| This 34 message thread spans 2 pages: 34 (  2 ) > > || |
|Open Warfare In Open-Source|
In-fighting over GPL Terms
| 9:56 am on Aug 22, 2006 (gmt 0)|
|On one side is Richard Stallman and his Free Software Foundation. When Stallman says "free" he doesn't mean price, he means freedom. He believes all software should be freely available to be modified by the public. And for him, this is nothing short of a moral fight. |
On the other is Linus Torvalds, the father of Linux. He and others in his open-source camp believe that freely sharing code simply produces the best software, but if other people want to hide their code, that's fine, too. Companies will just vote with their feet.
Business Week Article [businessweek.com]
| 2:05 pm on Aug 22, 2006 (gmt 0)|
Linus is right. There's no way anyone can force Microsoft, for example, to open-source their code. And I don't think they should, anyway. Open source is great for Linux, but for Microsoft, I think it would be a nightmere.
| 3:29 pm on Aug 22, 2006 (gmt 0)|
I like open-source, but it does have it's drawbacks. If someone has access to the source code then they can discover it's weaknesses.
The same does happen with proprietary software but this involves code hacking. If a piece of software that has become mainstream for example Internet Explorer was to release it's source you can bet the number of attacks and exploits would go through the roof. I'm not trying to take this thread into a MS/Open-source debate but I think it is an interesting point. If any application gets extremely poplar, then there will be people looking to find holes. That's the downfall of success.
As for open source and free software no one can force anyone to make their software free or open source. Tell me again how programmers will earn their wage? Software can be expensive to produce. This is very often reflected in the cost to buy or licence.
Open source is good, but so is the ability for the developer to choose.
| 3:45 pm on Aug 22, 2006 (gmt 0)|
Richard Stallman is right.
Well, I personally think that he's right. From what little I understand of things, Richard Stallman wants to insist that companies using it also have to make "open source" any other software/hardware that works with it. Is my understanding correct? If so I agree with this stance.
If I was a programmer that contributed to software I would be a tad annoyed to find that some companies using the software were two-faced - i.e. were more than happy to benefit from software written by others but were NOT prepared to share their software AND would actively sue anyone who they thought was infringing on their patents.
And lets not beat about the bush here - there are more and more ridiculous patents that are nothing more than common sense. Amazon's "one click" patient anyone? BT's patient on hyperlinks? At least the last one was dropped. These stop people from doing things they should be able to.
In short companies that develop software without RESTRICTING what others can do (via such patents) should be able to use and benefit from open source, while those companies who are trying to horde things to themselves and sue people left right and center for doing the obvious shouldn't be able to benefit from open source.
(P.S. isn't the title of the post a little bit melodramatic?!)
[edited by: TravelSite at 3:52 pm (utc) on Aug. 22, 2006]
| 3:56 pm on Aug 22, 2006 (gmt 0)|
Under most forms of the open source licence you can use the software freely in your own applications but you need to release any software that contains code from the original application, so if you use someone's code in your application then you need to release your code also.
| 4:20 pm on Aug 22, 2006 (gmt 0)|
I have to side with Richard Stallman.
The greatest thing about open source is that if a peice of software doesn't do something you want, or is broken in some way then you as the consumer have the abilty to 'fix' your product rather then waiting for a patch that may never come or an upgrade that won't include a fix to your issue.
Imagine if you bought a car and everytime it had a problem or an upgrade you had to wait for the manufacturer to do something about it. I like the fact that I can work on my car myself.
Even if my car isn't broken but I want to tweak it then I can. Why can't I tweak my MSword? I bought it. It is my product, I should be able to pop the hood, take look and tinker.
| 4:28 pm on Aug 22, 2006 (gmt 0)|
Open source programmers make money because not eveyone in the world can write code, and a lot don't want to know how to wrrite code. Also you can earn a lot of money doing support for your code. CentOs comes to mind for this.
I am bias as I am an open source developer, but trust me I do very well as far as earnings go, and all my clients get full access to all the code to their software, it is one of my selling points.
Also I can undercut all most all my compitition because I don't have license costs.
| 4:29 pm on Aug 22, 2006 (gmt 0)|
|Why can't I tweak my MSword? I bought it. It is my product |
With Word and many other MS and other developers products you don't actually buy the software, you are buying a licence to use it. This means you are agreeing to their terms before you install. Technically the program is still the property of MS you are just allowed to install and use.
In not in favour of this way of doing things, but it's just how it is.
| 4:57 pm on Aug 22, 2006 (gmt 0)|
|Open source is great for Linux, but for Microsoft, I think it would be a nightmere. |
No, it won't be a nightmare. It is the best move they can ever make to fight Google now. If Microsoft continue to stick to their old-fashioned business model they'll be a history in just 5 years!
| 5:02 pm on Aug 22, 2006 (gmt 0)|
Thanks for the clarification Mack.
I don't like it either.
I do see the value in closly guarding your secrets and methods of implmentation, but I see value in setting them free.
It was brought up that open code is more open to attcks because malicious coders will find the weak sopts and exploit them, but on the flip side if those exploits do exist and you open up the code the chance of a "good doer" pointing out the hole and maybe even writing a patch for it on his own time is very high and seems to be the norm in the opens source community.
This results in issues being found quicker and solutions found and done without more work from the original code authorer.
A very interesting debate, I see it both ways but because of my personal beliefs and my current profession I have to advocate for open source.
*edited for spelling
[edited by: Demaestro at 5:03 pm (utc) on Aug. 22, 2006]
| 7:29 pm on Aug 22, 2006 (gmt 0)|
|If Microsoft continue to stick to their old-fashioned business model they'll be a history in just 5 years! |
Nah, there's nothing old-fashioned about Microsoft's business model. Granted, there are things I would do differently if I were in charge, but the path from Windows 3.1 to XP (with the exception of ME) has been steady and significant improvements in stability and usability. In general, I think they're doing fine by the average workstion user.
| 7:56 pm on Aug 22, 2006 (gmt 0)|
It's my opinion that some things should definitely have to be open source...
SUCH AS THE APPLICATIONS USED TO COUNT VOTES IN THE US for example. I think that if the application is going to effect people on that huge of a scale, it really should be open source. I mean at least government stuff. I'd better not say anything more. lol
| 8:40 pm on Aug 22, 2006 (gmt 0)|
|Granted, there are things I would do differently if I were in charge, but the path from Windows 3.1 to XP (with the exception of ME) has been steady and significant improvements in stability and usability. |
It seems as if you don't get my point. Did I say that 3.1 is better than XP, more stable, or more usable?!
In general, business model is how do they earn money? What they charge for? What is the revenue structure and where does it come from?
All Google services and products are free with only one exception, you have to pay if you want to place your ads. MS, just on the contrary, offers every product and service for a fee... until recently (live.com) under pressure from competition. Yes, until recently, they've recognized the mere fact they have to change if they want to survive! They try to make some changes to their model, but what they do is too little and too late.
| 8:53 pm on Aug 22, 2006 (gmt 0)|
Quite a big difference between Google and Microsoft. MS is primarily a software company. Google is preferably a web company. Quite different markets. Both have products and sell their products.
It makes sense for MS to sell it's software, ad it makes sense for Google to offer most of it's services for free, because by making services free they are increasing their advertising portfolio. Google may for the most part be free, but someone pays for it.
| 10:08 pm on Aug 22, 2006 (gmt 0)|
|In general, business model is how do they earn money? |
To me, a business model is how the company provides value to their customers. Part of that is how much money they make in return.
I would prefer that Microsoft would stick to providing packaged software, and that Google would stick to search, advertising, and offerings similar to what they currently have. I believe in specialization, and idealistically speaking, there's plenty of room for MS, Google, and Yahoo (and for that matter, Dell, Sony, etc) to play together.
I agree that Microsoft's Web service offerings are too little and too late, and therefore I think that's wasted effort for them.
|recently, they've recognized the mere fact they have to change if they want to survive |
I don't think they have to change by offering some free services. They just want to have their hands in everything. They want to compete for the sake of competition.
| 12:11 am on Aug 23, 2006 (gmt 0)|
|I like open-source, but it does have it's drawbacks. If someone has access to the source code then they can discover it's weaknesses. |
The same does happen with proprietary software but this involves code hacking. ... If any application gets extremely poplar, then there will be people looking to find holes. That's the downfall of success.
No, that's the advantage of such success. These open-source applications are stable because they are open and everyone can see the code. Historically, being able to see the code translates into quicker fixes, not more exploits.
|If a piece of software that has become mainstream for example Internet Explorer was to release it's source you can bet the number of attacks and exploits would go through the roof. |
Without a doubt, but that's because it's been closed-source for the length of its existance. If IE were open-source to begin with, such exploits would be fixed much faster and it would be a safer product, because we wouldn't be reliant on a single software manufacturer (in this case, Microsoft) to find and fix the bugs. The interested programmers of the world could do it.
But to take a closed-source program, especially one that's already fairly buggy, and open it up overnight wouldn't be wise.
|Tell me again how programmers will earn their wage? Software can be expensive to produce. This is very often reflected in the cost to buy or licence. |
Tell me again why people keep implying that free software can't cost money? The "free" in "free software" means that you have the freedom to do what you want with the source code.. it doesn't mean that the software is necessarily "free" of charge. Free software may or may not be free of charge; so may proprietary software.
|Open source is good, but so is the ability for the developer to choose. |
| 12:27 am on Aug 23, 2006 (gmt 0)|
The discussion in this thread is going to the Open vs. Closed Source debate, but the article was talking about a discussion in the open source camp itself.
That discussion is between the two persons that have made open source successful. Linux and many open source projects are built with Richard Stallman's GNU C compiler and Linux relies heavily on the command line GNU utilities. On the other hand would Richard's software products not have been so successful, and his GPL License not so widely known, if Linux hadn't brought him such a large market share. Until linux came, the GNU project failed to have a good kernel and it was just a bunch of utility programs used on other operating systems to make life easier.
Now these two succesful open source developers divert in their way they see open source. Linus has a very pragmatic view and sees open source as the best method to produce good bug free code. The resulting program is free to use by others, also in propietary configurations. He doesn't want to tell others what to do with his source code. For Linus, the open source way of working is part of the development process, nothing more.
Richard has a more philosophic view. Open source for him means: "If I open my software source to you, and you wish to use it, you have to open your software source to me." This is not defined in the current GPL2 license in that way yet (the license talks mostly about distribution of source, not glueing with other applications), but the current discussion about the GPL3 version is about how far Richard's ideal situation can be implemented in the license.
This discussion is not new and has already resulted in many alternative open source licenses, like the FreeBSD and Apache versions.
Despite my warm feelings for Richard's ideal world, I think we have passed that station. If the GPL3 would contain this mutual publishing requirement, and GNU/Linux (as the OS is officialy called) would be shipped with such a license, all writers of closed software for this OS would be forced to either publish their source code, or move to another platform like FreeBSD or Minix.
In my opinion Richard Stallman is misusing the current situation that his license has become successful via an operating system for which he didn't write the kernel for. The threat that his compiler and utilities are needed to run it are an extremely complicating factor in this situation. If he decides that the compiler and utilities will be released under a GPL3 which doesn't allow coexistence with closed source software, Linus can either choose to accept that license too, with the result that many software developers are confronted with the necessity to open their source. Or Linus can decide to use another license strategy, but in that case his kernel will be without compiler and utilities, meaning the practical end of further development until a decent stable production compiler has been found and the source code has been ported to it, and until volunteers in the field have written all the basic Linux utilities we are used to, like the ls to get a directory listing.
Microsoft has not been able to threathen the position of Linux, but this war inside the open source community itself is IMHO able to cause a quick death of the Linux operating system.
| 12:57 am on Aug 23, 2006 (gmt 0)|
|all writers of closed software for this OS would be forced to either publish their source code, or move to another platform |
I *almost* agree that all software written for an open-source OS should have to be open-source. But not really. As an example, there's a nice little charting library that I use under Linux, which is commercial and closed. If I had the source, I could remove the licensing requirement and port and distribute it at will. While it would be nice if all Linux applications were free, I was happy to pay $100 for this library, and I want it to survive as a supported product.
| 1:08 am on Aug 23, 2006 (gmt 0)|
That $100 dollar utility is just a small example. I know personally of production machines in the $1000,000 range where the controller is running on linux. The manufacturers of these machines would with a mutual-publish version of the GPL3 be forced to either open their source--which often contains detailed information about the operation of these production machines which the competitors would like to know--or they would have to start completely new development on a new OS platform.
The same for all the smaller but wide spread appliances like the Tivo. When you have a network-ready disk on your desk, it is probably running some embedded Linux version.
| 1:38 am on Aug 23, 2006 (gmt 0)|
Yes, Lammert. I'm against trivial/frivolous patents, as TravelSite mentioned. But there large and valid secrecy concerns here.
| 1:42 am on Aug 23, 2006 (gmt 0)|
The most recent draft of the GPL3 can be found at the website of the Free Software Foundation at [gplv3.fsf.org...]
A quote from the draft:
|Some computers are designed to deny users access to install or run modified versions of the software inside them. This is fundamentally incompatible with the purpose of the GPL, which is to protect users' freedom to change the software. Therefore, the GPL ensures that the software it covers will not be restricted in this way. |
This effectively prohibits Linux to run in an embedded system, if it would be released under this version of the GPL3. For many embedded systems, user installation or modification of software would interfere with the functionality, so embedded systems producers HAVE TO protect their system to assure functionality.
| 3:54 am on Aug 23, 2006 (gmt 0)|
|Microsoft has not been able to threathen the position of Linux, but this war inside the open source community itself is IMHO able to cause a quick death of the Linux operating system. |
Probably not. While versions of GNU utilities released after the final publication of the GPL v3 is released will likely fall under that license, all existing versions fall under the current GPL v2 license and will continue to be licensed under that version even when GPL v3 is released. The GPL v2 licensed source can be forked off by others and continue to be licensed under the GPL v2.
Many of the GNU utilities that make a GNU/Linux system work, such as "ls" which you've mentioned, haven't changed (much) in years, so using the GPL v2 licensed versions of those doesn't present a problem.
If the FSF starts licensing everything with GPL v3 and no one uses it in favor of the previously-licensed versions, I think they'll figure out pretty quickly that they need to fix something on their end.
| 4:48 am on Aug 23, 2006 (gmt 0)|
You're better off with a BSD type license.
[edited by: trillianjedi at 9:18 am (utc) on Aug. 23, 2006]
[edit reason] TOS ;) [/edit] [/edit][/1]
| 6:16 am on Aug 23, 2006 (gmt 0)|
Fair is fair. If you build something which will only work with open source software you have built an extension to that open source software and it must be licensed as such.
If you made an open source calculator and I make an analysis package which relies upon your functions then I have extended your calculator to be an analysis package.
Just because the calculator may be physically distributed and actually installed seperately to the analysis system it doesn't meant that they are two different pieces of software.
Richard Stallman all the way.
| 7:49 am on Aug 23, 2006 (gmt 0)|
|Without a doubt, but that's because it's been closed-source for the length of its existance. If IE were open-source to begin with, such exploits would be fixed much faster |
We cant say that for sure...
The reason IE is such a popular target is because if the number of installs. The exploiters will always target the popular apps. If FF was the top browser in terms of market share then it would find it's self the victim of all sorts of unwanted attention.
| 1:13 pm on Aug 23, 2006 (gmt 0)|
I'm with Linus. If you put a lot of work into a project, you have a right to profit from it, and if that can best be accomplished through a closed-source scenario, so be it.
Now if someone could just tell me how I can really obfuscate my HTML and protect my images... ;)
| 4:55 pm on Aug 23, 2006 (gmt 0)|
I think allowing developers choice is in many ways critical for Linux as an operating system. If the rules where laid down like laws then developers may choose not to develop for Linux.
Developers like choice, choice is good for users so it's a win win.
I like open source, but I still like some commercial apps that run on Linux. If I couldn't get these apps for Linux then I probably wouldn't use it.
| 10:41 pm on Aug 23, 2006 (gmt 0)|
There have been some interesting points made in this thread, but I think we need lawman to put in an appearance! I certainly cannot profess to completely understand the legal ramifications of all the changes. However, I do have a couple of comments in reponse to points raised so far:
Firstly, as far as I can tell, the new draft doesn't change anyone's rights to develop commercial / closed source applications for GNU/Linux. They can even use the GNU toolchain to build it. This interaction is not what the 'intimate'/'complex' refers to. However, if your application includes GPL-licensed code, the rest of the code has to be made avaliable.
Secondly, private modifications are (and still will be) allowed. A company can modify an internally-used application to suit their needs without making the code publically avaliable, for example.
Thirdly, the contentious issue with embedded systems. The draft doesn't prohibit the use of Linux on embedded systems, it does mean that one must have the freedom to see and modify the code. I'll use an example to illustrate why this is needed:
Widgets Corp makes a new PVR. This PVR runs Linux, but it will only run code that has been digitally signed by Widgets Corp. The PVR does not come with an ad-skipping facility by default, since Widgets Corp decided to sell that functionality to its customers. Now, because the PVR will only run code that Widgets Corp signs, you and I cannot enable ad-skipping, even if we can write the code to provide that functionality ourselves.
The problem here is that many individuals have contributed to GPL-licensed code, knowing that people who used products based on this code would be able to benefit from seeing and modifying they code, much in the same way the original individuals have. By using hardware such that only signed code can run on the PVR, Widgets Corp have denied the users freedoms granted by the people who wrote the original code. Users may be able to 'see' the code, but since they cannot get it signed, they are effecively left with "look but don't touch". They are locked out.
Now, I may be off base with some of these assertions (like I said, it is my understanding), so if I am wrong, please do correct me and point me to the relevent parts of the draft.
| 11:40 pm on Aug 23, 2006 (gmt 0)|
|The draft doesn't prohibit the use of Linux on embedded systems, it does mean that one must have the freedom to see and modify the code. |
Yes, that is correct, but allowing to modify code is not possible with many embedded applications. In Europe, all devices and machinery must be CE compliant. That means that the manufacturer must assure that the device is operating as expected within predefined limits, safe to use, complying to specific laws and regulations, etc. This CE compliance is for the system as a whole, i.e. software and hardware. (Hardware is more than computer hardware, it can include control units, production machines or even complete production lines). The manufacturer is responsible for the proper working of the system and the CE certification gives him legal backup in case something goes wrong.
If the buyer would have the right and possibility to change the source code, the system may no longer be working as the manufacturer designed it and not be CE compliant, or safe anymore. FYI, I know systems where software on linux is used for precision control of hydraulic rams with 500 tons of power. Such software is closed for a very good reason.
| 12:39 am on Aug 24, 2006 (gmt 0)|
|This CE compliance is for the system as a whole, i.e. software and hardware. (Hardware is more than computer hardware, it can include control units, production machines or even complete production lines). The manufacturer is responsible for the proper working of the system and the CE certification gives him legal backup in case something goes wrong. |
I see your point, but surely they would only have to legally guarantee the system operating under code they [the manufacturers] wrote? Something like a disclaimer along the lines of: "You can use other code, but if you do you are on your own.". I dislike analogies since they are so often imperfect, but...
Standards are also set so that children's toys are safe. If I modified a pair of kids play 'scissors' to include real blades, I would be responsible for such modifications and any consequences that came about as a result. A blunt analogy perhaps, but it illustrates the point.
|FYI, I know systems where software on linux is used for precision control of hydraulic rams with 500 tons of power. Such software is closed for a very good reason. |
(I assume by 'closed' in this case you are referring to 'unmodifiable' software)
For what reason - can you clarify what exactly the reason is? I don't see a problem with requiring it be modifiable - just because it CAN be modified doesn't mean it will. Heck, most people are reluctant to modify the default configuration of their new computer [I'm looking at you, IE ;)]. The person who owns/operates/is repsonsible for the 500-ton hydraulic ram *should* obviously be very hesitant about modifying the controller code, but be able to if they absolutely need it. Still, my guess is that they would call the people responsible for the maintenance of the ram first if such a problem arose.
I may be failing to notice the proverbial elephant in the room, but can someone point out (explicitly) a reason for disallowing someone to modify what should otherwise be open code as intimated by lammert?
| This 34 message thread spans 2 pages: 34 (  2 ) > > |