar with Firefox and OpenOffice.org from Windows, or does it take away the motivation to migrate at all? Is porting Unix desktop software counterproductive? Does it even help Microsoft and damage Windows users?"
The other side of that is, if when I switch OS I have to switch ALL my apps at the same time I might now do it. However if I can switch apps 1 at a time using my current OS, when I get all of them switched and am using all apps that will run under linux (from a windows box) then switching to linux will be easier since i already know the apps and have my data files ported and such. It can make the switch easier.
If you want commercial enterprises to develop software for linux, three things need to happen:
1) A significant number of the potential users of the software need to either be on linux or plan on moving to linux within 6-12 months.
2) The development tools need to be useable now, with minimum fuss. Trolltech license for commercial software is a problem.
3) The software vendor needs to be able to insure that software cannot be installed on multiple machines.
4) Software vendors need to be able to distribute binary only packages.
5) Database access to MSSQL needs to Just Work. freetds is not fully stable yet, thus not useable in prod environment. Throw some resources that way someone please...
For those who wonder, here is trolltech commercial license terms for commercial use, from http://www.trolltech.com/products/qt/licensing.ht
License Pricing (per developer)
Platform Professional Enterprise Scripting
Single $ 1550 $ 2490 $ 3750
For the first year, after that:
renewal:
Single $ 510 $ 770 $ 1170
[Rant coming on]
Ex-squeese me? $4992 for 2 years for 1 developer? Heck, should just stay with MSFT and get their MSDN Software subscription... would be cheaper.
I know it costs money to develop commercial applications. But if I wanted to pay, I would pay MSFT, because they're CHEAPER!!!
Of course, I don't want to pay. That's why I python & php -> xhtml and let moz/firefox be my gui.
[Rant done]
Finally. for point 3 and 4, don't know really what do. It's never going to fly with unixers. I also don't see how someone with root can be stopped from copying the software at will.
(before someone jumps in with online software activation: Root, dns, change entry to vendor to point to 127.0.0.1 and write a lovely cgi that just replies: "you're ok to play" to the vendor software.
And if I need a code, well, I just look in the code. What? binary only? Ooops, sw not getting installed on my box to begin with anyway.)
"Piter, too, is dead."
Check out coLinux.
Installation is currently somewhat painful if you don't want to use a provided system image, but progress is definately being made.
Colinux + Xming (or your favorite X server for win32) = Windows and Linux applications running seamlessly side by side, with very little performance loss compared to running Linux natively.
(Don't let the last update of May on the front page fool you, check the snapshots for more recent updates, documentation also updated on the wiki site more often than anywhere else.)
coLinux can be installed as a system service that starts at boot. Put Xming in startup also, and on the Linux side add whatever you want to startup to contact the Xming session and go.
There's no fundamental reason why someone couldn't make a nice package that sets everything up automatically, it's just that so far as I'm aware, nobody has yet done any targetted application setup this way that I am aware of.
Cheers,
When you live in a sick society, just about everything you do is wrong.
Tax software is moving to a web based model. Let's face it - it's easier to control.
It's just accounting software now... And that's a big hurdle.
09 F9 11 02 9D 74 E3 5B D8 41 56 C5 63 56 88 C0
The vast majority of Windows apps are available only on Windows. You might be able to get some of them working under Linux if you invest in Crossover Office, WineX, etc., but that's a lot more work, not to mention somewhat iffy. So one way (some would say the only way) Windows is superior to Linux is in its ability to run Windows apps.
Losing your favourite apps is a big barrier to switching to Linux. But if people get used to using platform-independant applications, than the switch is a lot less painful.
For example, suppose Bob, a Windows user, uses his computer to run MS-Office, Outlook Express and Internet Explorer. Then for him to switch to Linux would require not only learning a new desktop but a new web browser, word processor and email client. For him, it's worth putting up with the worms and spyware (and paying for the associated removal software) just to avoid the hassle of making the switch.
If, on the other hand, he uses Open-Office, Thunderbird and Firefox, switching to Linux may well be worth it because he's familiar with those programs already and they are available under Linux.
Windows versions of Linux apps (FOSS or commercial) make OS compatibility less important and so reduce the cost of switching.
(As an aside, the original article gets a couple of things wrong. Firstly, Mr. Seigo says that few Windows users have decent development tools. If that's true, it's only because they don't want them. MS bends over backwards to provide development tools. They sell them (instead of giving them away) only because that allows competitors to exist and a wide variety of development tools means more developers. And that doesn't take into account all the FOSS tools that have been ported from *nix. Gcc does just fine compiling Windows code.
(Related to that, he also suggests that Microsoft could freeze out Firefox developers. This is highly unlikely because they can't do that without freezing out all third party developers. If they do that, they may as well just delete the Windows source code and fire all their programmers. The entire software industry will move to another platform, almost certainly Linux, and that's the end of the Windows hegemony.
(Microsoft has no choice but to tolerate FOSS applications on their platform because they need third-party developers. That's Windows' biggest selling point. Any dirty tricks to knock Firefox (or whatever) off of windows will work for maybe a week until someone compiles up a new version, but any commercial program that it breaks in the process will be out of commission for a lot longer. And each time Windows breaks an app, they remove one more reason to stick with it for someone.)
I switched to Linux. If I had never tried OpenOffice and the Gimp, I would have stuck with Windows.
I'm still having sound problems, but other that, I'm happy. And it's because of the applications that ran on Windows.
For geek dads: Contraction Timer
I totally agree. A simple example: I wanted to test drive Mantis recently. What I did was download PHP Triad on my laptop, deploy Mantis and try. I know that if we end up using it, it will end up on a Solaris or Linux box but it would have been completely unrealistic for me to do the trial on one of those machines. I would have given it a miss if I had been limited to a *NIX platform. The same goes for FireFox and OpenOffice. If those applications where only available for *NIX, we would not use them because we just can't have everybody on Linux.
Not really. Since I've got a Mac, and started developing with Cocoa, I have become more interested in other OpenStep platforms - GNUstep in particular. I am currently developing an OS X Jabber client which will be released under a BSD-style license. Once the UI is a little more stable, I plan to develop it concurrently on GNUstep (one of the few open source projects that has people that both know and care about UI design). The Jabber/XMPP framework I have written as a back-end should compile on GNUstep with no changes, and the UI should require little more porting than redrawing the Interface Builder .nib files in GORM.
I am TheRaven on Soylent News