Layered Open Source - A better open core model?By Sean Michael Kerner | November 30, 2011
From the 'Old Open Source Business Models Reborn' files:
The term open core is poison for an open source project today isn't it?
More often than not, the term open core refers to some crippled piece of open source software that only really works when paired with some type of proprietary shell that make the software usable.
Yet, the open core model – is still what many open source software firms (or at least, those that have raised venture funds) are embracing.
There is another way.
It's called the Layered Open Source Model, which is kind of like open core, but not nearly as evil. The first person I ever heard use the term (and likely the one who should get credit for 'coining' it is Emil Eifrem, the co-founder of the Neo4j open source NoSQL project.
Traditional open core is where the core project is open source and the add-ons are proprietary and commercial. For Neo, that's not the case. Their advanced and enterprise software editions are still open source, but under the AGPL.
"The AGPL basically says that if you're open source, than I'm open source. If want to contribute to the commons, awesome, you can use my AGPL software for free," Eifrem told me. "But if you want to take it off the commons and you want to make money, than we should make money too."
Layered open source is a variant to open core the difference is that all the software that Neo sells is still available under an open source license.
"I've never liked open core not because it's immoral or Richard Stallman hates it, I never like it because it's not a clear method," Eifrem said.
The trouble for an open core vendor is that on one hand they're telling customers that open source is great and produces superior software. On the hand the open core vendors are saying open source is great, except when it comes to the stuff they want to charge for, which will be kept secret. It leads confusion since if open source is a superior model, shouldn't it superior for the add-ons too?
Yes, I know, this does sound a bit like the dual-licensing strategy that MySQL pioneered. Just like open core now, dual-licensing also isn't loved by all today either, as people still worry that some the commercially licensed elements are somehow different than the open source ones.
I don't know if Eifrem's Layered Open Source model will ever take hold, there is too much money in open core already. That said, Red Hat is a Billion dollar company on (mostly) open source, so why can't some other (mostly) open source model work too?
Firefox 11 Gets Vibrator APIBy Sean Michael Kerner | November 29, 2011
From the 'Browser that Vibrates' files:
Mozilla developers have landed some interesting new mobile features for the upcoming Firefox 11 for Android release. The new features will enable the browser to take full advantage of underlying hardware features, including the ability to vibrate, use a camera, check battery status and send an SMS.
"Web apps, just like native ones, need a way to send feedback to various sensors - including notification vibes," Mozilla Developer John Hammink wrote. "The W3C spec lays a foundation, which we've implemented. Note that you'll need a device with a vibrator to test this one."
The Vibrator capabilities in Android is something that users benefit from today. Though the notification system users are alerted whenever a new Gmail message comes in, and if the ringer is set to vibrate, the phone vibrates for the message notification. With Firefox today, there is a notification mechanism for tabs (if a new item comes in) that could be linked to the Vibrator API for mobile, though it likely only makes sense for a tablet and not a small phone screen. I can also see this as a way to enhance the mobile browser based gaming experience with additional feedback.
In addition to Vibrator API, the new Battery API will integrate with battery status, which again is not something you normally thing off with a browser. But as Mozilla pushes forward with its' own mobile efforts like Boot2Gecko, the need to expose the underlying hardware becomes increasingly important.
Now the Camera API is another story - that's something that makes a whole lot of sense right away. With it you could perhaps use Twitter browser app (instead of the native app) and use the camera for example. SMS is another obvious API, and it's great to see it land.
Overall these APIs are not the sorts of things you would ever see on a desktop based browser. But as Mozilla (and everyone else) moves to an increasingly mobile world, the time has come for the browser to fully embrace mobile features.
You can check out the latest Firefox 11 for Android builds at: https://ftp.mozilla.org/pub/mozilla.org/mobile/nightly/latest-mozilla-central-android-xul/
Is CouchDB in Trouble?By Sean Michael Kerner | November 23, 2011
From the 'Not so relaxing news' files:
An interesting mailing list posting caught my eye the other day titled,
EOL for couchdb.
At first I did a double-take, couchDB is hitting end-of-life? How could that be?
As it turns out it is the CouchDB support that Ubuntu has that is going EOL and that's very bad news for CouchDB in my opinion.
Long story short is that Ubuntu - with UbuntuOne and other efforts - is the reason that I first ever even heard of the CouchDB and now Ubuntu is giving up on it.
"For the last three years we have worked with the company behind CouchDB to make it scale in the particular ways we need it to scale in our server environment," Ubuntu developer John Rowland Lenton wrote in a mailing list post. "Our situation is rather unique, and we were unable to resolve some of the issues we came across. We were thus unable to make CouchDB scale up to the millions of users and databases we have in our datacentres, and furthermore we were unable to make it scale down to be a reasonable load on small client machines."
So to recap, CouchDB doesn't scale enough and it's also too big for smaller devices. CouchBase, one of the leading commercial sponsors behind CouchDB should be plenty worried.
To be fair, Ubuntu leaving an upstream project for their own needs is nothing new. You need to look no further than Mark Shuttleworth's split from GNOME 3 with the Unity interface. The difference this time around though, is it's not just the community that Ubuntu is splitting from, but the commercial relationship too. It's one thing to have dis-agreements within an open source community, it's another not to be able to get a commercial vendor to help tailor a solution that will work.
CouchDB is of course more widely used that just Ubuntu. However, I would suspect that whatever Ubuntu does come up with as a solution for the Ubuntu 12.04 release will become a serious and significant competitor for CouchDB.
PHP 5.4 Hits RC1By Sean Michael Kerner | November 15, 2011
From the 'Not Far Now..' files:
The first Release Candidate (RC 1) for PHP 5.4 is now out, marking the end of the feature development phase of the next generation of PHP.
"No new features will be included before the final version of PHP 5.4.0," the PHP 5.4 rc1 announcement states. "The release candidate phase is intended as a period of bug fixing prior to the stable release."
I've been tracking PHP 5.4 development for awhile and from early on it's held my interest for at least one critical reason: performance. In November of 2010, Andi Gutmans, CEO of Zend said that PHP 5.4 could lower PHP's memory footprint by as much as 35 percent.That's a really big deal that will have immediate impact on the millions of PHP users out there.
Additionally PHP 5.4 marks the end of Magic Quotes (don't let the door hit you the way out!)
In PHP 5.4, developers will also be able to turn the
MB string on and off, so the multibyte support will be available without having to recompile PHP. According to Gutmans, that will provide a significant advantage to companies that want to have a common build. As opposed to PHP 5.3 which debuted in 2009, PHP 5.4 is more evolutionary and will also likely be easier for developers to adopt and benefit from.
"With PHP 5.3, if you want to take advantage of the namespace support, it essentially mandated a rewrite, or at least substantial changes," Zeev Suraski, CTO of Zend told InternetNews.com in an interview last month. "PHP 5.4 will be more evolutionary than PHP 5.3, so it won't mandate a rewrite of code."
It's been a long road for PHP 5.4, but now that it's at the release candidate stage, it's finally nearing general availability.
So Oracle - Are you Supporting Linux or Unix?By Sean Michael Kerner | November 15, 2011
From the 'Unix vs. Linux' files:
Oracle is in an interesting position. It is now a supporter of both Linux and Unix with their own Oracle Enterprise Linux and Solaris Unix operating systems. This past week, Oracle released Solaris 11 their first official Unix release and it made me wonder if the new Solaris is changing Oracle's position on Linux.
I asked Markus Flierl, vice president of software development at Oracle that question and the answer I got, was not what I had expected.
Flierl told me that Oracle's positioning on when to use Linux or Solaris is not set in stone. It's not something that is strictly defined by workload either.
"There is the skillset of the customer and the history of the customer. It's very hard to say here is a project where only this or that will work, there is definitely a lot of overlap," Flierl told me. "Quite frankly we're tyring to stay away from a positioning as any positioning we come up with will just lead to confusion."
Flierl added that the Linux vs Unix positioning is somewhat like Oracle's challenge in positioning x86 vs SPARC on the chip side. Officially speaking, Oracle has said that they will support both equally (though CEO Larry Ellison did make some remarks last month that he didn't care much for x86 in general).
For both the x86/SPARC question as well as Linux/Solaris question, Flierl sees momentum and history as being the driving factors.
"In most places it falls into place automatically because of the skillset and investment that a company already has in place," he said.
So, what does that mean? Does it mean that Oracle favors one OS over the other?
No, it does not, which is a good thing. Other vendors have positioned Unix vs Linux differently. I've spoken with HP at multiple points over the last 10 years on how they view Linux vs Unix, since they have HP-UX and also support Linux. On at least one occasion (officially at least), HP told me that Unix is more mature than Linux. Oracle didn't say anything like that.
Solaris vs Linux is a matter of choice, not necessarily a positioning of one versus the other as being more mission-critical or more feature-rich than the other. That's a good thing for Oracle.
For years, many Linux vendors made a lot of money by migrating people from Solaris to Linux and no doubt they'll continue to do that. The fact that Oracle isn't outright saying that they're looking to migrate their Linux users to Solaris, likely means that the flow to Linux will continue to be one-way.
Apache Harmony is Dead. OpenJDK Wins Open Source Java Battle.By Sean Michael Kerner | November 10, 2011
From the 'Good Ideas that Failed' files:
Remember Apache Harmony? The Apache Software Foundation led effort to build a fully open source implementation of Java?
Guess what? The project is now (mostly) dead and that's not good news.
In a vote, Apache Harmony developers decided to move Harmony to the Attic, essentially shelving the project that once held the prospect of being the leading light in the open source Java world.
Back in 2005, Apache's Harmony effort was seen as being essential for open source development. Things have changed since then, Sun open sourced Java, Oracle acquired Sun. And oh yeah both Red Hat and IBM now back the OpenJDK open source implementation of Java effort.
The ascendancy of OpenJDK has probably led some to think about why Apache Harmony might have been necessary. That said, one needs to look no further than Apache Tomcat to see why Apache really could have done something great, that is if IBM and others didn't abandon Harmony.
Functionally the OpenJDK works and is the home of open source Java, though the Apache Software Foundation itself still has issues with the JCP (which they quit). The failure of Harmony is not a failure of open source necessarily. Java today is open source, which it wasn't back in 2005 when Harmony got started. Harmony's failure has to more with OpenJDK's success.
At long last, Mozilla Releases Lightning 1.0 CalendarBy Sean Michael Kerner | November 10, 2011
From the 'Date and Time' files:
Back in September, I wrote about the pending release of the Mozilla Lightning Calendar. At the time, I had expected the 1.0 release to come out at the end of September - as it runs out I was off by over a month.
But hey, when we've all waited years, what's another few weeks between friends?
Over the course of the beta period, there have been over a million Lightning users and I suspect that the 1.0 nameplate will expand that number dramatically.
It has been very long time since the has been a 1.0 calendar from Mozilla, you'd have to go all the way back to the Mozilla/Netscape Suite when there was integrated mail to even come close. The 1.0 number however isn't really about product quality though, it's about Mozilla developers drawing a line in the sand and finally declaring the milestone.
In the era where many people rely on cloud based calendars like Google Calendar, I personally think that Lightning represents a breath of fresh air. Lightning (like most modern calendars) can and will share data with Google Calendar. It's not quite as seamless as it could be, but there is a decent help page to guide users through the process.
And for those of you that were not part of the first million users of Mozilla Lightning, check out this Mozilla video that will fill you in on what you're missing.
Eclipse Turns 10By Sean Michael Kerner | November 03, 2011
From the 'Time Flies' files:
The Eclipse project officially turns 10 today. Time sure does fly. It seem like just yesterday that I was writing about the fifth anniversary of Eclipse.
Eclipse started out as just an open source IDE that was open sourced by IBM, that's the genesis. The broader Eclipse effort really took off in January of 2004 with the creation of the Eclipse Foundation which has become home to more projects than I can count. When it comes to developer tools and tooling, Eclipse has become the defacto go too organization in the open source world.
It's the place where competitors collaborate on code. It's also not just a bastion of free code for freedom's sake, it's also the basis of many commercial products. I've been told time again by Eclipse and it's supporters that the Eclipse way is all about making open source projects that can then be commericalized in some way by members (or others).
Eclipse isn't just about Java either as it was five or ten years ago. I'm a big fan of the PDT Project, which delivers PHP Developer Tools and is the basis of the Zend Studio IDE.
Another thing that has changed in the last five years is the fact that Sun is gone and Oracle has always been a strong supporter of Eclipse. While I haven't seen wholesale shift of Netbeans developers to Eclipse, Oracle is clearly more friendly than Sun had been toward Eclipse and that's a good thing for Eclipse over all.
The fact that Eclipse has survived 10 years and has become more relevant than ever is a testament to both IBM for its early vision and the strength of the open source model for development.
Happy Birthday Eclipse!
Ubuntu One for Windows Bringing new users to Linux?By Sean Michael Kerner | November 01, 2011
From the 'Linux Tech for Windows Users' files:
How is Ubuntu gaining users? One way is by appealing to Windows users with Ubuntu One.
During the Ubuntu Developer Summit this week, Mark Shuttleworth, the founder of Ubuntu praised the Ubuntu One team for their recent Ubuntu One for Windows launch.
With Ubuntu One for Windows users on Windows platforms can sync files to their own personal cloud. Ubuntu One had previously been available for Ubuntu Linux and Android users.
So far the introduction of Ubuntu One for Windows has been a big boost, according to Shuttleworth.
"In the month that this was released we had 110,000 new users of UbuntuOne," Shuttleworth said. "25,000 of them had never used Ubuntu before. 25,000 users are coming to this platform just because of Ubuntu One."
What's not entirely clear to me is how many of those new users are actually using Linux (aside from the fact that the backend is all Linux). It's not clear whether or not Ubuntu One for Windows means more Ubuntu Linux users.
What it does do however is extend the Ubuntu brand and open the door for some (potential) Ubuntu adoption, which isn't a bad idea.
This isn't the first time that Ubuntu has made a direct appeal to Windows users. The Wubi effort that let Windows users install Linux inside of Windows (without re-partioning) was a previous attempt to go after Windows noobs.
With Ubuntu One though, I suspect that there can and will be a number of user that never touch Linux at all and just want the free cloud space that Ubuntu One provides (as an alternative to say Dropbox).