“Sheila Harnett, an engineer at IBM’s Linux Tech Center, told NewsFactor that what holds Linux back the most is the scalability issue. In the ongoing debate between open source advocates and proprietary software makers and users, it turns out that — to at least some extent — both sides are right: Open source does work, but only in some cases. With respect to enterprise computing, analysts agree that for smaller projects that do not involve mission-critical elements, there is room for open source software, such as Linux.” Read the rest of the article at OSOpinion.
Does Open Source Software Really Work?
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
I’m tired of hearing the question “Does open source software really work?” I mean, does Apache work? Does sendmail work? I dunno. What do you think? Does FreeBSD work? Ask Yahoo. I think the real question is “Is open source the answer to all software woes?” And the answer to that is “No”.
IT works for US (not the other weigh around). 150+ websites. not 1 phony billonly payper liesense CONtract in the place of REAL IT.
how much does IT cost?
The question “Does open source work?” is really a lame question. The question should be “Doest this software work? And is it reliable enough for what we need to do? If this is mission critical, is it stable, secure, etc.?” If the answer to these questions is yes, then go with it. Like I have said before, the typical corporate user could care less about open source. They only care about whether the software does what they need it to do or not. So open source software is no different than closed source software in this sense. If it can compete with closed source software in usability, stability, security, etc., than it will do well. If it can’t, it will fail. It really is that simple.
if sometimes open-source software can’t do it, then why do we need GPL? It just infects the “useful” closed software…
“if sometimes open-source software can’t do it, then why do we need GPL? It just infects the “useful” closed software…”
We don’t need GPL. Lets be honest here. There is not a single innovative piece of software that has ever come out of the GPL. Not one. Everything is simply a clone of something that already existed as commercial software. GPL has improved on some commercial ideas, but it has never produced anything innovative. There is a reason for this, and that is that when someone comes up with something truely innovative, they are not going to license it as GPL. Why? Capitalism my friend. They can’t capitalize on their idea with the GPL. They can’t get rich off of selling patent rights, etc.
Although we don’t need GPL, there is a definite need for open source and free software. For one, non-profits and such simply can’t afford to pay hundreds of thousands of dollars for software licensing. For them, products like MySQL and PostgreSQL for example are great.
The other reason is so that comp sci students can learn without Universitys having to pay outragious licensing fees for source code licenses to commercial operating systems and such. And even then, the usefulness of such commercial source code is limited because of all the NDAs involved. Science has a tradition of free information sharing among coleagues. This promotes learning. And this is one area where open source software fills a very important niche.
so what would be the best and friendliest licence?
Actually, I should change what I said in my last message a little bit.
We do need GPL because it fills an important niche in education. I allows comp sci students to colaborate and share ideas without the risk that someone’s idea will get stolen and packages up into commercial software.
“so what would be the best and friendliest licence?”
Personally, I think the best license for open source would be one in which the author distributes the source code with the following conditions:
#1: The end user may not sell a modified version of the product, or distribute a modified version of the product without the original author’s approval. Nor could they incorporate the code into their own product that they wish to distrubute without the author’s approval. However, the end user may modifiy the product as much as they wish for their own internal use.
#2: Exceptions to the above requirement would be made for educational use in which comp sci students and such are doing colaborative learning. However, said students would still not be allowed to sell the software or to distribute it publically as viable production quality software.
I’m sure a lot of people here will disagree with this. But I think this allows free software, colaborative education, and at the same time, protects the rights of the original author. The GPL doesn’t do that, and neither does the BSD license.
There is no such thing as the ‘best’ license in general – the definition of ‘best’ depends on what the author wants to achieve. Simba’s suggestion would work for authors who try to earn some money with a program; other programs which are written just for fun or for personal glory may be better served with a BSD-type license.
And to pick up another thread: even if Open Source does nothing but re-implement existing ideas, it still provides an essential service by establishing a set of programs which can’t be discontinued at the whim of a company exec or by market forces. And that is good for the customers (read: us) because they get more options to choose from.
looks like the license above is not “free” because the creator has control over the code so it’s not free software any more.
IMO actually, that licence is very favorable to developers
There’s an important distinction that I don’t hear made enough: OSS works great for “commodity” software, but not necessarily apps.
Everybody needs an OS, and everybody wants compatibility with open standards. Perfect place for OSS (and GPL IMO).
I really like the GPL and it’s “viral nature” as an OS license. I don’t like it for trying to make money off of innovative ideas turned into desktop apps through sweat & code.
Yeah, I’m glad that I’ve got xdvi, ghostview, etc… but if they didn’t exist (or if they weren’t that good) I’d likely be shelling out moolah for them.
We do need GPL because it fills an important niche in education. I allows comp sci students to colaborate and share ideas without the risk that someone’s idea will get stolen and packages up into commercial software.
What about a comp sci student that comes up with a great idea for their thesis or some other project and then wants to package it into commercial software themselves? If they collaborated with other students to some extent I can see a need to get the consent of those students before doing so, but I see the GPL as a real deterrant in that type of environment. The BSD license was created by a university and, imo, ideally suits the purpose of an academic environment. Then again, I may have a personal slant in that sort of matter because I went to a school that placed the GPL on all code written in that school and/or utilizing the school’s resources (though this was probably in part because it wasn’t very clear to the school whether anything compiled with GCC could be placed under another license, mind you this was a concern for a lot of developers at that time, and has since been clarified to some degree).
“looks like the license above is not “free” because the creator has control over the code so it’s not free software any more.
IMO actually, that licence is very favorable to developers”
It’s free in the sense that the end user is not paying anything for the software. And the end user is free to do what ever they want to the code for their own internal use. But at the same time, it protects the rights of the author in that it does allow them to maintain control over their own code.
I think it trys to strike a happy medium between restricting the end user’s freedoms too much and causing the author to lose control over their hard work, ideas, etc.
Personally, I think authors should maintain control over their own work (or if it was a work done for hire, the company that paid the author to develop the work should maintain control). But there are others who don’t feel that way. If they want to use the GPL, fine. But for me personally, I don’t like the GPL because once the author has GPL’d their product, there is no turning back. They have basically given up all control they have over that software.
“And to pick up another thread: even if Open Source does nothing but re-implement existing ideas, it still provides an essential service by establishing a set of programs which can’t be discontinued at the whim of a company exec or by market forces. And that is good for the customers (read: us) because they get more options to choose from.”
I don’t think I agree with this. After all, most people do not have the programming skills necessary to maintain source code. Most people are users and not programmers. With Microsoft or Sun, the end user can be reasonably assured that the company isn’t going anywhere anytime soon. With open source, what if the authors get bored and move on to something else? Now the end user is left with orphaned software. It doesn’t matter if they have the source code and can’t make heads or tails out of what it means.
“What about a comp sci student that comes up with a great idea for their thesis or some other project and then wants to package it into commercial software themselves?”
I’m pretty much refering to undergraduate studies here where the stuents are mostly just learing about OS design and such. In other words, they are not out on the frontier. They are mostly learning about concepts that other people have already figured out and implemented. With frontier projects that venture into the unknown, GPL is not a good license. I think these projects should be handled like any other scientific research in which multiple people are colaborating on the project. Basically, the ground rules should be laid out ahead of time. But in these types of frontier projects, decisions about source sharing, who gets credit for what, whether the work will be sold, patented, etc. have to be made on an individual per-project basis.
We don’t need GPL. Lets be honest here. There is not a single innovative piece of software that has ever come out of the GPL. Not one. Everything is simply a clone of something that already existed as commercial software.
No link between GPL and what you call “commercial software” (that is proprietary software). GPL and free software came after most innovative software were written. And since then (and probably before), commercial software didn’t bring anything innovative either.
Anyway, there’s no possible relation between a license and wheter the software is innovative or not.
There is a reason for this, and that is that when someone comes up with something truely innovative, they are not going to license it as GPL. Why? Capitalism my friend.
Sure, capitalism is a real problem. It is the context. If you see no innovative GPL software, the problem does not come from the GPL, but from this context.
>Anyway, there’s no possible relation between a license and wheter the software is innovative or not
riight… if it is an innovation that can bring profit in any way, it will be developed and maintained. therefore the innovator must have control over it. if there is no “reward” for the work done, the work won’t be done…
Don’t know why you call this open source license. Actually it couldn’t be free sofware license. The user is just not free, you take away all of his freedom.
Free software is not about egoism and profit. If you want the excessive rights the law currently gives to authors, well then it’s not a problem, just do proprietary software. I don’t see why you want to call your license an “open source” or a “free software” on since there’s nothing free (as in freedom), nothing to share, etc. That’s just proprietary software, nothing new.
“No link between GPL and what you call “commercial software” (that is proprietary software). GPL and free software came after most innovative software were written. And since then (and probably before), commercial software didn’t bring anything innovative either.”
I don’t agree with this at all. I can think of several innovative software technologies that have come out after the GPL.
1. Java
2. Sun’s dynamically patchable kernel
3. CORBA
4. Cold Fusion (of which PHP and ASP are basically copies).
5. HTTP itself
These are the ones that came to mind in about 30 seconds. I can think of several more as well.
With Microsoft or Sun, the end user can be reasonably assured that the company isn’t going anywhere anytime soon. With open source, what if the authors get bored and move on to something else?
Others will continue. While with MS or Sun, no one will be allowed to do so.
“Sure, capitalism is a real problem.”
Define what you mean by this. Capitalism is not the root of the problem anymore than the GPL is. Human nature is the root of the problem. Fact is because of human nature, history has proven time and time again that capitalism is the only system that works. Socialism sounds great in theory. But it has failed every single time a nation has tried to implement it. So if you don’t want to blame the GPL, don’t blame capitalism either. Capitalism is a byproduct of human nature.
Then you’ll easily find innovative free softwares if these are examples of what you call “innovation”.
“Then you’ll easily find innovative free softwares if these are examples of what you call “innovation”.”
Name one innovative product that has come from GPL and I am sure I can probably find a pre-existing product that it was a clone of.
can you name some of those??
Define what you mean by this.
Capitalism does not work, or at least does not work for cooperative projects.
Human nature is the root of the problem. Fact is because of human nature, history has proven time and time again that capitalism is the only system that works.
No, nothing can prove that, and nothing can prove anything about human nature. So human nature is never an argument.
Socialism sounds great in theory. But it has failed every single time a nation has tried to implement it.
So what ? This is completely off-topic (and wrong) : What do you mean by socialism ? Yes it may be wrong, even worse than capistalism. Nations failed to implemtent it ? So what ? Off-topic.
So if you don’t want to blame the GPL, don’t blame capitalism either.
You didn’t say why.
Capitalism is a byproduct of human nature.
Obviously, it’s not.
riight… if it is an innovation that can bring profit in any way, it will be developed and maintained. therefore the innovator must have control over it. if there is no “reward” for the work done, the work won’t be done…
Lots of people who create free softwares do not get rewards for their work, but they have do it, maintain the project, etc. Same thing for innovative software, even if people may be more tempted by proprietary software. As I said, this is because of the context, you consider the need of profit as a “normal” situation. But it’s just context. And talking about innovation doesn’t imply a particular context, unless you say it.
I’m pretty much refering to undergraduate studies here where the stuents are mostly just learing about OS design and such. In other words, they are not out on the frontier. They are mostly learning about concepts that other people have already figured out and implemented.
I still don’t agree much. Undergraduate students tend to have more time on their hands (yeah, it’s a generalization, bound to be shot full of holes), and some amount of commercial software has come from the ideas and work of undergraduate students (Napster is one obvious example that shows both the possibilities and the real need for students to continue their studies…).
With frontier projects that venture into the unknown, GPL is not a good license. I think these projects should be handled like any other scientific research in which multiple people are colaborating on the project. Basically, the ground rules should be laid out ahead of time. But in these types of frontier projects, decisions about source sharing, who gets credit for what, whether the work will be sold, patented, etc. have to be made on an individual per-project basis.
I definitely agree, though the GPL is definitely an option, and should be considered. For the most part, though, universities have a certain amount of ability to flex control over what licenses are used by students, and I think that placing the GPL into this domain has a few problems that people should be aware of before they really push them to do this. Usually universities are fairly capable of balancing these types of decisions on their own, but at times they can be very ignorant of the real world and what effects their actions might have on it. Sometimes people that have been in the university system for a long time get a real utopian view of things, or a very paranoid view of the outside world, and it sometimes becomes evident in decisions made by academic institutions (again, though, that’s generalization and isn’t always the case, perhaps even isn’t the majority case).
>Lots of people who create free softwares do not get rewards for their work, but they have do it, maintain the project, etc.
don’t you think that quality might suffer if there is no money involved in some way or any other reward? I mean, if it is a hobby, why work so hard?
don’t you think that quality might suffer if there is no money involved in some way or any other reward? I mean, if it is a hobby, why work so hard?
I think it really depends on the people involved. Different people have different interests in the code that they write. Many software developers aren’t content to only work on the software they write for a living, so many do contribute to open source software (whether it is GPL, BSDL, public domain, or whatever), write closed-source freeware/shareware, or just write programs for their own use.
The basic idea is that if there’s something in an open source program that you don’t like, you can either change it yourself or ask someone to change it for you. It doesn’t really require anyone to work very hard on it, just that people get what they want out of it, and if someone does work hard on it, it’s because they want to, and inevitably whatever part of the program they work on will be better off because of it. Certainly there are also people getting paid to write software that aren’t putting everything they could into it, as well.
At the same time, I can’t agree with the previous poster’s assertion that people creating free software aren’t getting rewards out of their project. If the work they’re doing isn’t rewarding enough, they should stop working on it. If they’re a maintainer of that piece of software, they need to pass it on to someone else (and possibly take a lesser role in the project if they wish). Volunteer work can be a thankless job, but if the results of the work are not enough, then you should consider doing something else.
but the whole problem is ensuring that the volunteer’s work has been done with high quality. I’m sure there are a lot of high quality volunteer coders. but looking at all the linux gui(for example), I’d say that the quality isn’t that good, although the people who created it are dedicated. here, the quality has suffered because there was no money invested in programmers and research, and the competition has won.
maybe the whole problem isn’t the money reward for the coders, but the funding fot the project?
don’t you think that quality might suffer if there is no money involved in some way or any other reward? I mean, if it is a hobby, why work so hard?
Because it’s exactly what you want… Because there are many people… Look at the Linux distros, Debian the non-commercial one, is one of the most stable/high-quality. Currently the highest quality systems I see are free. Same thing for softwares, but there are some fields with only proprietary softwares.
About your other post about GUIs… I don’t see what’s lacking, or a quality problem… well, compared to Windows or Mac of course. Actually I think none of them are that good.
About innovation. Perl, cocoon, gcc, ant (maybe, dunno), Freenet, CDDB (was opensource first), system+software upgrades “à la apt-get”. There are lots of them but as I said, they came “after” so most used software already existed (that doesn’t mean the free ones are “clones”, most of them are far better than some “original commercial” one — which is more important, todays softwares, or 20yr old ones ?). But for all of them, as I said, I don’t consider them really innovative, but just “as innovative as” the given examples.
“No, nothing can prove that, and nothing can prove anything about human nature. So human nature is never an argument.”
I would suggest you study some evolutionary theory, and possibly some game theory. There is overwhelming evidence here. And we see it in virtually all living organisms. Not just humans.
“So what ? This is completely off-topic (and wrong) : What do you mean by socialism ? Yes it may be wrong, even worse than capistalism. Nations failed to implemtent it ? So what ? Off-topic.”
It’s not off topic, and it is not wrong. Socialism has always failed in every nation that has tried to implement it. And it is not at all off topic because the GPL is based on socialist ideas. Its the “everybody works for the good of the community idea.” And as I said, there is plenty of scientific evidence in all species that things just don’t work this way.
“Obviously, it’s not.”
Obviously, you are wrong. Study some evolutionary theory. Read some writings of the famous psychologists. Capitalism is in human nature. Humans are no different than any other species in this aspect. Like all species, we are in competion with each other for a limited number of resources. Why do you think it is so hard to get a job? If you get offered a dream job are you going to say “Thanks, but I know this other guy wanted it really bad, so why don’t you give it to him instead.”? I hightly doubt it. You are going to take that job.
Capitalism is in human nature. Those who suceed in capitalist ventures can increase their status, wealth, etc. And it is in human nature to do that.
“I definitely agree, though the GPL is definitely an option, and should be considered.”
Sometimes it is an option. Othertimes it isn’t. The reason is that frontier research like this is often funded by commercial companies with the agreement that said company will get exclusive rights to the products of that research.
For example, a commercial company may fund research on a new drug that is being conducted at a University. If the University makes a break through, that company will get exclusive rights to the drug that is produced.
here is a good example: Gnutella
Gnutella is open, etc etc…
but…
Gnutella was developed by Nullsoft. then they realized that the responsibility behind all the piracy that can go on is too big for them to handle, so they just gave it away
“Perl”
#1: Perl is not licensed under the GPL, so its already disqualifed based on what I asked you to do. #2: Perl is nothing more than an invention created because somebody didn’t like TCL which already existed. Sorry, no dice with Perl.
“gcc”
What can you possibly consider to be innovative about GCC? It’s just a C compiler (and not a very good one at that. It generates poorly optimized code). C compilers have existed since the early 70s.
“Freenet”
Once again, it’s not GPL. And it’s not innovative. It’s based on a paper that someone wrote. It’s also based on technology developed by Napster.
“Cocoon”
Again, not GPL. And besides, it’s nothing more than the application of pre-existing technolgies that have been used in Java and such to XML.
“ant”
#1: Not GPL. #2: What’s so great about it? It’s just a build tool written in Java. All it does is improve on pre-existing products.
“CDDB”
#1: Again, although it may have been open source, it was not GPL. #2: What’s so innovative about querying an online database with a CD code number to obtain title/track, etc.?
“à la apt-get”
So what? BSD did this long before Debian did. Even Windows Update Utility came before apt-get.
Try again. Not only are none of these products innovative (they are all simply improvments of things that had already existed), but only 2 out of the 7 products you mentioned are even GPL.
“Does Open Source Software Really Work?”
If you have to ask this question you probably shouldn’t be using
it. Its for some people and not for others.
For some projects an OSS licence has been very successful.
Apache for one (it dominates the web server market).
As for the GPL “infects” useful computers and software argument
I remember a similar argument being used by older developers
when windows first started gaining popularity. Windows is a virus.
It infects computers and software. Fast forward a few years and
oh my. MS is well etablished and a contender to its licensing
schemes GPL has come along and is growing in popularity. Same
“virus” argument again. Is this a flash back or what?
“#1: The end user may not sell a modified version of the product,
or distribute a modified version of the product without the
original author’s approval. Nor could they incorporate the code
into their own product that they wish to distrubute without the
author’s approval. However, the end user may modifiy the product
as much as they wish for their own internal use.
#2: Exceptions to the above requirement would be made for
educational use in which comp sci students and such are doing
colaborative learning. However, said students would still not be
allowed to sell the software or to distribute it publically as
viable production quality software.”
Similar license already exists. See the old Linux is obsolete
thread. This is the beginning of Linux where Linus fought with
Andy Tanenbaum over what he could do and not do with Minix (and
patches obviously piss Linus off).
Two funny notes here:
1) Linus recieves a “F” from Andy because he uses a monolithic
kernel design, something Andy opposes. Now so many years later
we can see Linux (a hobby) kicked Andy’s Minix into the closet
in the educational arena and is working on MS Windows/Solaris/
etc in the commercial secene. Who really has the F now?
2) The reason Linux started in the first place is because
Andy was using a License similar to the one you just
described Simba. What would have happened if Andy had
distributed his source under the GPL or a similar open source
license? Linus would have developed his system using Minix
as the base. The GPL is NOT useless. It just scares the
uneducated and the people who don’t want to change (old
engineer syndrom).
http://www2.educ.umu.se/~bjorn/mhonarc-files/obsolete/
“Anyway, there’s no possible relation between a license and
wheter the software is innovative or not.”
Then why are you arguing against the GPL? Go use and develop
for Windohs if you want to *L* No one is stopping you.
“Name one innovative product that has come from GPL and I am sure I can probably find a pre-existing product that
it was a clone of.”
Same argument follows for most commercial software. MS is a
good example. OS2 (IBM), Lisa (Apple), their hardware (various
small companies they out right shafted like Goldtouch), etc.
Everyone knows the Pirates of Silicon Valley story by now.
MS ‘stole’ their designs from Apple.
Apple ‘stole’ some of their designs/ideas from Xerox.
Xerox is just stupid.
I know for a fact that I saw “groups” in Open Source desktops
before MS released them in XP/2000. Someone will probably telling
me that some system like Amiga or Apple used them before that.
I could go on and on….
Innovative/completely fresh ideas come out under all licences
albiet they are rare. At least the innovations that do not involve
similarities to previous works. Of course I’m starting to
get the feeling that something is innovative only if its
released by 1) a company and 2) that company happens to be MS.
Besides sometimes innovative isn’t whats required to be
successful. Look at Amiga for example. They were slaughtered.
“Capitalism does not work, or at least does not work for
cooperative projects.”
Capatalism does work. The system does work. We have the old
Soviet Bloc to prove that. The USSR had more then enough
resources and people to prove communism works. They failed.
The same holds for socialist democracies like Sweden.
It doesn’t work as well as the form of Capitalism we use here.
Why? They use all their money to set up a social saftey net and
make it impossible for companies to fire people the employ for
example.
That makes all the companies there ultra conservative and
makes it easier for people to sit at home and collect welfare
then go out and work. Innovation suffers because of it and
taxes go up. Your society feeds off itself.
Then again Social Darwanism doesn’t work either. You need
a balance between both realms. A little FDR here and a little
John Adams there.
Open Source was a product of a Capitalist society wether you like
it or not.
Anonymous:
“don’t you think that quality might suffer if there is no money involved in some way or any other reward?
I mean, if it is a hobby, why work so hard?”
Why not? Maybe someone wants to get it right for a change.
Frustrated people will do frustrating things (like write a
*nix clone or desktop from scratch).
“but the whole problem is ensuring that the volunteer’s work has been done with high
quality. I’m sure there are a lot of high quality volunteer coders. but looking at all
the linux gui(for example), I’d say that the quality isn’t that good, although the
people who created it are dedicated. here, the quality has suffered because there was
no money invested in programmers and research, and the competition has won.”
MS has high quality software?
You forget. A volunteer usually enjoys what they’re doing thus
they volunteer (and if nessicary learn along the way). The
software improves as it goes along (just like a commercial
project does). As for the Linux GUI being primitive and
backwards I have to disagree. I’ve been using KDE 2.2.2 for a
while and its pretty nice. I like it better then the XP
interface for example. The only thing I miss with Linux is
gaming support (the newest games anyways) and the ease of install
(in most cases). Then again I’m a college student and I don’t have
much time for games. So its probably a boon.
Anyways I can see why people freak out when the word GPL is mentioned.
Have you seen RMS? He looks like Marx come back from the dead. Big
bushy beard and everything. You have to realize though that
hes only one part of the equation. A lot of people in the GPL
community are actually pretty moderate. Linus for one. He has
said repeatidly that commerical systems are better in certain
areas like GUIs (this is changing tho), etc. Go use what you
want to use. I like Linus’ viewpoint on the whole thing.
I think the question should be rephrased to “Can corps. save money
using Open Source software?”. If used properly I’m sure it could.
Just because some corps fail at it doesn’t mean it won’t work for
me. It works quite well actually (and saves me money because I
don’t haveto spring for a $300 Windows disto) 🙂
As for the GPL stiffles the ability to inovate…. Well I would
say the same thing about MS and their buisness tactics.
“1) Linus recieves a “F” from Andy because he uses a monolithic kernel design, something Andy opposes.”
Monolithic kernels do have their problems though. Andy did have some valid points. For example, there are certain things that operating systems such as Plan 9 can do that Linux will never be able to do because of its monolithic kernel. There really are many limitations monolithic kernels have that microkernels do not have.
“but looking at all the linux gui(for exampe), I’d say that the quality isn’t that good, although the people who created it are dedicated.”
The problems in the Linux GUI are not the fault of the Linux developers or of the GUI developers. The problems are in that the GUI’s rely on an obsolete framework (X Windows) to power their backend. KDE is a damn nice desktop. But X Windows limits its power. Lets be honest here. X Windows is an idea whos time has long since past.
“Then again Social Darwanism doesn’t work either. You need
a balance between both realms.”
Social Darwinism was a complete misapplication and misunderstanding of Darwin’s ideas. Darwin himself would have strongly objected to the concept of social Darwinism.
“Monolithic kernels do have their problems though. Andy did have some valid points. For example, there are certain things that operating systems such as Plan 9 can do that Linux will never be able to do because of its monolithic kernel. There really are many limitations monolithic kernels have that microkernels do not have. ”
True and Linus did admit that microkernels were a nice concept
in his argument with Andy “from a theoretical (and aesthetical)
standpoint”. If the Hurd kernel had been released on time
he says he wouldn’t have even bothered with Linux. He chose simplicity and because
of it the GNU project has a nice kernel and a lot of software
as well.
Its intresting to note that most of the systems that top out
the charts at sites like netcraft use the traditional monolithic
kernel design. BSD is a good example (4 years is the
max I’ve seen and it was achieved with a BSD system). Saying
that a monolithic system is bad is not a reasonable argument
considering the evidence points to the contrary. The concept
might be outdated but those monos are still running strong.
Also Microkernels have their own problems as well performace being a
major one.
Of course if your implementing a real time system you would
probably want the pre-emptive abilities of a micro-kernel
architecture. When it comes to emmbeded systems and/or
RTOS micro-kernels are king. Other then that probably not
the best choice for now.
Besides can you consider Linux or Windows 2000/XP a true microkernel
or monolithic kernel? They both seem to lean towards a side
but are not completely aligned on one side. Linux uses
modules and 2000 runs a lot of stuff in kernel space for
performance reasons.
From what I’ve read Plan 9 isn’t a microkernel at least according
to the developers of the system. They seemed pretty opposed to the
micro kernel concept. It is a transparent distributed networking
system tho.
My question is why isn’t it considered a microkernel?
It seems to share concepts similar to a micro kernel.
Someone mentioned it doesn’t have the message passing
overhead of something like Mach? If someone could explain why
its different I would be thankful 🙂
VSTa which is similar to Plan 9 is a micro kernel.
BTW: Its the 21st century and hurd is *almost* usable.
And yes I really wish they would have given social darwinism
a different name but hey… thats the one that stuck :0p
OK, I’m trying to fathom the mind that can come up with:
– Perl is not innovative because it was made by someone dissatisfied with TCL.
Have you ever actually worked with Perl? Or even read about Perl? Find me the propietary technology that even *compares* with Perl, not to mention whether it came first.
– PHP and ASP are just copies of ColdFusion.
Not so. PHP is actually a programming language, not just a set of tags. ASP is a server plug-in environment that supports multiple languages. They are each different.
– Not one innovation ever came out of the GPL. Not one.
I thought we were discussing open source in general, but either way the argument ignores many issues. Innovations in software are not discrete items, where someone wakes up and says “Eureka! I will build something truly different, taking NO concepts from any other software!”. They are gradual, incremental steps in one direction or another, building upon common frameworks, adapting ideas from one place or another. Name me the last true innovation in software design by a commercial shop. In every case, what really happens in programming is that programmers constantly learn from as many different sources as possible. Closed-source companies are CONSTANTLY feeding in every bit of knowledge they can gain from open source concepts or university research. Open source developers are constantly returning the favor. Microsoft has enriched it’s software so many times from BSD, MIT (Kerberos, etc…), XML (an open protocol) that if you took out all concepts coming from open development, I’ll wager no Microsoft product would work at all. Apple’s latest greatest “innovation” was to take the work of FreeBSD and many other open source initiatives, and put a nice pretty wrapper on it, for which they are now winning all kinds of awards. How about finding me the closed source database that has the special datatypes and object inheritance that was developed in PostgreSQL years ago? Oh, and we are hearing that Microsoft is trying to build a filesystem backed by SQL. Well, that was an open source initiative with “pgfs” back in 1997. Also, open source innovation is happening on many smaller levels that are ignored by the popular IT press. Just browse http://www.freshmeat.net, and http://www.sourceforge.net sometime, and see if your brain doesn’t get twisted a bit.
In truth, open source and proprietary software actually help feed each other. Without any open source or standards, the computing world would fragment so badly that almost nothing would get done. Without any “monetary gainware”, open source would lose most of it’s benefactors, and thus would not be able to sustain itself. In fact, most open source developers also happen to work for proprietary software firms. So in the end, the relationship is symbiotic. Best said in the inimitable (and innovative) words of Perl creator Larry Wall: http://www.oreilly.com/catalog/opensources/book/larry.html
Capitalism:
Not a single evidence. Only utilitarist propaganda. Saying “Capitalism is in human nature” is just ridiculous.
GPL:
Oh you know that GPL is not free software. Since you didn’t know commercial software and proprietary software were different, I assumed you were talking about free software. Anyway you’ve shown you didn’t know anything of Perl, gcc, Freenet, etc. With the same kind of arguments, there’s nothing new in Java and other examples you gave.
“Have you ever actually worked with Perl? Or even read about Perl? Find me the propietary technology that even *compares* with Perl, not to mention whether it came first.”
Yes, I have worked with Perl. And personally, I don’t like it. It is a poorly designed language held together with duct tape. Python is infinitely superior to Perl.
But as far as Perl being innovative, it isn’t. It is just another scripting language that improves on TCL and was intended to be a replacement for AWK (something I think it failed at. AWK is still going strong).
“Not so. PHP is actually a programming language, not just a set of tags. ASP is a server plug-in environment that supports multiple languages. They are each different.”
#1: ASP came out before PHP. PHP was a direct answer to ASP. So it still is not innovative. It’s simply an open source answer to a commercial product. And once again, I might add that it is not even a very good one. ASP is far more versatile than PHP is.
“Eureka! I will build something truly different, taking NO concepts from any other software!”
When I say innovative I am talking about breakthroughs in technology that no one has ever done before. For example, Sun’s dynamically patchable kernel That was innovative. And it was also a breakthrough. It took a great deal of work and research to be able to sucessfully patch a running kernel without causing it to crash. And also, this is the first time anyone has ever done such a thing.
Java is another good example. It took a great deal of work and research to successfully design a VM that could read the same language on multiple platforms. Java is innovative.
Where is the innovation in Perl? It is just another scripting language (and one that as I already stated, is poorly designed in my opinion). There is nothing innovative about Perl. The author simply designed what he through was a better version of what had already been done a long time ago. There is nothing groundbreaking about that.
“Not a single evidence. Only utilitarist propaganda. Saying “Capitalism is in human nature” is just ridiculous.”
Like I said… The evidence is overwhelming. Obviously you have not studied evolutionary theory or psychology.
“Since you didn’t know commercial software and proprietary software were different, I assumed you were talking about free software.”
I do know the difference between proprietary software and commericial software. You simply have a hard time following directions. That’s all. I specifically asked For GPL examples and you didn’t give them to me. Don’t try to pass the buck here. The obvious answer here is that you were not aware that Perl was not licensed under the GPL.
“Anyway you’ve shown you didn’t know anything of Perl, gcc, Freenet, etc.”
I have? Why don’t you try backing up your accusations? I state again: What is innovative about GCC? It’s just a C compiler and it isn’t even a very good one. It doesn’t optimize well.
Truely the mark of immaturity. Resorting to personal insults such as “You’ve shown you didn’t know anything…” and then not giving a single supporting sentence for your statement.
Me: “And to pick up another thread: even if Open Source does nothing but re-implement existing ideas, it still provides an essential service by establishing a set of programs which can’t be discontinued at the whim of a company exec or by market forces. And that is good for the customers (read: us) because they get more options to choose from.”
Simba: “Most people are users and not programmers. With Microsoft or Sun, the end user can be reasonably assured that the company isn’t going anywhere anytime soon. With open source, what if the authors get bored and move on to something else?”
Having a company stick around does not mean that its software will. Software is abandoned in both camps, but with open software there is at least a chance that somebody else will pick it up (after all, some users _are_ programmers), or the user can pay a programmer to pick it up.
In the end it’s a trade-off, like with everything, and the user has to choose according to their priorities. And open software gives the user another option to chose from.
And to pick up another thread:
Simba: “Java is another good example. It took a great deal of work and research to successfully design a VM that could read the same language on multiple platforms. Java is innovative.”
Old hat. The p-Code system did exactly the same thing, up to and including having special processors built to run the bytecode directly. Java however deserves credit for re-invigorating the idea of platform independent binaries and is probably the first commercially successful VM-based system (it helped that modern hardware is fast enough to make this feasible).
“Having a company stick around does not mean that its software will. Software is abandoned in both camps, but with open software there is at least a chance that somebody else will pick it up (after all, some users _are_ programmers), or the user can pay a programmer to pick it up.”
Yes, but commericial companies have customers to answer to. So even if they do discontinue a product, they will usually phase it out gradually and help their customers with the transision to the new product that is replacing it. With open source, things often don’t get phased out gradually. The author just suddenly decides he or she doesn’t want to work on it anymore and they move on to something else. Also, with open source, the user will not get any help with transisioning to something else like they will with commercial software.
– #1: ASP came out before PHP. PHP was a direct answer to ASP.
Oh? So ASP came out in 1994? I don’t think so. OK, so that was just the beginning of PHP, and it didn’t really become a mature product until 1996/97. As I recall, that’s about the time ASP became a, erk…, mature product also.
– ASP is far more versatile than PHP is.
Versatile if you don’t mind being restricted to ONE operating system, and ONE webserver. Versatile if you don’t mind being restricted to only dealing with Microsoft component architecture. As I said, the two products are different. But PHP can call COM, and it can call Java objects, and it can also integrate with more third party applications than any other web scripting environment except for… Perl. I have worked with both, and I have not yet seen a problem solved by ASP that couldn’t be solved by PHP, more quickly, and with more range of possible solutions. How about an example?
And you didn’t really deal with my main point in my previous post, which is that most commercial innovations at some point got ideas from open R&D somewhere, and vice versa, which is actually how any advance in knowledge works. I am not in any way “putting down” proprietary software! Some very fine things have come out of proprietary software.
The main difference is one of perception, I think. Companies with a lot of money can of course find it easier to dramatize this or that breakthrough, while open source developers tend to just plug along making many small incremental improvements. But even in “complete open-source space”, such as Apache, there are “big” innovations. Look at Apache 2.0. Find me the commercial webserver that implements the filtering concept as completely and elegantly as this. Find me the commercial webserver that lets me choose between process forking, threads, or hybrid approach, or even mix them on the same system.
Ok, sorry for the rant. I am just tired of hearing this same old argument about how open source has not *ever* delivered an innovation. I would even accept an argument that *more* innovation comes from commercial source, but it’s only a relative thing, rather than a complete absence of innovation.
The original article that spiked this whole debate asked the rather trenchant question “Does Open Source Software Really Work?”. Of course it does, or companies like IBM wouldn’t be scrambling to “position” themselves in respect to it. Again, a smart symbiosis is the best answer.
Wow this is just turning into one big holy war! Joy!
Arguing which language is better overall and is more innovative is just stupid. Each language has its own strengths and weaknesses.
For example a good one is Java vs. C++. Yes Java is
portable but a lot of people feel that portability is it’s one claim to fame. For example, it doesn’t support pointers or unsigned ints (in order to maintain portability and the pureness of the language). Anyone who works with network software knows that sometimes these things are nessecities (unsigned ints in particular). Ask if it would be possible to add unsigned ints to java (on any of the major java usenet groups) and you’re going to be FLAMED by people who will accuse of you trying to subvert java and turn it into
kitchen sink similar to C++. They’ll say stuff along the line that your a poor programmer and using unsigned ints is a cheesy hack. One problem. A LOT of networking software uses unsigned ints (that cheesy hack). You have a compatibility issue there. The nice portable java language
is not compatible with your network software. You can either
call some external C++/C/whatever code breaking the portability of your code (which is the reason you were using Java in the first place) or just forget it and go C++ all the way (which will end up being faster then the java anyways). Trade offs like everything else.
Same goes for all these other languages:
Perl is great for quick hacks but you probably want to use
Python for large projects (like prototyping a C/C++ proggie).
Here is a good Perl vs. Python arg:
http://www.perl.com/pub/a/language/versus/python.html
(and yeah the new line thing throws me off sometimes too).
Here is a great article on PHP vs. ASP:
http://php.weblogs.com/php_asp_7_reasons
Same with Verilog and VHDL. Verilog is great for quick hacks but you probably want to use VHDL for larger projects.
There is no one be all, end all language in a paticular domain.
We would have still been in the icey grips of Cadance
if the DoD hadn’t bothered to develop the open standard of VHDL.
VHDL wasn’t a new fresh idea (like Verilog) yet it revolutionized the hardware industry to some extent. Cadence’s distribution model failed forcing them to open up Verilog to compete with the standardized VHDL. Cadance like so many other industries and hopefully soon Microsoft lost their strangle hold on a market. Hardware
became much easier to design and less expensive to do so. Hardware flourished and prices dropped. Everyone benefited.
“In truth, open source and proprietary software actually help feed each other. Without any open source or standards, the computing world would fragment so badly that almost nothing would get done. Without any “monetary gainware”, open source would lose most of it’s benefactors, and thus would not be able to sustain itself. In fact, most open source developers also happen to work for proprietary software firms. So in the end, the relationship is symbiotic. Best said in the inimitable (and innovative) words of Perl creator Larry Wall: http://www.oreilly.com/catalog/opensources/book/larry.html“
This I can agree with. You can’t completely eliminate one or the other and do you really want to? I mean most of us like giving our software away for the benefit of the community but you can’t always make a living doing that (tho it would be nice). 🙂
All you Catholics out there have a happy Good Friday BTW.
> do you really want to?
RMS, FSF, GNU… they do…
>>RMS, FSF, GNU… they do…
Ok here is the funny thing:
RMS, the FSF, GNU all have nothing to do with Open Source.
They identify themselves as “Free Software” and sometimes
they work with Open Source advocates. Other then that they
identify themselves as a completely different movement. They’re a stonger version of the free software movement.
In order to be approved by the GPL you have to effectively:
1) Users are free to use the program for any purpose.
2) Users are free to examine the source code to see how it works.
3)Users are free to distribute the program to others.
4)Users are free to improve the program.
Have a problem with it? Use an OSI (weaker then GPL) license:
http://www.opensource.org/licenses/
There only about 30 of them. I’m sure you can find one that
suits your needs.
Or use some proprietary license if you want to keep it closed.
The only reason I can find that people would be ranting about how horrible GPL is because
1) something is GPL’d that they can’t use in their software
because it would require them to release their software
GPL’d in return. A good example of this is KDE using
GIMP code in their desktop system.
2) they like the idea of having GPL advocates helping them
with their work but don’t have the support or intrest
because their software isn’t GPL’d.
3) or the Free Software community is a market they want to
exploit but they need a GPL approval to do it which
means they would have to submit something substantial
to do it.
4) In some way a GPL’d product or the FSF movement
threatens/provides competition to one of their
projects.
http://www.itworld.com/AppDev/350/LWD010523vcontrol4/
Then of course if 1) nothing good has ever been released under the GPL 2) software advocates and developers who support the GPL suck 3) the FSF market really isn’t worth exploiting then you shouldn’t have a problem with GPL. Microsoft shouldn’t have a problem with the GPL.
I think FSF people are people that are tired of seeing stuff like the whole Source Force fiasco/hijacking happening with the source code they worked so hard to produce. If you noticed for quite a while the software the drove Source Forge was under an Open Source License of some sort of Open Source license luring people in to help with its development. Later on they (VA Linux that is) convinced the people who had been developing the software to sign over their rights only to watch the next version turn into a proprietary closed source commercial project with not one dime going to the people who helped work on it.
Not only that but several of the features provided by SourceForge are not so featureful….
One complaint is that the sourceforge system locks user code/project details into it. No one can modify it since its a proprietary now.
http://www.fsfeurope.org/news/article2001-10-20-01.en.html
You’ll notice that a lot of projects are starting to move to Savanah.gnu.org (its based on older open SourceForge code that miracously wasn’t erradicated by VA). Why? People don’t like being ripped off.
Source Forge still claims to be GPLd yet I have a hard time finding any of the source code for it other then that of
an older/modfied version of it avalible at Savannah.
http://savannah.gnu.org/projects/savannah/
So you see Anonymous, your fears are unfounded. It’s still possible to subvert the work of even GNU loyalists to your own needs. I mean VA Systems has done it.
You can’t blame RMS for trying to stop that.
All mistakes in the last post are mine. Sorry about the *L*
“Versatile if you don’t mind being restricted to ONE operating system, and ONE webserver.”
Ever heard of Chilisoft? ASP is available for UNIX.
It’s more versatile because it doesn’t restrict you to one language. It’s more versatile because of ASP.NET. PHP doesn’t have anything similar. It’s more versatile and easier to work with because it has ADO.
“I have worked with both, and I have not yet seen a problem solved by ASP that couldn’t be solved by PHP, more quickly, and with more range of possible solutions.”
I highly disagree with this. There are things that can be implemented in ASP with far fewer lines of code than is required in PHP.
“The main difference is one of perception, I think. Companies with a lot of money can of course find it easier to dramatize this or that breakthrough, while open source developers tend to just plug along making many small incremental improvements. ”
You still haven’t mentioned a single innovative product that has come out of the GPL. There simply aren’t any. It really is that simple. And the reason there aren’t any is that when someone comes up with something truely innovative, they patent it and such so they can capitalize on it.
“But even in “complete open-source space”, such as Apache, there are “big” innovations. Look at Apache 2.0. Find me the commercial webserver that implements the filtering concept as completely and elegantly as this.”
Once again, what is innovative about it? So it implements filtering better than other servers. So what? It’s hardly a breaktthrough like things such as dynamically patchable kernels are.
“Ok, sorry for the rant. I am just tired of hearing this same old argument about how open source has not *ever* delivered an innovation.”
I didn’y say that open source has never done anything innovative. The Berkeley TCP/IP stack for example, was pretty innovative. What I said was that nothing innovative has ever come out of the GPL. And I still maintain that this is true. No one who has posted here has yet to come up with an innovative GPL product. Even if I give it you that Perl is innovative, and Apache 2 is innovative, neither one is GPL. The way the GPL is written discourages using it for innovative software.
“The original article that spiked this whole debate asked the rather trenchant question “Does Open Source Software Really Work?”. Of course it does, or companies like IBM wouldn’t be scrambling to “position” themselves in respect to it.”
Open source works if the software does what I need it to do. It really is that simple. And this is what the open source advocates don’t seem to understand. Like I have stated multiple times, most busiensses could care less if the software they use is open source or not. All they care about is whether it does what they need it to do. Open source developers waste far to much hot air on espousing the benefits of open source, and they don’t spend nearly enough espousing on the benefits of particular pieces of open source software.
If I am a business, I don’t want to hear about why open source is superior to closed source. I could care less. Instead, I want to hear about why OpenOffice is a better product than Microsoft Office. Open source is not an issue. Product features and usefulness are the issue.
“For example, it doesn’t support pointers or unsigned ints (in order to maintain portability and the pureness of the language).”
Why does Java need pointers? Many people would consider this an advantage. Java has automatic memory management which pretty much eliminates the need for pointers. The lack of pointers also eliminates what is probably the most common and difficult to track down source of seg faults, illegal operations, etc. Not having to worry about memory management can literally make the difference between development time of 3 years and 3 months.
“1) Users are free to use the program for any purpose.”
Is this a good thing? I’m not so sure it is. After all, I don’t want Linux controlling nuclear reactors anytime soon. But unline other licenses, the author of a GPL product cannot specifically forbid the end user from using it for that purpose.
“The only reason I can find that people would be ranting about how horrible GPL is because”
I can think of at least two other reasons:
1: GPL is bad for the economy in general because it hurts commercial business.
2: GPL is guilty of the exact same thing that its advocates like to complain about Microsoft doing: ie, stealing ideas created by someone else, copying them, giving the resulting product away, and thus hurting the person or company that spent all of the R&D money to come up with the idea.
“I have worked with both, and I have not yet seen a problem solved by ASP that couldn’t be solved by PHP, more quickly, and with more range of possible solutions.”
I highly disagree with this. There are things that can be implemented in ASP with far fewer lines of code than is required in PHP.
Like any other language out there. Some things are faster in one then in another. What if I told you I could beat both of you with a few lines of Lisp? Would that impress you?
Probably not because 1) I’d probably spend just as much time as you in ASP/PHP programming in LISP if not more 2) no one other then my would really understand what was really going on.
Just the number of lines of code in a program are not the determinative factor here. Its a balance of things including ease of use, understandability, and power.
From my experience PHP was natural to learn since I have a background in perl among other things. The learning curve was very short and learning to use libraries like Fast Templates was fairly simplistic. I have no reason to use ASP right now. I might learn it one of these days but not right now.
“You still haven’t mentioned a single innovative product that has come out of the GPL. There
simply aren’t any. It really is that simple. And the reason there aren’t any is that when someone comes up with something truely innovative, they patent it and such so they can capitalize on it.”
I’m curious Simba…. Define innovative.
Is it something that you can patent (ie it has no prior work)? RTLinux is patented and its GPL’d as well.
http://www.gnu.org/press/2001-09-18-RTLinux.html
http://www.openpatents.org/license/
That isn’t it? What is?
Anyway you define inovative Simba we’ll find a way to counter your statement that GPL software is not innovative and yet commercial software is.
Oh BTW the “Innovative according to Simba” doesn’t work.
“Once again, what is innovative about it? So it implements filtering better than other
servers. So what? It’s hardly a breaktthrough like things such as dynamically patchable
kernels are.”
Wether you define it as “Innovative according to Simba”
or not Apache still dominates the market. Personally I consider Apache innovative. Of course we could argue with each other until your blue in the face. Its not
going to get anywhere.
Open Source works and in some cases dominates. I think this is what has you all riled up.
“The way the GPL is written discourages using it for innovative software.”
Not true. Just more FUD. Refer to the Open Patent used by RTLinux.
“Open source works if the software does what I need it to do. It really is that simple. And this is what the open source advocates don’t seem to understand.”
Oh but open source advocates do understand this. If the product doesn’t do what they want it to do they modify it so it does what they what. Again refer to Linus vs. Tanenbaum. Linus didn’t use one line of code from Andy’s software. Now he has a powerful base that he and others can use and modify as they see fit. If you can’t find any use for it stop bitching and contribute 🙂
“Like I have stated multiple times, most busiensses could care less if the software they
use is open source or not. All they care about is whether it does what they need it to do.”
And if implemented correctly I’m positive there are a lot of cases where OSI projects
like Apache can be more flexible, secure, and cost effective then commercial implementations.
Of course it all depends on what a person needs tho and their are always exceptions.
A good example of commercial software would be something like Circuit Maker or Orcad 🙂 Both commercial apps that I like. Dreamweaver is another one.
Other then that I’m going to try my hardest not to waste a dime on a buggy piece of crap that I can’t at least fix myself.
“Open source developers waste far to much hot air on espousing the benefits of open source,
and they don’t spend nearly enough espousing on the benefits of particular pieces of open
source software.”
And commercial software vendors don’t? Personally, I’m absolutely amazed at how much the Open Source/Free Software community has grown and this is because people communicate and trade ‘hot air’. Good ideas are contagious Simba. Open Source is a good idea (well at least to me).
“If I am a business, I don’t want to hear about why open source is superior to closed source. I could care less. Instead, I want to hear about why OpenOffice is a better product than Microsoft Office. Open source is not an issue. Product features and usefulness are the issue.”
Well I’ve used OpenOffice for a while and it does everything I need. It works for me. Oh and its cheap. Thats something to think about for any start up.
“Why does Java need pointers? Many people would consider this an advantage. Java has automatic memory management which pretty much eliminates the need for pointers. The lack of pointers also eliminates what is probably the most common and difficult to track down source of seg faults, illegal operations, etc. Not having to worry about memory management can literally make the difference between development time of 3 years and 3 months.”
And Java suffers the same problems Lisp did with Garbage Collection. Mainly 50% more memory is used then in programs that do manual memory management. Also again you can’t do some of the niffy things that you can do with C++.
http://www.jelovic.com/articles/why_java_is_slow.htm
A trade off of security for flexability and efficency. This could go on and on…. Its a holy war. Its stupid.
Every language is going to have its trade offs.
I know I would rather use Java in a critical enviroment
then C++.
“Is this a good thing? I’m not so sure it is. After all, I don’t want Linux controlling nuclear reactors anytime soon. But unline other licenses, the author of a GPL product cannot specifically forbid the end user from using for that purpose.”
Well they’re thinking of putting it into shuttles so why not into nuclear reactors 🙂
http://flightlinux.gsfc.nasa.gov/
Oh boy I smell Microsoft propaganda in here…..
1: GPL is bad for the economy in general because it hurts commercial business.
The GPL and other OSI licenses are just that Licenses and legal ones at that (unlike Borlands Kylix and C++ Builder licenses). Last time I heard the FSF was considered a company. An eccentric one but still a company. If anything the GPL and OSI licenses provide competition to other companies and capitalism works. People improve their products or product distribution methods to compete (or they don’t and they suffer). If you can’t stand the heat in the kitchen, then get out.
2: GPL is guilty of the exact same thing that its advocates like to complain about Microsoft doing: ie, stealing ideas created by someone else, copying them, giving the resulting product away, and thus hurting the person or company that spent all of the R&D money to come up with the idea.
They take what they legally can just like any other company does. I’ve never heard of the FSF threatening to punish a PC distributor because it was introducing a competing product (MS vs. Gateway), setting up roadblocks in their system so a competitors software wouldn’t work with it, or completely railroading a company and blatantly stealing their patented ideas (MS vs. GoldTouch).
You have a long ways to go before you can even consider GPL in the same boat with MS.
As for the wasted R&D, thats not even remotely believable. Most companies have their legal departments check the GPL or any other open source licenses out before they use it in any major pieces of work last time I checked. Just like you said Simba they use what they want to use. No one is forced to use the GPL or any OSI license (unless they’re using GPLd code in which case they must GPL it in some cases just like a company would charge another company money for using its code).
If you don’t like the GPL later on you can always switch away from it and use a different license for a later version. The only catch is that the earlier versions have to remain GPLd. Tux Racer did this and so can you.
heck, that was a long one!
Eugenia, please keep this thread on top for a while longer, I have a feeling this is going to have 100+ posts.
——————————————————–
don’t yo all think this fight for “freedom” is worthless?
never will all soft be 100% free (as in speech). and I don’t see the customer needing the source code so much.
yea, he can fix bugs, but then, the support for software he purchased is gone, because he altered it. imagine this:
you buy an open (GPLd) version of rational rose. you get the source code with you, and you add a feature or fix a bug.
what if you changed the initial functionality? rational soft can’t support it anymore, so you just wasted $400(+) on support that is useless.
it would have been better to give it away than sell it. and buying a (somewhat) mission critical app like that definitely reqires tech support. if there is no tech support, the quality suffers.
what if it was closed? you file a bug report or request a feature, then rational soft does the job and tech support is available. more than that: they test it and they are sure that it will work 100% now.
“Anyway you define inovative Simba we’ll find a way to counter your statement that GPL software is not innovative and yet commercial software is.”
When I think if “innovative”, I think of something that has to be something more than simply a new twist on an old idea. That is what filtering in Apache is. A new twist on an old idea. That is also what Perl is in my opinion.
Then there are things like Sun’s dynamically patchable kernel. That is something that sets the kernel apart. Nothing else can do anything like that.
“Oh but open source advocates do understand this.”
If they understand it, than they really need to do a better job of marketing it. Like I said, as a business, I don’t care if you think OpenOffice is better than Microsoft Office because it is open source and MS Office is not. But I do care if you think Open Office can meet my needs better than MS Office in terms of economics, features, etc. (BTW, I like OpenOffice. And I do use it for most things. But the fact that it is open source did not figure into my decision in any way. It was primarily a features vs. cost analysis. It had the features I needed and it didn’t cost me anything. Microsoft Office does have more features than OpenOffice, but if they are features I don’t need, why spend money on them?)
“Not true. Just more FUD. Refer to the Open Patent used by RTLinux.”
The open patent seems rather useless to me. After all, what good is it? I can’t legally charge people to use GPL’d code, nor can I legally prevent them from using that code in their own products provided that the products it is used in are also GPL’d. I don’t see the point in a patent like this.
“Oh boy I smell Microsoft propaganda in here…..”
Not propeganda. Just perfectly valid points about GPL software. I didn’t even say I agree with them. I just pointed them out as perfectly valid reasons why people might not like the GPL.
“They take what they legally can just like any other company does.”
Just because it is legal doesn’t mean it is ethical. The courts ruled for example that Microsoft did not do anything illegal when it borrowed ideas from the Mac GUI for Windows. But there are a lot of people out there who still complain that Microsoft basically stole Apple’s idea. And yet these are the same people who seem to have no problem with the fact that KDE and GNOME both stole many ideas from the Windows interface (and the Mac interface). All I am saying here is don’t set a double standard. Don’t complain about Microsoft doing what you your yourself are doing the very next day.
“As for the wasted R&D, thats not even remotely believable. Most companies have their legal departments check the GPL or any other open source licenses out before they use it in any major pieces of work last time I checked.”
I think you misinterpreted my statement. What I meant by this was that a company will spend millions of dollars on R&D, develop an idea, release a commercial product, and then some open source author will create a GPL clone of said product. Like it or not, this does hurt the original business that spend all the money on R&D. And it borders on the unethical.
“And Java suffers the same problems Lisp did with Garbage Collection. Mainly 50% more memory is used then in programs that do manual memory management. Also again you can’t do some of the niffy things that you can do with C++.”
This a valid point. But many would argue that with memory as abundant and cheap as it is today, it’s not an important issue. In most cases, the ability to deploy an application in 3 months instead of 3 years far outweighs the increased memory usage now that memory is about worth its weight in sand.
“Wether you define it as “Innovative according to Simba”
or not Apache still dominates the market. Personally I consider Apache innovative.”
I don’t consider Apache to be innovative for a few reasons. After all, it’s just a web server. Sure it implements some ideas better than other web servers. But at the same time, there are some ideas it doesn’t implement so well. For example, the fact that its not threaded (I know this will change with 2.0, which will be partially threaded, but still not fully threaded) is a performance bottleneck. None the less, I do think Apache is a better web server than IIS. But once again, if I am a business, open source doesn’t figure into that decision at all. Intstead, features, cost, etc are what make that decision.
Bottom line is this: Open source authors need to do a better job of marketing their products based on their own technical merits. Because using “it’s open source so it is better than so and so’s product” is not working.
Red Hat seems to be finally understanding this for example. Look at their adds in trade journals. They talk about why Red Hat Linux is a good server OS. They talk about its technical merits. But not once in these adds will you ever find the word “open source”. Why? Simple. The target audiance of these adds doesn’t care whether it is open source or not. They only care about why it is a better choice for their server than Windows is based on technical merit, features, economics, etc.
Simba, I was going to ignore this thread, but I just can’t take this last one… (I will even ignore the many other issues you have not-so-gracefully sidestepped 😉
Hello? Apache was “just a webserver” back in 1996. At that point, it was an improvement on the innovative (also open source) NCSA server. But since then, it has evolved to become the most versatile and extensible webserver out there. If versatility is not innovation, then what do you want it to do? Cook your beans?
“Apache is just a webserver” is like saying Unix is just a data storage mechanism.
But furthermore…
Do you even know what the filtering concept is with Apache 2.0? I don’t think any commercial webserver has this, but even if they do, none have implemented it in quite as sweeping a method as Apache 2.0 has, where any output from any module can be passed to another module for further processing AS IF it was a source file. You can have a whole chain of filtering directives, each handled in its own native mode, without worrying about the origin of the data. This IS an innovation, and I didn’t see it being offered anywhere else, and I can’t wait to start using it.
“(I know this will change with 2.0, which will be partially threaded, but still not fully threaded)”
Sorry, wrong again. Apache 2.0 will again do something no other webserver does. As I said earlier, you will be able to configure it to be fully threaded, Or, you will be able to configure it for pre-forked processes (which has some stability benefits). Or, even more interesting, you can configure it so that you fork processes which spawn separate thread groups (this can isolate certain dangerous tasks from taking down the whole webserver). Or you can configure mixtures of processes and threads.
Sounds like innovation to me, but even if we disagree on what the word means, your remarks tend to be most disingenuous indeed. YOu may know a lot about certain things, but please stop blurting out spurious opinions on areas you abviously haven’t researched. And it is not a holy war with me at all. I don’t care whether something is commercial or not. I am only interested in technical merits and good workmanship. I just hate disinformation masquerading as opinion.
Anonymous:
“don’t yo all think this fight for “freedom” is worthless?”
In the words of Wallace/Mel Gibson:
Frrrrrrrrrrrrrrrrrrrrrreeeeeeeeeeeeeeeeeeeeeeeeeeeddddddddddddddddddoo ooooooooooooommmmmmmmmm eeeehhhh
“never will all soft be 100% free (as in speech).”
And this I agree with. The is why I like the opinions of Linus so much. The will always be places where commercial software
is better a better choice then open source and visa versa.
“and I don’t see the customer needing the source code so much.
yea, he can fix bugs, but then, the support for software he purchased is gone, because he altered it. imagine this:
you buy an open (GPLd) version of rational rose. you get the source code with you, and you add a feature or fix a bug.
what if you changed the initial functionality? rational soft can’t support it anymore, so you just wasted $400(+) on support that is useless.
it would have been better to give it away than sell it. and buying a (somewhat) mission critical app like that definitely requires tech support.” if there is no tech support, the quality suffers.
What if you have inhouse support? Why is OpenBSD and Linux so popular with groups like the NSA and in general people that are
security conscious? If I were truly concerned about the quality of a piece of software I was going to implement in a critical environment
wether it needs to be secure or stable I’m going to want to be able to audit the source code (with my own programmers) to ensure it does
what it says. If you find some bugs you can report them back to the vendor and pray they fix them. If they refuse to you can void
your warranty (the warranty was useless anyways) and fix them yourself. What if you want to add a new feature and the vendor refuses to add
it you can void your warranty and add that feature.
If anything releasing the source with a warranty proves that you really feel your product is stable enough/secure enough to use.
Remember, the peer review process is at the heart of the scientific method and Open Source encourages it.
Here is a good paper on the subject http://skyscraper.fortunecity.com/mondo/841/documents/99-184.html
It was written in the argument that Open Source would benefit the USAF (a critical industry/organisation).
As for those that feel Open Source projects are incapable of a firm structure and development environment take a look at Debian, Linux,
Apache, etc.
I know Debian uses a democratic environment to coordinate the development of it’s system.
Obviously this isn’t for everyone. Remember, no one forces you to use open source in your product, develop with it, or to buy from
a vendor that uses open source. Thats your own decision.
“what if it was closed? you file a bug report or request a feature, then rational soft does the job and tech support is available.
more than that: they test it and they are sure that it will work 100% now.”
In an ideal world yes. If “rational soft” really does this then I applaud them. A lot of times vendors push if off to the last minute,
are to slow to respond (if they even do in the case of the 2000 root xploit), deny it exists, or claim its a feature and tell the users
to stop complaining. I heard several vendors are trying to make illegal for security sites to
release info on holes in their software or slow down the release process.
Why? Because they’re to sloooooooowwwwww to respond, want to blow it off, or just don’t care.
Debian doesn’t bitch when someone finds a whole in their system and and publishes info on the
exploit on the web. They just fix it (and in most cases they’re not getting paid to do it).
Same follows for OpenBSD which has an impressive security record.
Simba:
“When I think if “innovative”, I think of something that has to be something more than simply a new twist on an old idea. That is what
filtering in Apache is. A new twist on an old idea. That is also what Perl is in my opinion.”
Thus anything that uses four wheels to travel on a road is not innovative (after all ox driven wagons were first and anything else
would be a twist on that idea). Ford was a lame hack right? Furthermore, anything that uses a currently existing infrastructure to operate
is not innovative. I know for a fact there are several cars that I would call innovative or that have innovative features in them that
travel on hiways just like any other car does. Your “Apache is just another webserver” and “Perl is just a rehash of Tcl, awk, etc” arguments
follow this bogus philosophy.
Apache:
Check the changelogs and compare them to other servers. Apache has been around since 1995 and
pretty much the second oldest (and most popular) webserver on the web. I’d be surprised if
they didn’t release some innovative features.
Perl:
Here is what Perl was first at according to its creator:
http://www.wall.org/~larry/pm.html (Perl the First Postmodern Language) 🙂
I love Perl what can I say? 🙂
The open patent seems rather useless to me. After all, what good is it? I can’t legally charge people to use GPL’d code,
nor can I legally prevent them from using that code in their own products provided that the products it is used in are
also GPL’d. I don’t see the point in a patent like this.
Actually read the GPL compatible license FSMLabs uses. Just because you can’t fathom how to use something doesn’t mean other
people can’t.
“Not propaganda. Just perfectly valid points about GPL software. I didn’t even say I agree with them. I just pointed them out
as perfectly valid reasons why people might not like the GPL.”
Are these view points any different when considering two separate companies competing with each other in a similar market? One
company releases a popular feature and another emulates it in some way to compete. Both companies release original
ideas to compete driving the market forward but they also clone old ideas too. Just business Simba. Why setup a double standard
saying the only companies using closed source can compete in the market while companies using open source can’t?
“Just because it is legal doesn’t mean it is ethical. The courts ruled for example that Microsoft did not do anything illegal when it borrowed ideas
from the Mac GUI for Windows. But there are a lot of people out there who still complain that Microsoft basically stole Apple’s idea. And yet
these are the same people who seem to have no problem with the fact that KDE and GNOME both stole many ideas from the Windows interface
(and the Mac interface). All I am saying here is don’t set a double standard. Don’t complain about Microsoft doing what you your yourself
are doing the very next day.”
Let me ask you a question Simba…. Why can’t a company patent the concept of a book or a wheel (well in most cases)? Because its so
obvious and commonly used that the whole idea is absurd. The same follows for most of the Windows GUI or a command line interface.
A lot of prior work has been done by Amiga, OS/2, Mac, Englebart, etc. I highly doubt anything about
the MS GUI is innovative (doesn’t stop them from trying to patent it). But in general your right it
doesn’t matter. Its nit picking.
What upsets me is that Microsoft has often shafted companies and stolen legitiment patents from them (or tried to). The GoldTouch vs. Microsoft case is
the one that always comes to my mind. Goldtouch was basically the first company to develop an ergonomic mouse (at least of its type). They patented it and contacted
Microsoft on terms of getting a driver for their mouse in the next release of Windows. Microsoft essentially used the exact same design and
released the mouse a while later as the Intellimouse Pro. GoldTouch took Microsoft to court but was essentially forced out because it couldn’t
afford the legal costs.
Microsoft raped GoldTouch and they’ll rape any other company (especially if they’re a small one) later on wether it’s the FSF or not.
GoldTouch isn’t the only one that suffered:
http://www.vcnet.com/bms/departments/innovation.shtml
Your argument in general agaist the GPL is quite different. Your saying that the GPL is bad because it provided competition to other companies, hurts them,
and should be abolished except in universities or government areas (or something along the lines).Correct?
This is non-sense. Its like me marching up to Microsoft and telling them to shut down because they provide competition to several other
companies using general ideas. Stealing patented ideas on the other hand and using unethical business practices to further your
products are a completely different ball game and partially the reason for the lawsuit MS is now facing.
As for the bitching on general ideas that will always be occurring on all sides. Get use to it.
“I think you misinterpreted my statement. What I meant by this was that a company will spend millions of dollars on R&D, develop an idea,
release a commercial product, and then some open source author will create a GPL clone of said product. Like it or not, this does hurt the
original business that spend all the money on R&D. And it borders on the unethical.”
Is a clone by an company who doesn’t use Open Source or the GPL any better? As far as I know most
things can be cloned provided 1) the source code isn’t ripped of someone else (copyright infringement)
2) the said clone isn’t replicating the behaviour of the thing it is patenting (pantent infringement)
So the solution to your problem:
Get a lawyer and patent your idea Simba (if its possible). The GPL can’t touch it until it becomes general knowledge (in 20
years providing it doesn’t get extended). However, watch out tho Microsoft will.
“This a valid point. But many would argue that with memory as abundant and cheap as it is today, it’s not an important issue. In most cases,
the ability to deploy an application in 3 months instead of 3 years far outweighs the increased memory usage now that memory is about worth it
s weight in sand.”
And it could go one be saying the outrageous memory usage and lack of speed is one of the reasons Java isn’t used in oh say most
commercial games and large time sensitive applications like a RTOS. Its just like the Energizer Bunny. It
keeps going and going……. :-p
Drats. Next time I’ll weed out those annoying CRs *L* Sorry about that.
>What if you have inhouse support?
So what you are saying is that Open Source is good for inhouse support? I agree. But should the software derived from the inhouse supported soft be open source? I guess not always. Especially if it is mission critical. But GPL forces that “always”.
If there was no GPL, only BSD, LGPL etc… this whole freedom fight would be so bad??
>And it could go one be saying the outrageous memory usage and lack of speed is one of the reasons Java isn’t used in oh say most
commercial games and large time sensitive applications like a RTOS<
I’m a Java zealot, plz don’t start
“Sorry, wrong again. Apache 2.0 will again do something no other webserver does. As I said earlier, you will be able to configure it to be fully threaded, Or, you will be able to configure it for pre-forked processes (which has some stability benefits).”
The last time I researched this, the plan was to make Apache 2.0 partially threaded, but not fully threaded. They must have changed their minds since than if your statement is correct.
But even so, you simply reenforced the point I was trying to make. You pointed out technical benefits of Apache that I was not aware of. You pointed out technical reasons why it is better than IIS. That is what open source needs to do more of. Instead, most open source developers seem to content to simply say “it’s better because it is open sourc e.” And that simply doesn’t work when you are trying to get your product into business and corporations.
“Thus anything that uses four wheels to travel on a road is not innovative (after all ox driven wagons were first and anything else would be a twist on that idea). Ford was a lame hack right?”
Well, actually if you look into this, you will find that Ford was hardly the company to produce the first automobile. Ford simply invented the assembly line and mass production and thus made it affordable for the average person.
But the comparision is invalid anyway. There are a lot of things that set the automobile apart from the horse drawn carriage. We are obviously talking about a quantitative measurement here about how similar it is to the original.
“Let me ask you a question Simba…. Why can’t a company patent the concept of a book or a wheel (well in most cases)? Because its so obvious and commonly used that the whole idea is absurd.”
Actually, you are wrong about this. A company can’t patent the book or the wheel because they can’t prove they invented it. Assume for a minute that neither the book or the wheel existed today. If some company invented it, they would definately be able to patent it.
“GoldTouch took Microsoft to court but was essentially forced out because it couldn’t afford the legal costs.”
I’m going to guess they would have lost anyway, since as Apple found out, “fashion and design” trends cannot be patented. After all, how many companies did Apple try to sue when they came out with translucent gaudy colored computers?
As far as Microsoft stealing legitmately patented ideas, that is what the courts are for. For exampe, Microsoft attempted to do this with Stack Electronics with disk compression in DOS. Stack called them on it and Microsoft ended up removing disk compression in DOS 6.21 and paying Stack a rather large some of money to put it back into DOS 6.22. Granted there is no way to no for sure, but Stack probably came out ahead on this deal. If they hadn’t sold patent rights to Microsoft, Microsoft probably would have come up with a legal (but not necessarily ethical) way to implement disk compression that used different compression algorithms than what stack was using. In this case, Stack would have gotten nothing out of the deal.
“Your argument in general agaist the GPL is quite different. Your saying that the GPL is bad because it provided competition to other companies, hurts them,
and should be abolished except in universities or government areas (or something along the lines).Correct?”
That’s not what I said. I specifically said that I don’t even necessarily agree with the ideas I mentioend about why people might not like the GPL. But they are valid reasons not to like it. I simply said if I were writing software, I would not use the GPL because I don’t want to lose that much control over my software.
“This is non-sense. Its like me marching up to Microsoft and telling them to shut down because they provide competition to several other companies using general ideas.”
The difference is that Microsoft is selling products. You are right, it would be nonsense for you to tell Microsoft to shut down for providing competition to other companies. But if Microsoft started giving away Office for example, it would not be nonsense for you to walk up to Microsoft and tell them that they have to stop giving Office away for free because they are probably violating product-dumping laws. If Microsoft were to start giving Office away for free to all comers, that would probably be illegal. And it would also be highly unethical.
In the case of GPL, it’s not illegal to give the software away for free because its not operating as a business and so it cannot be considered product dumping. But that doesn’t necessarily make it any more ethical than if a commercial company was doing it.
“So the solution to your problem:
Get a lawyer and patent your idea Simba (if its possible).”
This doesn’t really work. For example if someone can find a different way to implement the same idea, they may not be violating any patents. Like I said, Stack was probably better off licensing the technology to Microsoft. Otherwise Microsoft would have come up with different algorithms for doing disk compression, and they probably would have been perfectly legal.
Why doesn’t PNG have to pay licensing fees to Unisys (by the way, I think the GIF patent is stupid.)? After all, arent’t they doing basically the same thing? Once again, the compression algorithms and methods are different. So they get around any patent violations.
“And it could go one be saying the outrageous memory usage and lack of speed is one of the reasons Java isn’t used in oh say most
commercial games and large time sensitive applications like a RTOS. Its just like the Energizer Bunny. It
keeps going and going”
Java is not the perfect solution for everything. (I would argue that speed is a non-issue though now that we have JIT compilers and such. There are JIT compilers that can produce Java programs that are only about 5% slower than native C++ programs.
But anyway, in most cases, speed is not a major concern either since the average program spends most of its time idle and the typical CPU spends most of its time twiddling its thumbs and losing its sanity because it doesn’t have anything to do. (Of course, this is not true with game engines for example).
“So what you are saying is that Open Source is good for inhouse support? I agree. But should the software derived from the inhouse supported soft be open source? I guess not always. Especially if it is mission critical. But GPL forces that “always”.”
It also needs to be pointed out that as software becomes easier to use and such, in house support is rapidly being phased out by even major corporations in favor of out sourcing to a company that specializes in supporting software. It’s just not cost effective for them to have full time support in house because they don’t need it as often as they used to. Out sourcing is becoming much more common these days in all areas of IT.