OpenJDK Governing Board Minutes: 20011/4/21
Dr Andrew John Hughes
gnu_andrew at member.fsf.org
Mon May 9 11:55:06 UTC 2011
On 9 May 2011 08:32, Fernando Cassia <fcassia at gmail.com> wrote:
> On Sun, May 8, 2011 at 7:05 PM, Dr Andrew John Hughes
> <gnu_andrew at member.fsf.org> wrote:
>> I also personally wouldn't contribute to non-copyleft projects either,
>> something I already
>> made clear years ago when Apache Harmony started, so whatever you are trying to
>> imply with this comparison is flawed.
>
> I don´t think my comparison is flawed.
>
> Mozilla (the Mozilla Suite) was originally -and for a number of years-
> released under a triple license: MPL/GPL/LGPL. This allowed Netscape
> Corp to build a commercial browser (Netscape 6.0 to 7.2) on top of the
> open source code, and package it with propietary code if they so
> wanted (commercial spellchecker, or the AIM sidebar tab written in
> XUL, etc).
>
> So if you contributed to Mozilla.org, your code would not only end up
> in the Mozilla 1.x browser suite, but also on Netscape 6.x - 7.x
>
> Now, for some reason you say this approach irks you, and that you want
> no piece of your contributions ending up in a commercial product with
> a different license.
>
> This is what I don´t understand. I never saw any Mozilla.org developer
> complaining that their work would end up being part of the Netscape
> 6.x / 7.x browser. And that was my comparison.
>
Yes, and I agree with and follow all that. But I don't see how it has
any relation to what I was saying about my own personal choices in
contributing to FOSS projects. Sure, Mozilla developers have made
their own decision about what they are happy to have happen to their
code. Just as I've made mine. What exactly was your point?
> you said "Making it part of OpenJDK under the OCA also means
> contributing to Oracle's proprietary products and this is why I
> personally would not make any
> significant contribution of work (as in complete new features like
> Mario mentions, rather than fixes) to OpenJDK."
>
> Oracle is doing the same Mozilla did for years. Trying to build a
> commercial product AND an open source project, both at the same time.
> Why can´t they? After all, they´re the one putting incredible
> resources (bandwidth, servers, and manpower, aka programmers on
> Oracle´s payroll) to advance the project. Why can´t they set the rules
> they see fit (like Mozilla) to have an open source project and
> release, and a commercial product with a different license at the same
> time?.
>
They can set whatever rules they wish. It doesn't mean anyone else
has to contribute to or agree with it. You only have to look at the
existing low contribution rate and the fact that there is a bunch of
stuff in IcedTea which can't go into OpenJDK to see how successful the
current system is.
They are not trying to "build a commercial product AND an open source
project". They provide proprietary binaries gratis and are attempting
(and failing IMHO) to run a FOSS project. I have no problem
contributing to a FOSS project but I do not see why my contributions
should be used to enhance Oracle's proprietary product. It's not so
much about multi-licensing as about why I should give Oracle
non-reciprocal rights to my work. There is no benefit for the
contributor in doing this.
I don't buy the resource argument, because it's simply not true that
Oracle are "putting incredible resources... [in] to advanc[ing] the
project". They have a lot of resources for working on Oracle's JDK
product, no doubt, and they have clear motivations for doing that.
Due to Sun's choice to release a significant proportion of this code
under the GPL, work on it is now more visible than it was before, but
advancing a FOSS project is not the main motivation for this. Just
look at OpenJDK6 if you want to see how much Oracle contribute when
there is no corresponding proprietary product to motivate their work.
They have one engineer reviewing patches and a few engineers do bother
to backport changesets to it. But the majority of backporting is done
by Red Hat employees. This process is inefficient because, as we
don't write the original changesets, we have to try and spot
appropriate fixes going into OpenJDK 7 and make decisions as to
whether to backport them. If Oracle were truly as focused on the
success of OpenJDK itself as you claim, considering backporting to
OpenJDK6 would be a standard part of committing new patches,
especially when Oracle employees know which patches have come from the
proprietary JDK6 project but we don't.
> Please correct me if I´m wrong but you seem to be on a crusade against
> multi-licensing, which is at the heart of many open source projects,
> including QT, and asterisk, just to name two.
>
Please don't try and generalise my comments. I'm talking specifically
about OpenJDK, not every FOSS project in existence.
> Regards,
> FC
>
--
Andrew :-)
Support Free Java!
Contribute to GNU Classpath and IcedTea
http://www.gnu.org/software/classpath
http://icedtea.classpath.org
PGP Key: F5862A37 (https://keys.indymedia.org/)
Fingerprint = EA30 D855 D50F 90CD F54D 0698 0713 C3ED F586 2A37
More information about the discuss
mailing list