Stallman Urges EC to Stop Oracle Acquisition of Sun(eweek.com)
eweek.com
Stallman Urges EC to Stop Oracle Acquisition of Sun
http://www.eweek.com/c/a/Database/Stallman-Urges-EC-to-Stop-Oracle-Acquisition-of-Sun-581470/
22 comments
keep in mind that Torvalds unilaterally declared that applications that run on top of Linux are exempt from GPL.
That text is just there to allay FUD, it's not a necessary part of the license in any way.
This is in effect LGPL, not GPL.
No it is not, not in any way, shape, or form. If you distribute something that dynamically links with the kernel, the GPL applies to you in full force.
To the kernel, applications running in userspace are data. They may link with libc to make syscalls, but that just means they throw interrupts that the kernel catches. Your userspace programs aren't derivative works of the kernel any more than your documents are derivative works of the last application you used to edit them.
That text is just there to allay FUD, it's not a necessary part of the license in any way.
This is in effect LGPL, not GPL.
No it is not, not in any way, shape, or form. If you distribute something that dynamically links with the kernel, the GPL applies to you in full force.
To the kernel, applications running in userspace are data. They may link with libc to make syscalls, but that just means they throw interrupts that the kernel catches. Your userspace programs aren't derivative works of the kernel any more than your documents are derivative works of the last application you used to edit them.
I see stuff like this and I wonder about two things:
1) there is another open source (BSD) database - postgresql
2) why should it matter, since the whole purpose of the GPL is to keep Oracle from "destroying" MySQL. Didn't the GPL protect the users in this case?
Here's the letter to the EC written by Stallman et al the OP is referring to: http://keionline.org/ec-mysql . Shocking of eWeek to leave out the link.
The key paragraphs imo are:
"As only the original rights holder can sell commercial licenses, no new forked version of the code will have the ability to practice the parallel licensing approach, and will not easily generate the resources to support continued development of the MySQL platform.
The acquisition of MySQL by Oracle will be a major setback to the development of a FLOSS database platform, potentially alienating and dispersing MySQL's core community of developers [...]"
No license can fully protect users from developer inactivity or not knowing where to go when they have problems. In this case the problem is particularly acute because MySQL AB has been known as the go-to place for a long time, and this may no longer be viable in the event of Oracle's acquisition.
EDIT: Simon Phipps (Sun) wrote an eloquent piece about the dual licensing aspect: http://www.webmink.net/2009/10/remarkable-reversal.htm
The key paragraphs imo are:
"As only the original rights holder can sell commercial licenses, no new forked version of the code will have the ability to practice the parallel licensing approach, and will not easily generate the resources to support continued development of the MySQL platform.
The acquisition of MySQL by Oracle will be a major setback to the development of a FLOSS database platform, potentially alienating and dispersing MySQL's core community of developers [...]"
No license can fully protect users from developer inactivity or not knowing where to go when they have problems. In this case the problem is particularly acute because MySQL AB has been known as the go-to place for a long time, and this may no longer be viable in the event of Oracle's acquisition.
EDIT: Simon Phipps (Sun) wrote an eloquent piece about the dual licensing aspect: http://www.webmink.net/2009/10/remarkable-reversal.htm
I am in awe of the thinking that a commercial option is needed for MySQL and that Stallman agrees with that. It just seems so counter to all the comments on the GPL.
I don't really see the support argument, a lot of software gets orphaned. Companies need to have plans for this kind of stuff. They are better off here because they might be able to hire people to fix compatibility issues until they can migrate to another platform.
I think "major setback to the development of a FLOSS database platform" is kinda bs given the existence of postgresql.
I don't really see the support argument, a lot of software gets orphaned. Companies need to have plans for this kind of stuff. They are better off here because they might be able to hire people to fix compatibility issues until they can migrate to another platform.
I think "major setback to the development of a FLOSS database platform" is kinda bs given the existence of postgresql.
I completely agree with your first paragraph. It is amazing to see Stallman, the very founder and long time advocate of "Free" software and the GPL, admitting that a closed and open approach to software is needed. That seems very different than what he has always preached.
Then again, perhaps I have not listened to his speech in enough depth.
Then again, perhaps I have not listened to his speech in enough depth.
Regardless of whether there is a 'commercial' option for the source code itself, there is plenty going on with GPL-based development and an alternative for InnoDB led by Monty Widenius (Monty AB's development of the Maria Engine should lessen the concern due to Oracle's purchase of InnoBase and their patent ownership / 'original author' status). It's already been discussed at length, so I'll just say I'm not convinced it's in Oracle's best interest to try to kill off InnoDB in order to decrease consumer options for MySQL. I guess we'll see what they do if the EC approves.
But the market for 'commercial services' around MySQL is doing just fine -- Percona and MontyDB being the most obvious (more here: http://lwn.net/Articles/329626/). Not an endorsement, but our team decided to start using Percona and it's been OK; over the last couple years we didn't end up using MySQL AB's support that much anyway.
But the market for 'commercial services' around MySQL is doing just fine -- Percona and MontyDB being the most obvious (more here: http://lwn.net/Articles/329626/). Not an endorsement, but our team decided to start using Percona and it's been OK; over the last couple years we didn't end up using MySQL AB's support that much anyway.
Where is Richard Stallman and what did they do to him?
Now, on a more serious tone, a program dies when nobody uses or is interested on it. MySQL will inevitably die someday, replaced by something better, much like MySQL itself replaced mSQL. The GPL will make its source available forever, protecting users who may be interested in continuing it.
As for PostgreSQL, it saddens me to say its BSD licensing provides no such protection. If people lose interest in it, it will probably evolve some more in the form of EnterpriseDB and the source may be lost or locked up forever.
Isn't Drizzle a worthy successor to MySQL?
Now, on a more serious tone, a program dies when nobody uses or is interested on it. MySQL will inevitably die someday, replaced by something better, much like MySQL itself replaced mSQL. The GPL will make its source available forever, protecting users who may be interested in continuing it.
As for PostgreSQL, it saddens me to say its BSD licensing provides no such protection. If people lose interest in it, it will probably evolve some more in the form of EnterpriseDB and the source may be lost or locked up forever.
Isn't Drizzle a worthy successor to MySQL?
> program dies when nobody uses or is interested on it ... The GPL will make its source available forever ... As for PostgreSQL, it saddens me to say its BSD licensing provides no such protection
A software application also dies when you can't use it for your purposes anymore, even when you're willing to pay-up. This letter from Stallman actually highlights the hypocrisy behind GPL (not LGPL, just GPL).
PostgreSQL's source code will never be lost, and it saddens me that many people perpetuate such a myth.
If people lose interest in it, there's always going to be an archive out there with the latest released version. And I find it hard to believe that people would prefer EnterpriseDB over PostgreSQL since the BSD license is what made it shine in the first place.
MySql and PostgreSql are actual threats to Oracle's profit margins. Guess which one they picked to kill.
A software application also dies when you can't use it for your purposes anymore, even when you're willing to pay-up. This letter from Stallman actually highlights the hypocrisy behind GPL (not LGPL, just GPL).
PostgreSQL's source code will never be lost, and it saddens me that many people perpetuate such a myth.
If people lose interest in it, there's always going to be an archive out there with the latest released version. And I find it hard to believe that people would prefer EnterpriseDB over PostgreSQL since the BSD license is what made it shine in the first place.
MySql and PostgreSql are actual threats to Oracle's profit margins. Guess which one they picked to kill.
"MySql and PostgreSql are actual threats to Oracle's profit margins. Guess which one they picked to kill."
Interestingly, thanks to the GPL, they can't kill MySQL as long as someone cares to maintain it.
Interestingly, thanks to the GPL, they can't kill MySQL as long as someone cares to maintain it.
The dead and forgotten code issue will affect BSD or GPL in the same way. When people lose interest, it will disappear except for maybe some archives. I would say the BSD license might last longer given that a company might pay more attention to it since it can live in their code.
The GPL provides no protection from disinterest and (given the article) ticked off companies. If MySQL were BSD, it wouldn't really matter what Oracle did, there would be no commercial license problems and it could thrive in a different home more easily. Drizzle is still GPL with no possibility of a commercial license unless they rewrite all the code. https://answers.launchpad.net/drizzle/+faq/135
The GPL provides no protection from disinterest and (given the article) ticked off companies. If MySQL were BSD, it wouldn't really matter what Oracle did, there would be no commercial license problems and it could thrive in a different home more easily. Drizzle is still GPL with no possibility of a commercial license unless they rewrite all the code. https://answers.launchpad.net/drizzle/+faq/135
But if no one pays attention to a given piece of code, is it really worth to maintain it?
Truthfully, as much as I respect the work, no. I don't think code is like literature where it might be relevant and understood years after creation. The best that can be hoped for is that the hidden gems live on in other code.
I'm sure if Oracle abandons MySQL then after some initial confusion developers and users will settle on 1 or 2 forks and continue from there. Some are already forking MySQL to scratch their own itch.
I guess the main people it hurts is commercial software developers that paid a license fee to not have to open source their code. None of the forks would be able to do the licenses.
I'm confused as to why the European Commission is involved with this at all. Anyone cares to explain?
Also, "Web 2.0 companies such as Google, Yahoo and Amazon.com". Good picks for web2.0 companies… :)
Also, "Web 2.0 companies such as Google, Yahoo and Amazon.com". Good picks for web2.0 companies… :)
Because the EC is a market in which Oracle intends to make a substantial portion of its income ?
Anti-trust actually has some meaning here, it stands for not being able to own a substantial portion of the market and then being able to gobble up your competitors (directly, or, as in this case indirectly).
American companies invariably explain this is a lack of 'free market', but it is one of the few cases where I think that some government involvement in business is actually a good thing.
The EC got to sharpen it's knives in this respect on a very unwilling telco industry, and they are fairly sharp when it comes to stating how far is too far.
Whether Oracle has crossed the line or not in this respect is hard to tell, IBM still has Informix/DB2, there are other open source alternatives but I think a fork of MySQL is another way of resolving this.
But then there would have to be a pre-sun style company behind it to give it legs, otherwise corporations would never go for it.
Anti-trust actually has some meaning here, it stands for not being able to own a substantial portion of the market and then being able to gobble up your competitors (directly, or, as in this case indirectly).
American companies invariably explain this is a lack of 'free market', but it is one of the few cases where I think that some government involvement in business is actually a good thing.
The EC got to sharpen it's knives in this respect on a very unwilling telco industry, and they are fairly sharp when it comes to stating how far is too far.
Whether Oracle has crossed the line or not in this respect is hard to tell, IBM still has Informix/DB2, there are other open source alternatives but I think a fork of MySQL is another way of resolving this.
But then there would have to be a pre-sun style company behind it to give it legs, otherwise corporations would never go for it.
From the Open Source perspective there is Postgres and the mysql code
On the commercial side there is IBM with DB2/Informix and Microsoft with SQL Server
This doesn't seem like directly gobbling up your competitors..
On the commercial side there is IBM with DB2/Informix and Microsoft with SQL Server
This doesn't seem like directly gobbling up your competitors..
Oracle, postgres and mysql are not in the same space. That leaves informix/DB2.
Not being in the same space is maybe not enough for the anti-trust people though, to them 'databases' are 'databases'.
Not being in the same space is maybe not enough for the anti-trust people though, to them 'databases' are 'databases'.
And I present to you our version of Dr. Robert Stadler
Impressions to take with a pinch of salt as it has been written after a long day:
when reading previous comments, the word "inconsistancy" comes to my mind. Those comments seem to be written in a unpationated way. Their authors seems rather for free licencing... but STILL!! In my understanding, they may be ultra summarized in "tss, what this guy(stallman) is doing is unsuitable to reach the free licencing goal". AKA he is inconsistant.
I find it rather rare to get this kind of reaction. Usually when fsf people, it is quite pationated, up to religious.
PS: note the humor "free licencing" term to avoid the ridiculous "FOSS" acronym
when reading previous comments, the word "inconsistancy" comes to my mind. Those comments seem to be written in a unpationated way. Their authors seems rather for free licencing... but STILL!! In my understanding, they may be ultra summarized in "tss, what this guy(stallman) is doing is unsuitable to reach the free licencing goal". AKA he is inconsistant.
I find it rather rare to get this kind of reaction. Usually when fsf people, it is quite pationated, up to religious.
PS: note the humor "free licencing" term to avoid the ridiculous "FOSS" acronym
I think Stallman has always been something of a pragmatist, in the sense of preferring widespread FOSS adoption even at the expense of some ideological purity. That's why the LGPL exists, after all.
(NB: I think dispassionate/passionate are the words you're looking for.)
(NB: I think dispassionate/passionate are the words you're looking for.)
The time has come for a reevaluation of what constitutes real open source. I believe GPL should be excluded from it, and the MySQL business model is the classic example.
To those who answer "Linux", keep in mind that Torvalds unilaterally declared that applications that run on top of Linux (calling Linux via the standard libraries) are exempt from GPL. This is in effect LGPL, not GPL.
MySQL specifically takes the opposite tack: any application that uses the database using standard libraries comes under GPL, for distribution purposes.
LGPL and GPL are very different animals.