apache quitting JCP?
phil swenson
implementers of Java specifications are not upheld by the JCP
Executive Committee to the limits of the EC's ability. The lack of
active, strong and clear enforcement of those rights implies that the
JSPA agreements are worthless, confirming that JCP specifications are
nothing more than proprietary documentation."
https://blogs.apache.org/foundation/entry/statement_by_the_asf_board1
what are the implications of this (if any)?
Kirk
However, there are still a few outstanding questions and Apache is one of them. I don't have an answer except to say that it appears that Apache instead of trying to work with Oracle and engage with them has decided to threaten to throw the toys out of the pram. I would be a shame if they did this however, the JCP would go on though it would be weaker without that strong open voice. That said, there are some radical elements in the ASF that seem to want to bash Oracle (Sun) for the sake of bashing them.
My conversation with those in Apache that I know was to ask the question, has someone broken a legal agreement to which the answer from ASF's POV is yes. But in reality, the question is answered with OpenJDK licensing. They can fork OpenJDK and they are clear. Yet the refuse due to some ideological position that was encouraged by IBM (in their battle against Sun for things they didn't like). Another point, if someone broke a legal agreement there are legal remedies. Yet no one at ASF will stand up and say why after I don't know how many years of whining about the problem, they haven't used any of those legal remedies. And as much as I may or may not agree with them, the whining is getting really really old. I won't name names but some have been posting a lot of FUD about various things that Oracle is planning on doing. To be fair, I don't completely know what Oracle is planning on doing with Java but then neither do they and until just recently, their blogs about Oracle have all been speculation and conjecture and all based on dubious facts and in some case fiction.
So, here are some facts,
1) Oracle is hiring people to work on the JDK so Java is getting more resources than it ever had
2) They are putting everything into the OpenJDK and it will remain there
3) Oracle is working hard to deliver 7.0 on time, something that Sun wasn't going to be able to do.
4) Oracle will have to deal with Sun promises that were impossible to deliver on.
5) Oracle is a corporate that has legal obligations which is takes seriously which some times prevent it from speaking
6) Oracle has a corporate culture (some what like Apple's) where all PR is carefully managed through a PR department. They are slowly learning that this doesn't work so well when dealing with community but they are learning.
Pure speculation on my part. Oracle will announce in the near future that they will pick up development of the JDK for Mac OSX. My basis for the speculation is that Oracle doesn't want to give any opportunity for MS to claim a better cross platform story than Java has.
Look to Devoxx in a few days where Mark Rienhold will be speaking on a panel about the future of Java. Stephan has put a few interesting characters on that panel so I expect it will be fun as well as informative.
Regards,
Kirk
> --
> You received this message because you are subscribed to the Google Groups "The Java Posse" group.
> To post to this group, send email to java...@googlegroups.com.
> To unsubscribe from this group, send email to javaposse+...@googlegroups.com.
> For more options, visit this group at http://groups.google.com/group/javaposse?hl=en.
>
Fabrizio Giudici
> Well, from my perspective, ASF is making a mistake.
(the NetBeans community) I can say that the communications, which
weren't exceptional four years ago, constantly improved ilife n the last
part of Sun and Oracle didn't step backward (with the exception of some
general policies). I think that there is some good will and it's still
worth while talking.
Clearly, Apache has got many reasons to be angry, with Sun first and now
with Oracle. From the community point of view, the lack of an alternate,
ASF-based implementation such as Harmony is a missed opportunity. Given
that, I think it's also clear that Harmony wasn't really driven by the
community, as people started the RIP mourning five minutes after IBM
withdrew; so in the end IBM was relevant. I've said many times, that I
don't think it's feasible a complete JVM totally maintained by the
community and the Harmony presumed ending sort of demonstrated it (yes,
I know, the ASF said they will still work on it, but I'm very skeptical;
let's have a recap point six months from here).
Sure, Oracle probably played some Machavelli politics with the latest
JCP elections, but the system on the whole worked as the alleged puppet
candidate was not elected; furthermore, Apache received 95% of votes and
I'd see it as a sort of betrayal if they quit.
Given that, I'm not against wars in principle. It depends on the
strength of your weapons and army. Frankly I don't see our advantage in
declaring war to Oracle, and just minor troubles for them. It would be a
pity.
> Look to Devoxx in a few days where Mark Rienhold will be speaking on a panel about the future of Java. Stephan has put a few interesting characters on that panel so I expect it will be fun as well as informative.
Too bad I won't be there... :-( Looking forward to read some blogs.
--
Fabrizio Giudici - Java Architect, Project Manager
Tidalwave s.a.s. - "We make Java work. Everywhere."
java.net/blog/fabriziogiudici - www.tidalwave.it/people
Fabrizio...@tidalwave.it
Casper Bang
is done bending over. What else can they do, they have limited options
here against large corporations and their ulterior motives.
Chris Adamson
> what are the implications of this (if any)?
with it.
I've long since tired of ASF's obstructionism, and this election was
the second time I've voted against their re-ratification. If you
think Sun screwed them over, fine, but what good has come from the "no
on all Sun JSRs" position? Enough drama already.
Casper Bang
> with it.
either Java is an open standard, or it's not! And if it's not, then
Java's greatest asset (being an open standard) falls under the
category of false marketing and Oracle should be held accountable. If
Apache were a large corporation with lots of cash, they could sue...
but unfortunately in the US, money and lawyers are the name of the
game. Bring on the drama, enough corporate bullying.
Fabrizio Giudici
>> A round of applause from my little home office if they follow through
>> with it.
> Yeah you are making that very clear here and via Twitter. Listen,
> either Java is an open standard, or it's not! And if it's not, then
> Java's greatest asset (being an open standard) falls under the
> category of false marketing and Oracle should be held accountable. If
> Apache were a large corporation with lots of cash, they could sue...
licensing scheme that implies legal protection. Thus, I assume that they
are perfectly equipped for lawsuits (and since the question arose from
the old Sun times, they could sue Sun in a moment of weakness). Thus, I
completely agree with Chris that there were no legal basis for anything.
Now, I've just read at DZone that they're menacing a vote against Java
7. Now I declare myself in the bag of people that got tired of this.
This point is lost, let's move on to next one. We're been complaining
with Sun for years because of the stall, and now that we have a roadmap
we're going to boycott it? Crazy.
Sure, Java is not an open standard. In the perfect world it would be,
but a working, vibrant open technology that is not an open standard is
better than nothing.
Stuart McCulloch
On 11/09/2010 10:35 PM, Casper Bang wrote:I don't take this. Apache, like FSF, is the inventor and promoter of a licensing scheme that implies legal protection. Thus, I assume that they are perfectly equipped for lawsuits (and since the question arose from the old Sun times, they could sue Sun in a moment of weakness).
A round of applause from my little home office if they follow throughYeah you are making that very clear here and via Twitter. Listen,
with it.
either Java is an open standard, or it's not! And if it's not, then
Java's greatest asset (being an open standard) falls under the
category of false marketing and Oracle should be held accountable. If
Apache were a large corporation with lots of cash, they could sue...
Thus, I completely agree with Chris that there were no legal basis for anything. Now, I've just read at DZone that they're menacing a vote against Java 7. Now I declare myself in the bag of people that got tired of this. This point is lost, let's move on to next one. We're been complaining with Sun for years because of the stall, and now that we have a roadmap we're going to boycott it? Crazy.
Sure, Java is not an open standard. In the perfect world it would be, but a working, vibrant open technology that is not an open standard is better than nothing.
--
Fabrizio Giudici - Java Architect, Project Manager
Tidalwave s.a.s. - "We make Java work. Everywhere."
java.net/blog/fabriziogiudici - www.tidalwave.it/people
Fabrizio...@tidalwave.it
--
You received this message because you are subscribed to the Google Groups "The Java Posse" group.
To post to this group, send email to java...@googlegroups.com.
To unsubscribe from this group, send email to javaposse+...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/javaposse?hl=en.
--
Cheers, Stuart
Casper Bang
> but a working, vibrant open technology that is not an open standard is
> better than nothing.
long time - it took Google to throw some new logs on the fire yet
Oracle is standing with a fire hose. I admire your optimism, that we
can just humour the firemen and enjoy the pretty fire, but I remain
unconvinced that Apache's "obstructionism" is to blame for the sorry
state of Java... Sun handled that all by themselves. On the contrary,
I feel obliged to point out just what Apache has done for the Java
community: http://projects.apache.org/indexes/alpha.html
Oracle can just grant them a license, as to their contractual
agreement, and things roll again. How hard is that?
Chris Adamson
> I admire your optimism, that we
> can just humour the firemen and enjoy the pretty fire, but I remain
> unconvinced that Apache's "obstructionism" is to blame for the sorry
> state of Java... Sun handled that all by themselves.
not say that ASF's obstructionism is responsible for the "sorry state
of Java" (and I didn't say _that_ either… on the server and on
Android, Java is doing great). Frankly, I don't think the ASF's JCP-
related actions have had a substantial impact at all, except to
discredit them in the minds of a small minority of developers, myself
included.
-Chris
Greg Reddin
> it appears that Apache instead of trying to work with Oracle and engage with them has decided to threaten to throw the toys out of the pram.
I would submit that Apache has been trying to work with Oracle and Sun
for several years and the process has gained nothing. Maybe it's time
for a different strategy.
> That said, there are some radical elements in the ASF that seem to want to bash Oracle (Sun) for the sake of bashing them.
If you're going to make an accusation like that, you might want to
back it up with some facts.
> My conversation with those in Apache that I know was to ask the question, has someone broken a legal agreement to which the answer from ASF's POV is yes. But in reality, the question is answered with OpenJDK licensing. They can fork OpenJDK and they are clear.
No they can't. Apache cannot fork OpenJDK and release it under the
Apache License. If we can't release it under the Apache License, what
would be the point?
> Yet the refuse due to some ideological position that was encouraged by IBM (in their battle against Sun for things they didn't like).
Ideological position? Encouraged by IBM? So, Apache's desire to
release software that has no usage restrictions is an ideological
position? If so, then it's a good ideological position to have, IMO. I
don't think you can say Apache's position is unduly influenced by IBM.
> Another point, if someone broke a legal agreement there are legal remedies. Yet no one at ASF will stand up and say why after I don't know how many years of whining about the problem, they haven't used any of those legal remedies.
Right. We could risk the solvency of our entire foundation by suing a
multi-national corporation with tons of legal muscle. The ASF has so
far decided that this is not a course of action we should take.
Greg
Greg Reddin
> I've long since tired of ASF's obstructionism, and this election was
> the second time I've voted against their re-ratification. If you
> think Sun screwed them over, fine, but what good has come from the "no
> on all Sun JSRs" position? Enough drama already.
The only recourse Apache has within the JCP is to use EC and JSR votes
to bring about change. The foundation's position has been that a spec
lead who proves they are willing to break the JSPA agreement cannot be
trusted. We can't implement those specs in good faith knowing the
project's users will be free to use the software. I'm not sure what
other course of action there is within the process.
Greg
Greg Reddin
<fabrizio...@tidalwave.it> wrote:
> I don't take this. Apache, like FSF, is the inventor and promoter of a
> licensing scheme that implies legal protection. Thus, I assume that they are
> perfectly equipped for lawsuits (and since the question arose from the old
> Sun times, they could sue Sun in a moment of weakness). Thus, I completely
> agree with Chris that there were no legal basis for anything.
Whether there is a legal basis and whether a lawsuit is pragmatically
a good idea are two completely different questions.
> Now, I've just
> read at DZone that they're menacing a vote against Java 7. Now I declare
> myself in the bag of people that got tired of this. This point is lost,
> let's move on to next one. We're been complaining with Sun for years because
> of the stall, and now that we have a roadmap we're going to boycott it?
So they gave us a new roadmap, which is the same as the old roadmap.
Actually no. For Apache, the new roadmap is a dead end. So I guess we
should just move on. I'm not sure what you'd have Apache do: stay in
the JCP even though it's proven itself to be unwilling to enforce its
own agreements?
> Sure, Java is not an open standard.
I think this is the reality we are facing. It's unclear to me what
that means for my personal interests in the Java community.
Greg
Kirk
On Nov 10, 2010, at 12:13 AM, Greg Reddin wrote:
> On Tue, Nov 9, 2010 at 1:12 PM, Kirk <kirk.pe...@gmail.com> wrote:
>> it appears that Apache instead of trying to work with Oracle and engage with them has decided to threaten to throw the toys out of the pram.
>
> I would submit that Apache has been trying to work with Oracle and Sun
> for several years and the process has gained nothing. Maybe it's time
> for a different strategy.
Trying to work with or just crying foul if ASF doesn't get everything it's way?
>
>> That said, there are some radical elements in the ASF that seem to want to bash Oracle (Sun) for the sake of bashing them.
>
> If you're going to make an accusation like that, you might want to
> back it up with some facts.
It's my opinion that some people have continued to blog FUD even when clear statements of intent have been made. I see no rational reason for doing so unless one was just interested in creating more FUD. Others, not myself, have called them freetards..
>
>> My conversation with those in Apache that I know was to ask the question, has someone broken a legal agreement to which the answer from ASF's POV is yes. But in reality, the question is answered with OpenJDK licensing. They can fork OpenJDK and they are clear.
>
> No they can't. Apache cannot fork OpenJDK and release it under the
> Apache License. If we can't release it under the Apache License, what
> would be the point?
Well, it would be forked and you could do what you wanted and run a tck and not face field of use restrictions.
>
>> Yet the refuse due to some ideological position that was encouraged by IBM (in their battle against Sun for things they didn't like).
>
> Ideological position? Encouraged by IBM? So, Apache's desire to
> release software that has no usage restrictions is an ideological
> position? If so, then it's a good ideological position to have, IMO. I
> don't think you can say Apache's position is unduly influenced by IBM.
If ideology blocks perfectly acceptable solutions.... And the bit about IBM is from conversations with people inside IBM.
>
>> Another point, if someone broke a legal agreement there are legal remedies. Yet no one at ASF will stand up and say why after I don't know how many years of whining about the problem, they haven't used any of those legal remedies.
>
> Right. We could risk the solvency of our entire foundation by suing a
> multi-national corporation with tons of legal muscle. The ASF has so
> far decided that this is not a course of action we should take.
Justice for the rich and not the right.. yeah?
Regards,
Kirk
Michael Neale
oracle says. Think of it like an inflammed wound - when you have
inflammation - small irritations become quite painful - at least until
you heal.
So we need healing...
when I get that feeling I need... wait what?
mP
for it) for the TCK so Apache could get harmony verified and so on ?
We all know that Sun refused to give the TCK for free, but what if IBM
had paid the fee and so on. Did this ever happen ? - im not sure i
have ever heard that particular news event...
Casper Bang
> not say that ASF's obstructionism is responsible for the "sorry state
> of Java" (and I didn't say _that_ either… on the server and on
> Android, Java is doing great).
open technology" we shouldn't throw the baby out with the bathwater.
That Java is currently in a bit of a sorry state were my own words
inferred from all the recent bad news, and a general acceptance that
Java is failing to provide assistance to developers fighting
increasingly complex applications. It might be doing OK on the server
so far, but it's getting heavy competition and is generally on the
decline [http://goo.gl/KASe7] - in spite of the significant focus
given to it by Android.
I guess I just wish you'd argue for why we should allow Java to be
pushed as an open standard when clearly it's not. You're doomed if you
abide by the 3 existing official profiles (Apache Harmony) and you're
doomed if you step out of bounds of them (Android). How do you not see
that as a problem?
Vince O'Sullivan
> Others, not myself, have called them freetards..
clean. If this a measure of the Oracle position and what Apache are
facing in the JCP then I'm not surprised they've had enough and walked
out.
Fabrizio Giudici
>> Please be careful not to put words in my mouth, or Fabrizio's. I did
>> not say that ASF's obstructionism is responsible for the "sorry state
Well, the point is: Apache was right in that try, but _they lose_. My
point is not that one should not oppose to Sun/Oracle policies, but that
one should do in a constructive way. What did ASF get in a few years of
obstructionism with Sun? Nothing. And what do you think they're going to
get now, with a stronger Oracle and furthermore backed up by IBM? Only
noise. I agree on insisting on a point, as far as that point can be
still saved. Not when it has been lost and it would be better to focus
energy on something else. If they sued Oracle because of the alleged law
infringement when Sun did not release an unconditioned TCK, I'll back
them. But I don't see the point in making more difficult the release of
Java 7.
Fabrizio Giudici
>
>
> bogus logic - there are people who invent and promote defence systems,
> but that doesn't mean they're personally ready to declare war on someone
>
> I don't take this. Apache, like FSF, is the inventor and promoter of a
> licensing scheme that implies legal protection. Thus, I assume that
> they are perfectly equipped for lawsuits (and since the question arose
> from the old Sun times, they could sue Sun in a moment of weakness).
ASF. It's a bunch of people that, beyond producing code, also design
licenses and promotes them. Ok? Now, designing a license is a matter of
lawyers, not engineers. Any software foundation that proposed a license
model and wasn't able to defend it in court would be just laughable. FSF
and ASF are not.
Fabrizio Giudici
>
> should just move on. I'm not sure what you'd have Apache do: stay in
> the JCP even though it's proven itself to be unwilling to enforce its
> own agreements?
not.
>> Sure, Java is not an open standard.
> I think this is the reality we are facing. It's unclear to me what
> that means for my personal interests in the Java community.
1. It has never been, and we're still here. So, it's bad, but not
totally bad.
2. Is it .Net an open standard? .Net is the most direct competitor.
3. Alternatively, one should move to the completely community based
platforms such as Python, etc...
Casper Bang
exactly one vote and no veto right (unlike Oracle in the JCP).
Microsoft's implementation is however not open source, while other
implementations are. Given that you can write applications in C# and
deploy to iOS and Android, this standardization is clearly not just
theoretical of nature.
> 3. Alternatively, one should move to the completely community based
> platforms such as Python, etc...
starting to believe they are on to something.
JodaStephen
wrote:
> point is not that one should not oppose to Sun/Oracle policies, but that
> one should do in a constructive way. What did ASF get in a few years of
> obstructionism with Sun?
issue both before and after the open letter. Some of that is private
which is why it can seem like obstructionism for no benefit. As such,
you really only have my word that the approach taken did have the
opportunity to suceed.
> Nothing. And what do you think they're going to
> get now, with a stronger Oracle and furthermore backed up by IBM? Only
> noise. I agree on insisting on a point, as far as that point can be
> still saved. Not when it has been lost and it would be better to focus
> energy on something else.
progress can be made, and the Java SE 7 vote is the marker point to
them moving on (leaving the JCP).
Feel free to disagree with the choice of leaving the JCP - I
personally encouraged a more moderate position.
> But I don't see the point in making more difficult the release of
> Java 7.
sadly. Some thoughts:
(1) Most JSRs only need 50% support, platform JSRs need 66%, so its an
easier pressure point.
(2) The platform JSR is a pressure point that has been used before.
(3) Every other member of the JCP EC (except Sun) signed up to the
ASF's position and agreed to use Java SE 7 JSR as the pressure point.
(4) Java SE 7 is indirectly connected to Java SE 5, the JSR in dispute
(as they are both Java SE)
Feel free to have an opinion on the validity of any of these. Its
mostly just a suitable line in the sand (agreed by everyone in the JCP
EC).
Sun could have tried to force through the Java SE 7 years ago. But
they chose not to, probably because they weren't strong enough. Now
IBM has changed its opinion, the game is different. Unless the earth
moves in the next couple of weeks, Java SE 7 will pass and the new
world order starts.
Stephen
Stuart McCulloch
On 11/09/2010 10:58 PM, Stuart McCulloch wrote:You seem not to understand what is a Software Foundation, such as FSF or ASF. It's a bunch of people that, beyond producing code, also design licenses and promotes them. Ok? Now, designing a license is a matter of lawyers, not engineers. Any software foundation that proposed a license model and wasn't able to defend it in court would be just laughable. FSF and ASF are not.
bogus logic - there are people who invent and promote defence systems, but that doesn't mean they're personally ready to declare war on someone
I don't take this. Apache, like FSF, is the inventor and promoter of a licensing scheme that implies legal protection. Thus, I assume that they are perfectly equipped for lawsuits (and since the question arose from the old Sun times, they could sue Sun in a moment of weakness).
--
Fabrizio Giudici - Java Architect, Project Manager
Tidalwave s.a.s. - "We make Java work. Everywhere."
java.net/blog/fabriziogiudici - www.tidalwave.it/people
Fabrizio...@tidalwave.it
--
You received this message because you are subscribed to the Google Groups "The Java Posse" group.
To post to this group, send email to java...@googlegroups.com.
To unsubscribe from this group, send email to javaposse+...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/javaposse?hl=en.
--
Cheers, Stuart
Fabrizio Giudici
>> 2. Is it .Net an open standard? .Net is the most direct competitor.
> Yes it is, first under ECMA and then ISO/IEC - where Microsoft has
> exactly one vote and no veto right (unlike Oracle in the JCP).
> Microsoft's implementation is however not open source, while other
> implementations are. Given that you can write applications in C# and
> deploy to iOS and Android, this standardization is clearly not just
> theoretical of nature.
platform.
Miroslav Pokorny
> 2. Is it .Net an open standard? .Net is the most direct competitor.Yes it is, first under ECMA and then ISO/IEC - where Microsoft has
exactly one vote and no veto right (unlike Oracle in the JCP).
Microsoft's implementation is however not open source, while other
implementations are. Given that you can write applications in C# and
deploy to iOS and Android, this standardization is clearly not just
theoretical of nature.
Please be more accurate in your answers, if dotnet was open then by definition Microsoft would also need to be open as they are the defacto implementation and by definition define the standard. The c# language may be open and a standard but who cares about that when the entire class libraries are not available under a free license. Any language no matter how wonderful is worthless without classlibraries. Therefore it is simply wrong to say dotnet is an open standard, just because some visible parts are open does not make the entire thing open.
If the platform as a while were open then mono would only need to write the runtime and could leverage the Microsoft donated class libraries. Occassionally they can use some bits but they have not yet been able to do this. In the past these was more evident as large portions of the class libraries was incomplete and they even had a link with graphs and figures showing which packages were completed and so on. This is not as visible now that they are catching up and (almost) completed past versions. if dotnet is truely open plz explain why mono still write class libs...?
> 3. Alternatively, one should move to the completely community basedYeah that's what the Scala people are doing. Recent events have me
> platforms such as Python, etc...
starting to believe they are on to something.
--
You received this message because you are subscribed to the Google Groups "The Java Posse" group.
To post to this group, send email to java...@googlegroups.com.
To unsubscribe from this group, send email to javaposse+...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/javaposse?hl=en.
--
mP
Carl Jokl
time as a .Net developer it seemed to me that .Net was quite coupled
to Win32. Many of the APIs and concepts seemed to mirror those of
Win32. In the case of Mono (without knowing the details) I would guess
that Mono either has to bend Linux into adapting to a Win32 way of
doing things or have a different set of API's for Mono for certain
parts of the platform. The Windows.Forms library is a good example. It
is really just a wrapper around the native stuff. I suppose AWT is too
but Windows.Forms leaks the underlying implementation through. For
example, there is a method or property to get the Win32 window handle
pointer. There are a number of classes that implement a IWin32
interface (possibly where this kind of digging down to native Win32
can be done). Granted this is just one library from .Net but it
demonstrates that it is really practically impossible to port the
entire .Net framework in its present form to Linux or other similar O/
S platforms. This may however not be an issue for those which don't
need the bits which cannot be ported.
I believe that Mono only exists by special permission from Microsoft
and so I don't think just anyone can port .Net.
The shame is I used to believe Java was open or at least a good deal
more open than it is proving to be. It is a shame too that Java is
being put to shame by the kind of innovation going on in the .Net
world. I think the ethos is a bit different with .Net. There isn't so
much genuine care about being cross platform but rather just to make
developer's lives easier.
Fabrizio Giudici
>
> a good idea are two completely different questions.
on complaining about a point asserting that it's not legal, while the
absence of any lawsuit implies that there are no basis for a legal action.
On 11/09/2010 10:58 PM, Stuart McCulloch wrote:
Sigh... it's the difference between theory and practice. Just because
someone develops a license doesn't necessarily mean they are best placed
to prosecute in court. Sure they could be called as an expert witness on
that particular license, but that doesn't make them prosecutors - not
all lawyers are equal! (or in other words... who's better at flying a
plane, the person who designed it or a pilot?)
This is how the ASF defines itself:
... The Apache Software Foundation provides organizational, legal, and
financial support for a broad range of open source software projects. ...
What does that "provides legal support"? Isn't it as the FSF, that sees
also as the entity that sues people infringing the GPL? As Boeing, they
seems to be both the person who designs and the one who flies the plane.
Casper Bang
> platform.
3 silos, namely JME, JSE or JEE, with the two former containing UI
library. Microsoft cut the cake differently, and while they did not
standardize nor mandate any UI library (i.e. WinForms). However, it's
not just the language that's ISO/ECMA standardized, the whole platform
actually is (CLR, core libraries as well as language). How well do you
think a mandated cross-platform UI library worked out for JSE?
> Please be more accurate in your answers, if dotnet was open then by
> definition Microsoft would also need to be open as they are the defacto
> implementation and by definition define the standard. The c# language may be
> open and a standard but who cares about that when the entire class libraries
> are not available under a free license. Any language no matter how wonderful
> is worthless without classlibraries. Therefore it is simply wrong to say
> dotnet is an open standard, just because some visible parts are open does
> not make the entire thing open.
is open". That's you (and Fabrizio) extrapolating your understanding
of the Java platforms, to .NET. If you actually read ECMA-334, you'd
see it covers the C# language. ECMA-335 covers the runtime (with
bytecode and common library specification).
> if dotnet is truely open plz explain why mono still write class libs...?
implementation"? The Mono guys are writing class libraries because you
can not magically generate code out of a specification. Some parts on
top of .NET, like the DLR (Dynamic Language Runtime) is in fact open
source under Apache and can be readily used by Mono though.
Fabrizio Giudici
>> But I'm not talking only of the language - I'm talking of the whole
>> platform.
> > From the Java world, you are used to "the whole platform" being one of
> 3 silos, namely JME, JSE or JEE, with the two former containing UI
> library. Microsoft cut the cake differently, and while they did not
> standardize nor mandate any UI library (i.e. WinForms). However, it's
> not just the language that's ISO/ECMA standardized, the whole platform
> actually is (CLR, core libraries as well as language). How well do you
> think a mandated cross-platform UI library worked out for JSE?
>
application is a whole set of stuff. I'm refining my question: is
something like EJB, Servlets/JSP and JPA part of the things that are
standardized? I've understood that something like AWT/Swing isn't.
Second question: what's the equivalent of the JCP? In the .Net world,
who decides the evolution strategies?
geir
On Nov 9, 2:12 pm, Kirk <kirk.pepperd...@gmail.com> wrote:
[SNIP]
>
> However, there are still a few outstanding questions and Apache is one of them. I don't have an answer except to say that it appears that Apache instead of trying to work with Oracle and engage with them has decided to threaten to throw the toys out of the pram. I would be a shame if they did this however, the JCP would go on though it would be weaker without that strong open voice. That said, there are some radical elements in the ASF that seem to want to bash Oracle (Sun) for the sake of bashing them.
>
The ASF has been working with Sun, Oracle and the rest of the JCP for
*years* trying to get this issue resolved as patiently and
constructively as possible. Don't forget that Oracle was one of our
strongest public supporters of our efforts to secure a TCK for Apache
Harmony, so it should be clear that what we're trying to get is
rational and legitimate. Oracle now has different business
motivations behind the switch in their position (as do EC members
now...)
Remember also that Doug Lea quit the JCP over this issue, and you know
that he's not a "radical element".
But Oracle has take a position regarding their obligation to provide a
TCK license for Java SE that is counter to the their contractual
obligations as a signer of the JSPA, and counter to the public promise
that Java is an open ecosystem.
> My conversation with those in Apache that I know was to ask the question, has someone broken a legal agreement to which the answer from ASF's POV is yes.
excuse from EC members who change their position on this issue - that
it's really a contractual dispute between the ASF and Oracle and the
EC has no standing.
> But in reality, the question is answered with OpenJDK licensing. They can fork OpenJDK and they are clear. Yet the refuse due to some ideological position that was encouraged by IBM (in their battle against Sun for things they didn't like). Another point, if someone broke a legal agreement there are legal remedies. Yet no one at ASF will stand up and say why after I don't know how many years of whining about the problem, they haven't used any of those legal remedies. And as much as I may or may not agree with them, the whining is getting really really old. I won't name names but some have been posting a lot of FUD about various things that Oracle is planning on doing. To be fair, I don't completely know what Oracle is planning on doing with Java but then neither do they and until just recently, their blogs about Oracle have all been speculation and conjecture and all based on dubious facts and in some case fiction.
FUD too.
The fact that one can fork OpenJDK is irrelevant. Apache Harmony
started before OpenJDK, and arguably OpenJDK happened as a response to
the rapid progress that Apache Harmony was making.
One of the core values of Java is that there are multiple
implementations of specs. The Java ecosystem is strong because of
that basic principle - that users can choose among multiple
implementations of the same spec, where the TCK guarantees (to a
limit) that those implementations will behave the same way.
It's nice that you can choose between multiple implementations of JPA,
of the servlet spec, etc - choice is good, as it leads to innovation.
Imagine if no one was allowed to implement the EE spec, that we had to
use the RI from Sun. Don't you think the Java ecosystem would be
poorer because of it?
While you may think of the differences between the Apache License and
GPLv2 as some sort of "ideological difference", there are legitimate
and real-world differences, such as lack of clear patent protection
under GPLv2 (did you ever stop for a second and wonder why Sun didn't
put OpenJDK under GPLv3 which has modern patent handling? Have you
thought about why Google has to claim estoppel vs just pointing to
clear patent grants in OpenJDK in their defense of Oracle's patent
suit?) as well as the differences in how downstream users can use the
software in proprietary products. Do you see Oracle shipping any
GPLv2 code? Have you asked Oracle why they will continue with their
closed-source distribution of the Java RI and components to customers
like IBM, SAP, HP? Have you asked if IBM will be now shipping their
JRE with GPLv2 code?
Licensing matters.
But putting that aside, if we as an ecosystem lose the ability to
implement JSRs (where "implement" means create, test with TCK, and
distribute under our own terms...) then the whole JCP becomes little
more than an Oracle Customer Advisory Panel and I don't think that's a
good thing.
We already have one major single-company-controlled technical platform
in .NET. I don't think the world needs a second one. Lets keep Java
open.
geir
geir
On Nov 9, 6:49 pm, Kirk <kirk.pepperd...@gmail.com> wrote:
> On Nov 10, 2010, at 12:13 AM, Greg Reddin wrote:
>
> >> it appears that Apache instead of trying to work with Oracle and engage with them has decided to threaten to throw the toys out of the pram.
>
> > I would submit that Apache has been trying to work with Oracle and Sun
> > for several years and the process has gained nothing. Maybe it's time
> > for a different strategy.
>
> Trying to work with or just crying foul if ASF doesn't get everything it's way?
It's been *years* Kirk, and Oracle as a vocal and strong supporter of
our position. You seem to be working really hard to spin this as if
the ASF is just being impulsive and irrational, but as the VP JCP at
the ASF that has been leading this effort, I can guarantee that a lot
of sweat, blood and tears went into this battle.
When Sun was refusing to give the ASF the TCK, Oracle's firm opinion
was that Sun was wrong.
Sun was doing it for internal business reasons. Oracle is now doing
it for internal business reasons.
Now Oracle is wrong.
"Internal business reasons" of one company shouldn't dictate the
governance of a specification system that purports to be open.
That's what this issue really is : Does the JCP EC have any
independence from Oracle's business interests? Is the Java ecosystem
'open', by any practical definition of the word?
If so, we're staying - we have a strong Java tradition at the ASF, and
we want it to continue. If not, then the ASF won't be used to
"openwash" the JCP system, which will become little more than a
customer advisory board for Oracle.
>
> >> That said, there are some radical elements in the ASF that seem to want to bash Oracle (Sun) for the sake of bashing them.
>
> > If you're going to make an accusation like that, you might want to
> > back it up with some facts.
>
> It's my opinion that some people have continued to blog FUD even when clear statements of intent have been made. I see no rational reason for doing so unless one was just interested in creating more FUD. Others, not myself, have called them freetards..
message, this is just FUD too...)
>
>
>
> >> My conversation with those in Apache that I know was to ask the question, has someone broken a legal agreement to which the answer from ASF's POV is yes. But in reality, the question is answered with OpenJDK licensing. They can fork OpenJDK and they are clear.
>
> > No they can't. Apache cannot fork OpenJDK and release it under the
> > Apache License. If we can't release it under the Apache License, what
> > would be the point?
>
> Well, it would be forked and you could do what you wanted and run a tck and not face field of use restrictions.
that has happened? RedHat's certification of IcedTea doesn't count -
they really aren't a fork, just plugged the holes in the original
OpenJDK release. How about that new managed runtime system led by
Azul? Do they have a TCK yet? Have you read the license for the TCK
for OpenJDK? It makes it clear that the tested code must be
substantially derived from OpenJDK, which sorta negates your assertion
that "you could do what you wanted".
geir
Fabrizio Giudici
[snip]
Ok, so coming to present times, how do ASF think to change things now,
given that ASF (unfortunately) didn't succeed with Sun, at a time when
two large corporates (Oracle and IBM) backed that position, and now they
don't any longer? I think that, unfortunately, that battle has been
lost. There are two concepts: Java as an open standard, and Java as an
open platform, substantially restricted to OpenJDK. As you can see, I
agree with you, Java is not an open standard: we're left with the
OpenJDK. As you said, the only alternative, .Net, is not better. So,
let's keep Java as good as it can be. I think that news and opinion
circulate very well in the community and the concept that Oracle is
taking as much as power as it can in the JCP is clear to most of us.
Given that, they don't have 100% complete control. Leaving the JCP would
give them more (of course, there are still Google and others, but if
anybody reasoned as Apache and withdrew from the JCP, we'd be delivering
total control to Oracle).
Given that, I understand that withdrawing from the JCP could be a
coherent move, if not the best one in my perspective. But voting against
Java 7? What would be the meaning of that? I prefer to have Java 7,
though not an open standard, in time rather than delayed in the hope to
achieve what can't be probably achieved any longer.