Alessandro has posted 2 posts at DZone. View Full User Profile

Oh Yes, The Sun Has Not Set ... Yet.

  • submit to reddit

I have just received this e-mail:

Hello fellow Project Administrators, It's with a sad heart that we have to announce that the domain will be shutdown as part of the consolidation of project hosting sites now that Sun is a wholly owned subsidiary of Oracle. Project Kenai has always existed as two different things: Kenai the infrastructure, and Kenai the website ( While it has come time to close the domain of, the infrastructure (which is already used under will live on to support other domains in the future. With this decision from Oracle to close the domain, it is now time for project owners to begin the process of migrating their repositories and content over to other locations.

It looks like Oracle's commitment to open source has a fresh start. Yes, I do mean to be provocative.

Published at DZone with permission of its author, Alessandro Santini.

(Note: Opinions expressed in this article and its replies are the opinions of their respective authors and not those of DZone, Inc.)


Martin Wildam replied on Wed, 2010/02/03 - 4:15am

I personally thought of trying Kenai as NetBeans recently added integration with it. From what I heard in the webcasts I am calm for those Sun products I rely on.

Of course Oracle cannot continue 100% the same way as Sun did - as Sun did not succeed with that set of strategies. So of course Oracle need to change some things.

What do you suggest, where should one host projects when now need to move away from Kenai?

Ryan Developer replied on Wed, 2010/02/03 - 8:25am

I'm happy that and Sun's public cloud were some of the only things that Oracle is cutting.   If you watch the Oracle developer tools video they say that didn't get the kind of usage they were hoping for.  At least GlassFish, NetBeans, MySQL, OpenSolaris, and a bunch of other Sun software are safe and getting increased investment.

Fabrizio Giudici replied on Wed, 2010/02/03 - 8:40am

What provocation? :-) Everybody understands that closing a forge is not relevant with the open source attitude of _products_ - that of course we will closely monitor. And it's quite obvious to guess what would have been Kenai's fate with IBM, being it so bound to the Sun cloud for which IBM has a direct competitor.

Alessandro Santini replied on Wed, 2010/02/03 - 10:43am in response to: Fabrizio Giudici

Ciao Fabrizio,

I do not necessarily agree that a forge is a lesser god compared to a product - if you can move to another forge, you can move to another open source product. This is the usual disturbing idea of "open source" == "free as in beer" off-the-shelf products.

As to IBM's approach, I would imagine they would have migrated the whole repositories onto their own cloud and possibly onto Rational Jazz... but that's only an hypotesis.

Alessandro Santini replied on Wed, 2010/02/03 - 10:54am in response to:


I'd stare at the window for three quarters and see what happens in the light of the cost that these open source products will bring to Oracle's balance sheet.

Presently Oracle has too many people looking at them and it would be too unwise to say "oh no, we will cut this and that." - surely not everything will be cut or dismissed, but I still se GlassFish, Netbeans and MySQL nice candidates - the first safer only because of technology leadership.

Reza Rahman replied on Wed, 2010/02/03 - 12:25pm


Please don't take offense to this, but I do hope you can acknowledge that you are stating the obvious here and I am not sure it adds much value at this point (indeed it may be causing some easily avoidable harm). I think it would be much more useful to actually work with Oracle to ensure they get the right constructive feedback when it is obviuos that it is actually needed to help make the best decision for both them and the community at large. The truth is that cutting other things before cutting these makes sound sense all around and they know that.

That being said, it is clear to me that you are expressing genuine conceen that is not malicious, so that is ultimately good (in moderation) :-).


Fabrizio Giudici replied on Wed, 2010/02/03 - 8:36pm

Well, but - and I'm saying this as owner of a dozen of projects at Kenai - and moving to Rational Jazz rather than GoogleCode is more or less the same annoyance. In the end, I'm loosing a few days of work (distributed in two months, even though I think I'll do that in a shorter time). If a FLOSS software I'm relying on would disappear in a day (BTW, a thing that is not possible) I'd have to redesign parts of my software.

In a broader perspective, I agree that Oracle's strategy could be to slowly kill things. Let's see, but let's also judge them by what they do, not by what we fear they could do.

Ryan Developer replied on Wed, 2010/02/03 - 9:04pm

I'd bet money on Oracle not dropping MySQL.  MySQL is what held up the merger for so long, and Oracle clearly wants it badly.   I also feel confident about the future of GlassFish, especially since they need a Java EE reference implementation, and because it is better to get money from these customers than to force them them into migrating to JBoss or Resin.   GlassFish and MySQL sell to a type of customer that Oracle didn't have before. These customers would never spend tens of thousands of dollars on middleware products.  Oracle can keep these customers and continually send marketing email about Oracle's other software, hardware and service offerings just as Sun used to do.

Reza Rahman replied on Wed, 2010/02/03 - 9:23pm in response to:


I agree it is obvious they have a real interest in GlassFish (and MySQL). Take a look here at what the Oracle guys are saying quite publicly: I'm not a betting man, but if I was, I think Oracle AS and JDeveloper will be dropped before GlassFish and NetBeans based on deployment numbers, let alone technical merit/monetary value.



Scot Mcphee replied on Thu, 2010/02/04 - 3:40am

The problem for Glassfish, as the first sentence of that ServerSide article states, is that Oracle view it as "non mission critical" (I believe Oracle used the terms "departmental" and also "reference implementation"). Technology just doesn't come into it. Not so long ago Oracle spent a -big- wad of money acquiring an app server (Weblogic) and then a stack of more money porting all its other products into it and branding the resulting mess^H^H^H^Hplatform Oracle Fusion Middleware 11g. Now not only do they have their third app server (OC4J/OAS, Weblogic and Glassfish), but the Sun product suite includes products that compete with various Fusion Middleware 11g products (portals, ESBs, and so on).

 So on one hand you've got a "departmental" application server, which you can either licence for free by using the open-source version, or buy support for the fancier version, and on the other, an expensive, full-stack-integrated (all the way to the IDE), fully branded platform that Oracle just invested a vast amount of money into, and are pushing it like crazy the past six months.And hte same sales team will sell both this licensed "departmental" Glassfish. And if you say the magic words like "need a fault-tolerant cluster" or "maybe we might build a portal" lo and behold you'll find the molto-dinero "Fusion Middleware 11g" solution installed all over your sorry arse quicker than you can say "can you please explain this per-core with special CPU-architecture-loading-factor licencing schema to me once again and why is it a different price if I upgrade my hardware without adding any additional cores and why won't you let me run it under Red Hat 5 running on a plain vanilla Xen hypervisor?". Even if you deliberately ask for Glassfish directly I bet you'll have to fight off the Weblogic borg absolutely tooth and nail to the last man as they repeatedly try to board your IT department.

Basically, after a year of not selling barely any Glassfish/Sun-One licences - because if you ask for any of the features that are in the licenced version and not the open-source one, you'll be pushed to Weblogic (and anyway, at ten times the price they'll prefer to sell you Weblogic as a default position) - Oracle will announce, "there's no sales in it", then probably ditch the licenced version completely, leaving only the open source version. Finally they'll cut the open source funding off and it will have to limp along without hardly any of the resources it formerly had.

It's a pity 'cos IMHO Glassfish is ten thousand times a better app server than anything Oracle ever produced, or even bought.


Reza Rahman replied on Thu, 2010/02/04 - 11:19am in response to: Scot Mcphee


As is also explained on the TSS thread by the Oracle/Sun folks, the "departmental" part is also mostly a mistake (and a bad one). If that wasn't the case, the Oracle folks would not have made the effort to clarify what they said and keep this kind of language out going forward. Just for some perspective, IBM uses similar terminology for WebSphere CE/Geronimo and I bet this is where their marketing guys got it from: No one argues that WebSphere CE/Geronimo are not production quality or that you can't get support for them if you need it. You'll notice that the Oracle application server page keeps such language out, not to mention the explicit commitment to keeping the GlassFish community vibrant:
While it's OK to keep our collective eyes open, it's not wise to keep making the worst assumptions possible, especially without facts backing it. If Oracle wanted to kill GlassFish, they would have done it already or donated the code to say, Apache or Eclipse. Let's not forget Oracle maintains the EclipseLink JPA 2 implementation as open source too and no one is arguing that it is not production quality.


Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.