Wednesday, November 22, 2006

Top spot on Javalobby ... and their newsletter

Javalobby has pinned a discussion about Magnolia powering their network as the top news item on their homepage, and in addition we were the top feature of their weekly newsletter. Needless to say, we are very happy about that. (Thanks Rick and Matt!).

Matt, Rick and I first talked about Magnolia powering their network of sites maybe 1.5 years ago. Back then, Javalobby decided it was to early to replace their custom software with a standard CMS like Magnolia. Since then, Magnolia has steadily improved, and JSR170 is much more on people´s minds than even one year ago.

We (Greg, me, Rick & Matt) had a two hour session today where we discussed the major requirements for their new infrastructure, which Magnolia can easily fulfil. Also Rick played around with their installation of the Enterprise Edition and enjoyed Magnolia´s ease-of-use.

I am excited about the fact that the best Java Community site likes what we do. Its up to them to see if Magnolia makes their lives easier or not - as stated in my last post, everybody in the CMS space has different needs. And replacing an existing system is always a major undertaking - to let go of all you have built yourself, to be at the mercy of something prefabricated - is usually difficult for any developer. But in the long run its inevitable.

By the way, this is just as true for Magnolia itself. We constantly discuss replacing parts of our product with new stuff that has been created by other projects. Its a lot of work, but it has significant benefits once its accomplished. For one, know-how of the product is not in the head of the one or two persons that developed it. And it might even be documented! In any way, its less work to maintain, and allows us to focus on the parts where we can add value.

Tuesday, November 21, 2006

Promises and products

If you work in a CMS area, you know that there are more than 1000 CMSs out there. Most of them are more or less irrelevant, some are hugely popular (for instance, Typo3 is huge in Germany), some are very very powerful, and each one has its own set of pros and cons.

No system can reasonably fulfil all client's requirements and still be usable. Each client has different content management requirements. One of the reasons we developed Magnolia is that larger projects always need customization or integration. We thought that making these really easy to do and maintain, this could be a significant advantage for Magnolia (which it is).

So if you look at selecting a new ECM and are looking in the domain JSR-170/open-source, you will naturally have to make a choice between the two dominant players in that space: us - and them. We get asked about the differences between Alfresco (them) and Magnolia (us) about once a day by a prospective client. I have never answered these, but a recent posting on Javalobby made me reply along the following lines:

  • On the surface both "support" JSR 170. Magnolia is build on the API, whereas A. has added support for that API at a late stage. A. locks you into their repository, while Magnolia is repository independent. Since your data is where the value sits, that's a pretty effective lock in. Magnolia runs with any repository out there: Jackrabbit, CRX, Exo(?), soon Oracle and any new compliant repository. You have the choice.
  • A. is coming from a DMS end, only now adding support for CMS. DMS and CMS are not integrated in any way. Magnolia is coming from the CMS end, where it shines. Its DMS and CMS are integrated seamlessly.
  • God is in the details, and paper is no substitute for a product. Magnolia has been in the works for more than 3.5 years now, and our interest was always to build the easiest to use CMS in the world. Our clients tell us that working with Magnolia is highly intuitive and easy. I doubt that anybody would say that about A´s GUI (we have hear that system integrators simply replace their GUI with something usable, which obviously would be driving up TCO considerably)
  • Magnolia ships with Sitedesigner, a tool that replaces client-side web development software like Dreamweaver or Golive. This is much more useful than it initially sounds - you have to do a project with it to realize just how amazing it is to work this way. It saves massive amounts of time, layout is versioned (!), and updates work without worries (because we guarantee that).
  • Magnolia ships with the only opensource cross-language business process engine on the planet: OpenWFE. Its one of the top ten projects on Sourceforge regularly, and is amazing in its flexibility and power. If you want to integrate your business with your web initiatives, and send workitems ("tasks") to agents written in python, java, perl, dotnet etc that's what you can do. A. on the other hand uses something much less flexible and powerful.
  • A. was founded with a big pile of cash and massive promises to clients and shareholders alike. They have to deliver, and the way its done is by spending on marketing and sales. Magnolia on the other hand has been an organic development. No piles of cash to spend on anything, really. Just trying to create the best CMS in the world with a couple of smart brains and lots of experience.
A. is very successful at the moment selling promises. Magnolia is very successful at the moment selling products. Choose, but choose wisely.

(This is my personal view, and reflects in no way the view of my company)

Monday, November 20, 2006

Magnolia EE out - and Javalobby in?

Its been a long while since my last post, and if you look at, you know why. We have managed to release Magnolia® 3.0 Enterprise Edition on November 15th, exactly three years after our first public release, together with the final release of Magnolia Community Edition and the first public release of Sitedesigner™, a tool to layout sites directly in the browser.

We have worked on Sitedesigner for more than a year. As they say: grass won't grow faster if you pull on it. Some things need time, and if we look back on what we had 6 months ago, we know why Sitdesigner needed that time. We have gone through endless tests and iterations and refinements to make Sitedesigner outstanding. It shows. The attention to details is what makes Sitedesigner a pleasure to work with. Try it.

Javalobby does already like the new Magnolia: There is an ongoing discussion if Magnolia would be a good foundation for their next generation community site. Join it!

Thursday, September 28, 2006

Magnolia Geek.Meet 2006

Last weekend saw the first ever Magnolia Geek.Meet, and it was a great event. We gathered in Mission21, a large protestant mission building that has 15'000 m2 of park around it and rents away meeting rooms, provides the catering and even is a hotel.

Attendees were Nico (Japan), Fabrizio (Italy), Chris Miner (Berlin), Greg (Brussels), Ralf (Stuttgart), Philipp, Sameer & myself, plus Pascal who came to say hi in the beginning & bye at the end.

I moderated the event. After a quick intro of who is who in Magnolia community-land, we started brainstorming. All topics then were voted for – each participant had 4 votes. We then proceeded to discuss the 4 main topics for roughly 90 minutes each, plus breaks. Rinse and repeat on Sunday.

Some of the big issues include
  1. Beanification (a word that stuck after initial laughter) - meaning: create more classes (beans), which means they can be documented and even tested (!)
  2. Introduction of Spring for various usages, including Spring Remoting (Fabrizio came up with a Spring way of doing basically any topic we discussed except documentation :-))
  3. GUI toolkit/AJAX framework/XML-RPC/XForms etc - the whole world of what we could change with respect to technologies used today. When we started with Magnolia there was no Ajax, so we wrote our own. Today many options exist. Decoupling of the admin interface. Possibly use Spring remoting for intra-application communication (Think: GUI talks to core) and Inter-Magnolia communication. Provide alternate ways of accessing the content (Webservices, CLI)
  4. Workflow Future: now that we have the basis, it is becoming clear that Magnolia's declarative dialog creation + openwfe and storage of workitems in JCR make it very tempting to create form & workflow-based front-end applications. What a great tool for SI's.
  5. Dialog-Refactoring - introduction of JSF for dialogs and trees etc.
  6. WebDAV for access of DMS through client file system browsers (Niko and Sameer start working on it, expect it working before the year is over)
  7. lots more (I'll update this when I have my notes around)
We decided to have more releases on a shorter time frame with well defined major tasks (basically, each topic we discussed will be the focus of a single release). Now that Philipp has finally finished the build mechanism in maven to do a complete distribution (i.e. we can hit "build" and get readily packaged everythings - from wars, to bundled tomcats, community edition, modules... whatever) it is much less work to actually prepare a release. We expect to have dot-releases (3.1, 3.2 ...) every 2-4 months now.

Also we went out Saturday night in Basel, had dinner at the Hasenburg (Rösti) and drinks at the Des'Arts (a bar right below Day's headquarter).

Everybody liked the event and I guess there will be a Geek.Meet '07. Expect the discussions to result in a roadmap of versions, available via jira any day now.

And finally some impressions for those that had to spend the weekend with Japanese girls at Game shows in Tokyo or went sailing at the baltic sea (who can blame you?).

(more photos)

Tuesday, September 12, 2006

New iTunes

I was probably the first guy back in highschool that had a walkman. I was a musician back then, having my own rock band, and music was what was surrounding me day and night, so to be able to listen to music on the go was revolutionary and elevating.

Being interested in music, I loved a web site that was publishing all song lyrics ( a community effort). What a great resource for quotes. Needless to say, it was shutdown soon enough by the music industry because of "copyright issues". I mean, come on "Love love me do, you know I love you" - I am not allowed to post this to a website?

Later, my company bought every employee a first generation iPod right when they were announced, back at the time that was a 11k CHF investment for us. I still got mine, by the way, but added Shuffle, a Nano and a Mini along the way. Needless to say, I have used iTunes since its very first day, and love it.

Now its already release 7 for iTunes. Congratulations. One of the great things is a new feature that lets you flip through your collection. You need the album artwork for that, and the first thing I did was to click the "fetch artwork" menu item. It works quite well, but I was wondering how they managed to convince the record industry to make them available.

Turns out they did not, or at least not in all cases. Several albums that are available in the store did not return their artwork. When you search for the album in the store, the artwork shows up, but in the iTunes cover area (lower left corner) it says "artwork available when purchased". Try it with "intensive care" by Robby Williams as a certainly popular example.

Steve (Jobs) does a fantastic job of bridging the gap between consumer's reality and that of the recording industry. But some things never change, right? I guess I have to use the old-fashioned approach and manually copy the album artwork for Robby Williams from some website like Amazon. Until then, the likelihood that I will listen to Robby is diminished due to the fact that his cover is empty when I flip through the 2652 songs I currently have on my laptop.

Who exactly has gained what by not releasing the album artwork?

Monday, August 28, 2006

Magnolia peak

For some reason, this week has seen a strange peak in Magnolia interest. Magnolia has reached a traffic rank of 13'057 on alexa (that means we are currently ranked the 13'057th place of all alexa-tracked websites on earth, which should mean pretty much any website).

Now maybe that doesn't sound so thrilling, but if you know that at the same time our main perceived competitor's rank isn't even available because they are not in the top 100'000 sites, thats not too shabby. Not that I want to make much out of it, but it feels good.

Also, we are likely reaching a cumulative 150'000 downloads in the next 2-4 weeks, another milestone.

What else is news? An impressive list of large companies is in contact with us, we will be closing a number of significant deals in the next few weeks, and today has seen some serious interest in Magnolia from a very large company. Lets keep our fingers crossed!

And I will be in Paris launching Magnolia France later in September, just after the Geek.Meet.

Now all I need is time to write all these press releases…

WebObjects will be open-sourced

This must have been one of the longest running tragedies in recent computing history. The short story is that WO (as WebObjects is known by its afficinados) is an outstanding piece of software if ever there was one, but never got off the ground commercially, was copied by Open-Source programmers, dropped its price from 50k to 600USD (IIRC) to free, and finally will be released into the open.

Parts of it (the Enterprise Object Framework) stem from the client-server time before the web as we know it was born. Apple (who bought it with its aquisition of NeXT) is using it to build their Web 2.0 suite of apps like .Mac or the iTunes music store, so the capability of it is proven beyond doubt. A couple of years ago I compiled a list of companies using it, and just about every major bank was on that list.

Alas, somehow WO never took off. Other frameworks copied the concepts or even the API. Just a view days ago, jope 1.0 was announced, which might have played a part in the decision to finally open-source WebObjects.

For me, this is several years too late. Would WO have been open-source in 2003, we might have used it to build Magnolia. Today, we are rewriting one of my last remaining apps of that time – an associative web shop that I wrote for the Kunstmuseum Basel – to Magnolia.

Congrats to Apple (and specifically Steve Jobs, who apparently wanted to open-source WO back in 2004) for finally doing it. I hope it picks up speed!

Tuesday, August 22, 2006

workflow state of mind

Today John and I sat together to do some final testing of his decision table implementation. As so often the problem was with the data (data is root of all evil), not the software.

During our obligatory lunch pizza I mentioned to him that I think that programming with and for workflows differs from "normal" programming.

As a "normal" programmer, if I change code, I can compile it, test it, run it and if it does what it should, thats fine. In other words, I am pretty free to change all aspects of my system. Granted, I need to make sure that if I change the API, I change the implementations that use my API, but the whole process is pretty much independent of a time component.

When you deal with workflows, the picture changes profoundly. Why? Because everything you do needs to be done in such a way that long running processes still work.

At first that sounds pretty trivial, or logical. But if you work on a project that has processes running for a year and longer, the impact of this requirement is quite massive. It means that whatever you do (to improve the system, to fix a bug), part of the system will still need to run with the previous, maybe buggy logic. It means you cannot just change things, you need to find a way to know what you are allowed to change, and what not.

It would be nice if a system understands that a "normal" programmer is in a different state of mind, and provides mechanisms to protect her. For instance, once you launch a flow, all definitions could be versioned, so that the programmer is free to change them later without affecting the running system.

Alternatively, instead of versioning, a tool could warn me that there are still running processes that use a definition that I have just changed.

In the context of OpenWFE, workflow definitions are versioned, and the workflow that you launch has a copy of its definition stored with the engine. The same is not true however, if you define (sub-)processes to be loaded at the startup of the engine (a library). These are not versioned, but instead are valid as long as the engine runs (they can be overwritten if you want, since in OpenWFE's Scheme-like semantics, a function/method/flow-definition is the value of a variable, so you can change that value at runtime). Thus, if you change them, and restart the engine, running workflows use the new version of the library, not the one that was present when the workflow was launched. So there is a difference between a sub process defined in a "normal" workflow definition and one defined in a library even when they are syntactically equivalent.

This difference is exactly the difference you need to take into account if you start working with workflow engines. A workflow programmer needs to be in a "workflow state of mind", and take into account that whatever he changes, there might be running processes that depend on previous versions of a what he is changing.

Tuesday, July 11, 2006

Why OpenWFE?

There was a post recently on the OpenWFE developer list asking why Magnolia choose OpenWFE. Here is why:

The decision was made quite some time ago, and it was based on my research and recommendations.

First of all, it was a license issue. I wanted to make sure we can change the license to something other than LGPL (Magnolia's current license) at any time should that prove of advantage. So all projects under GPL/LGPL were off my list. This reduced the list considerably.

Of the remaining projects the exact decision process is lost in time. However, I have looked at the features, the architecture, the perceived success/stability of the project (well, would I have known that John updates his software once a week I might have proposed a different engine :-)).

There were a couple of things at the time that attracted me to OpenWFE, one being that it implements the complete set of workflow patterns, one that its a top project at sourceforge (that makes it easier for me to silence critics), one being that it is cross-platform (that's a big advantage given the fact that there are no homogeneous systems in production).

Finally, I contacted John and found that he is a very likeable fellow (he sits opposite of me at the very moment). It is quite helpful if you can just call someone, or even meet in person, to discuss issues beyond "how to I open a terminal?". So the fact that OpenWFE is a European project and not based in the USA was definitely a big plus for us, because it makes communication very easy.

To sum up: license, features, architecture, success/maturity and support drove my decision.

Others will have other needs, and arrive at other results. Or at the same, but for different reasons.

Thursday, June 29, 2006

Making Javaref really successful

Javaref is the latest stab at making Javadocs useful – and of making a business out of utter lack of usability of the brainless default presentation that has haunted us for about a decade now.

Main ideas of Javaref:
  • search (that was a hard one to come up with ;-))
  • user profiles to restrict the libraries being searched (not especially way-out either)
Thats a nice start. Here are a few more ideas:

How about a mechanism that hocks into maven (which is used to build and document projects like magnolia). Such a plugin could use the maven-generated list of dependencies (and transitive dependencies) and automatically publish it as a project to Javaref (along with the source code or a link to the project's SVN code repository for Javaref to fetch at a later time).

The goal would be that its really easy for a project to have a profile at the Javaref site, and keep it up to date with each release. As a matter of fact, this should not be a profile (thats a user's domain) but a "project". Projects should be published and sharable (hey, viral, here we come!), and a user could have a list of projects he works on.

Now add a simple way to advertise the availablitity of the Javaref project at my project's site to drive my user base to your site (and give me back some of the advertisement cash you make in exchange) and you would have nailed it.

Come to think of it, if these guys don't built it, I might be tempted to build it myself ;-)

Monday, June 26, 2006

To unknown charters: Expedition Open Source

The last three years have seen preparations of the great expedition called "To unknown charters: Expedition Open Source".

Big ships have set sail long before us, heavy vessels with plenty of food to keep the crew happy. But the tides have changed, and big ships are of little use in shallow waters. Most will be sunken or their crew starved before they reach safe lands again. They set sail when the weather was high and the fish was plenty, but the winds have changed, and so did the fish.

The Magnolia has set sail, and we are picking up good speed - the winds seem benevolent to our undertaking. Our ship is small, fast and agile. Our crew is well trained and multi-disciplinary. Our equipment is versatile and our compass shows the direction of change, the winds of which fills our sails.

Less than a week ago we have spotted the biggest fish so far, and we netted it today. It may have been a small fish for the big vessels, but for the humble Magnolia, its a big catch, clearly showing that for the well-prepared, the exciting journey across the ocean is worth taking the risk.

May long she sail in happy waters!

TSS Java Symposium in Barcelona rocks!

I have been in Barcelona the last couple of days, attending the TSS Java Symposium - TSS's first in Europe. The event was excellent - the speakers were great: Simon Phipps keynote on the "The Zen of Free: Models for Understanding Open Source Software" alone was worth the trip to Barcelona. He really gets what is happening.

TSS Barcelona was also a great place to meet people, which is why I went there - I gave a talk about Magnolia with the goal of raising awareness in the Java community. I have met many interesting people and believe that my goal has been met – although it is always difficult to talk about a product/project at a technology conference.

It was a pleasure meeting all of you, and especially I'd like to say "Congrats" to Nitin and his team of pulling this thing off in just about 6 month's of preparation. I surely hope to be part of it again next year, maybe in Lissabon, Vienna or Prague?

Wednesday, June 21, 2006

Wireless access at conferences costing an arm and a leg

I am in Barcelona right now at the TSS Java Symposium (and should prepare my talk instead of blogging) but here is a rant: while the location is certainly great, charging conference attendees for wireless access stinks. The three day pass for the conference is 1700 USD. The hotel is 280,- or so a night (But I stay in a cheaper place, we are an open-source startup, right?). On top of that, the wireless access is 17.40 €€ (22 USD) per day. Do the math: if 500 people want to access their email for the three days the conference lasts, they collectively pay 11'000 USD. Thats mad.

Besides, you can probably fly to Barcelona for the same amount you pay for a single day of wireless access. If you ask me, competition in the wirless space is not working correctly, or else it could not be so expensive.

Tuesday, June 20, 2006

3.0 is out the door, and I am in Barcelona

We released Magnolia 3.0 today, and simultaneously migrated our website. The new website is awesome if I may say so myself, of course created with the new Template Designer we ship as part of the Enterprise Edition.

The site is featuring on its homepage a photo I did 2 years ago on the other side of the world. (guess which one ;-))

I have said so many things about Magnolia 3.0 on (about 100 pages of content currently published that I have written) that I won't repeat anything here. Well, one thing maybe: its featuring John's OpenWFE - this is a big step, the foundation of a system that will be amazingly in its applications. We have done many interesting things with it in the last 2 years, and I am very happy to see it move into Magnolia. Thank you John!

Finally, amongst all the crazyness, I flew to Barcelona right in the middle of the release, posting the news to all the web from my hotel room now. I'll be presenting Magnolia at the TSS Java Symposium - see you there!

Monday, June 05, 2006


Well. If you ever read my enthusiastic endorsement of Sala on LinkedIn you could have known. He is the guy that not only has great ideas, but he also follows up on them. I have watched his idea from the day one (no, I did not buy). It was always clear that there are two options - either it will explode, or it will sooner or later be irrelevant. Well, explode it did. The last time I visited the site before tonight - roughly two weeks ago - he had sold about 64 paintings. As I am writing this, 522 are sold, and I would not be astonished if the rest is gone withing 24 hours - after all, there is apparently a BBC interview coming up.

All of this is no coincidence, if you reads Sala's blog (one of his blogs, that is) you can very much understand how one thing adds to another. There was the funny death ad he copied from a newspaper that ended up on boingboing. And the applet he wrote to visualize Websites that attracted more than 100'000 unique visitors in just 4 days. Whatever Sala does is great.

Well, I am happy that its working out for Sala. After all, he was the person that wrote the first GUI of Magnolia and thus helped lay the foundation for its success.

I used his applet to visualize - very nice, I like visualization tools (right, John?)

Friday, June 02, 2006


After many years with obinary at Binningerstrasse 15, we move on. B15 was a great place in many (though not all) respects, and many things will be remembered long after we have left. I recall one meeting with all of the top executives of a large retailer (early 2000) where Pascal was presenting a solution downstairs (then still Switzerland's smallest cinema) while I was frantically hacking it together upstairs. Or the time when Cyrill fell through the roof when trying to install the server room air conditioning (luckily nothing happened to him). The neighbours statement "I knew this would happen, this is already the third time" - still makes us wonder why she did not say so before Cyrill fell through the roof. The parties we had after we took over the cinema-come-jazz-club - surely many will remember the Magnolia 2.0 release party! At some time we even rented a second office in the front house on the forth floor. A web cam connection to see what was happening downstairs resulted in a funny movie still out somewhere - Josh called downstairs about 10 times a day just to see them pick up the phone.

Well, its time to say goodbye. And about time. We spend the last two months in the stables ( a former video studio, see the background blue box) as an interims office - no windows. All in one room. Thats because we had to move out of the main room (see deserted picture above) before we could move into our new offices.

This is a rare foto by the way - it shows Philipp, Leonie and Daniel. Leonie is now in Rome - and Daniel has only worked a very short time at obinary.

Well, we moved today. Servers are up and running (thanks Michi and Pascal!), all communication system live (thanks Kurt!), the coffee machine is working (thanks to Philipp, I assume - Sameer usually drinks tea) and the table soccer has found a new home (thanks Fröde!).

Time to say hello to our new place in Maiengasse 30 that already feels so much better! Lights! Action! Maiengasse 30 - Magnolia 3.0. Here we come.

Now all we need is better weather to open up these large barn doors and enjoy amazing tranquility in the center of Basel.

Thursday, May 25, 2006

Open Source forces enterprise transformation: think value or die

I have just returned from the Tranformation & Innovation conference in Washington, D.C.

I was accepted to talk about "How to evaluate open -source communities" (download slides) and invited Daniel Hinderink of Typo3 fame to join me as a Co-Speaker.

We are coming from opposite directions regarding Open-Source - Typo3 is community driven, where Magnolia is vendor driven.

Initially, we thought that this was a great distinction in the context of evaluting communities. However, preparing for the talk we realized that there the difference between the two approaches are smaller than you might think. This has pretty profound implications, as I shall outline below.

While at first (see slides) the projects are of very different nature, a vendor-driven Open-Source project is soon under pressure to build up its community, else it will not be perceived as "real open-source" and be deprived of many of the opportunities and benefits real open-source brings with it.

A community-driven project on the other hand is "real open-source" from the start. This brings many challenges - the main one being that the project is "project-driven" i.e. contributions are always based on projects undertaken on behalf of customers. Lacking an entity that defines and finances fundamental (architectural) changes to the project, there is no "long-term thinking" possible. To overcome this challenge (and overcome it must be else the project will ultimately fail) some governing body must be founded and financial means must be available. Typo3 for instance has founded the Typo3 association, which is financed by its members, and thus has the means to do long-term planning.

Once such an organization is in place, structure and planning can be imposed on the project - very similary to what a vender-driven project has from the beginning. And once a vender-driven-project has gained sufficient traction in the community, the lines between the two start to blur.

On the other hand, a vendor-driven project might sooner or later choose respectively be forced to found a governing non-profit. Once that happens, the difference between the two approaches is gone. This is the most logical step to avoid the risk of a revolution, although it sometimes back-fires (see Joomla!).

Why "be forced"? We are all in a transition in the Open-Source economy, and nobody really knows where the journey will take us – think of Calvin and Hobbes "Transmogrifier". One thing is for sure: there is no going back. Open Source is one of Friedman's 10 flatterners of the world. In the long run, software companies can either embrace it or vanish - here is why:
  1. The "standing on the shoulder of giants" principle is especially true if these giants are readily available around the world for free
  2. whatever you do, if the value of your product lies in proprietary code, someone out there will build an open-source version of it – because of (1) that gets easier and cheaper every day.
  3. To meet that challenge, either you go down with your prices until you are out of business or go open-source fast, else any possibility to gain significant community mindshare is getting harder every passing day (especially if you are going from proprietary to open-source, where the community always suspects you of potential misuse of their contributions)
  4. Once you are "infected" by the open-source virus, and build up a great community, a lot of value will be created by the community. In the long run, that value will be higher than anything you can provide code-wise.
  5. Open-source is based on meritocracy. Once most of the (core-)code is written outside your company, the likelyhood of a revolution increases significantly unless you release most of the initial influence to a governing body that takes the real power distribution of the committers into account. (by the way this is why sugarCRM or mySql do all coding themselves, und therefore are no "real open-source" projects)
  6. If there is a revolution, you are either dead or part of it.
  7. If you give over control to a governing body, nothing is left of your crown jewels, so look where your value is. If there is none, the project might survive, but you won't.

This is a transition that will take place over time, and time is money. Hence the different approaches enterprises take today to make use of and possibly prolong that time:
  • upselling
  • dual-licensing
  • writing code themselves to buy time
  • trying to keep control of the project
It works, for some, and for the time being. However, one final thought – you are not the only vendor out there that considers to take the plunge into openizing your crown jewels. Once your competitor starts, you better move fast, or you'll be irrelevant before you know it.

Thats the power of open-source – it is a truly transforming force. It forces us to ultimatively think about value, and who would argue that this is not a good thing?

Thursday, May 18, 2006

Magnolia - BOF at the TSS Barcelona

I have just been accepted as a speaker for the TSS Java Symposium in Barcelona

I will present an introduction to Magnolia 3.0, directed at developers and IT Managers. Its focus is to get you up to speed with Magnolia as a product and as an open-source project. It includes a product demo.

You will learn
  • what Magnolia looks like
  • how standard content management tasks are achieved in Magnolia
  • why Magnolia is interesting from a technical perspective (pluggable anything, AJAX-powered GUI, JSR-170, OpenWFE integration)
  • who is using Magnolia (references, quotes and real-world data)
  • what exists community-wise and where to find more information
  • future directions (Spring, Freemarker, JSF, JSON, ...)

I would be especially interested to raise the project's awareness with leading java developers to get more great minds on-board for the project.

Hope to meet you there!

Friday, April 28, 2006

fresher than most

With Magnolia 3.0 around the corner, and a new positioning of our products coming with it, we have decided to open-source Magnolia For Documents, our JSR-170 / magnolia-based document management system. The code will probably be checked in later today.

One reason behind this decision is that we strongly believe that Magnolia 3.0 will be so compelling, that our market share will rise significantly anyways, and be further boosted by the open source DMS. Magnolia immediately is the best open-source WCM and DMS available, so this is going to be a fun ride.

We also decided to scrap the planned (commercially licensed) web edition. Instead the community edition will be free, too, and offer functionality that is way beyond anything out there in the open-source area today.

In short, there will be:
  • open source (no binary)
  • community edition - free (binary with mostly open source)
  • enterprise edition - commercial with enterprise features like LDAP and JSR-168 support
  • business process edition - commercial
More soon through our official channels of communication.

to Matt: magnolia eats its own dog food

In a recent post, Matt was under the impression that is not running on Magnolia.

Well, he is under the wrong impression. Magnolia is using its own CMS for its homepage, and has been doing so ever since before we released the first version (November 15th 2003), exactly since July 16th 2003.

We have built some pretty amazing things with Magnolia, including magnoliaQT (edit videos in your browser, anyone?), Magnolia for Business Processes (draw and document business processes directly in your browser), and Magnolia for Documents.

Magnolia 3.0 will see its beta in the next week, including workflow and versioning, JSR-168 support, JAAS/LDAP, meta-templates including RSS feed generation (and consumtion) and podcast-generation, asset-management, scheduling and an even better new GUI.

As far as kick-ass sites go, thats a very personal perception, - here is one I like, using a quote from Brandon Root (Portent Interactive, who built the site):

"Thanks, you guys made my day Magnolia! We just finished up the Amgen Tour of California,, and everyone loved it!

We were very impressed with how Magnolia handled under the load, even before I got caching working. We averaged about 580 hits a second, although that was an overall average and we probably had about 4 times that during lunchtime. This was all running on a single server per magnolia instance, an intel xeon 3.0 ghz processor with 2 Gb of ram."

I think that kicks ass pretty badly, handling up to 8 million hits per hour. Maybe it needs two minutes for installation instead of one, but boy, that one minute sure buys you a lot ;-)

Friday, April 14, 2006

Magnolia Marketing Marathon

Pascal and I spent two days in Ticino with the intention to take a break from our hectic office and prepare the marketing of Magnolia 3.0

Ticino is a fantastic place - the exact opposite of our office - nice view, amazingly quiet, a place were you can reset your brain and start to be yourself again, and be creative.

Unless you are constantly online, using skype to coordinate projects around the world, mobile phones while taking a stroll to the plaza in Locarno and generally do business as usual, which is of course what we ended up doing instead.

Nevertheless, it was great being there, and yes, the pictures really show 40 gramms of red hot chili peppers and a handful of garlic to go into a brilliant fish curry we prepared for dinner on the first evening.

We still have managed to structure our new website and come up with a couple of great ideas.

And the magnolias where in full blossom (just that I did not take a single snapshot). What a beautiful view.

Magnolia's Multi-Server Advantage

Magnolia has built-in clustering capabilities. These come in the form of a pluggable publish-subscribe mechanism that allows to have any number of subscribers for a publisher.

Magnolia by default runs on a multi-instance setup - one authoring and one public instance - where the public instance is subscribed to the authoring instance.

This architecture has a lot of advantages. Two instances provide flexibility, performance and security. It separates concerns, scales well and allows creative solutions for interesting problems. Most of all, it allows to run Magnolia on two (or more) separate servers serving different purposes, namely to author content in a secured environment and to publicly serve the content to site visitors.

Stability of the public instance. Authors have the liberty to do whatever they need to without affecting the public instance. If something they do is so severe that the server needs to be restarted, this is not a problem. This even extends to developers that need to add new templates and other configuration changes that require a restart. All of this can be implemented and tested on the authoring instance without affecting the public instance which is visible by the user community.

The authoring instance needs system resources to facilitate editing content, locating and managing data, and constantly checking permissions. Often, you also output more log information, which costs additional resources.

The public instance is only concerned with generating cached pages and then serving those pages when requested. With this separation, content editors are free to tax this system as much as they like, developers can put in hot fixes required by content editors, and none of this impacts the performance or availability of the public instance.

Commonly, you want the authoring instance to be behind a firewall, and the public instance accessible from the internet. There are mainly two reasons why this setup makes sense - one is to make sure that no one can publish content by directly manipulating the public server, and the second one is to make sure that information is only then available to the public when it should be. A typical example is the publishing of a press release or financial information. Authors can edit the information behind the firewall and only once the publishing date has come, this information is pushed to the public server. So even if - for whatever reason - your public server is compromised, only public information is found on the system. At the same time, since no authors need to access the system directly, you can secure the public server tightly.

Due to this architecture, you can easily run the two instances on two different servers (or more instances on more servers). For instance, one of our customers has the need to publish information to 70 standalone servers for security reasons. Each of these machines needs to be accessible "from within" if their location is cut of from the outside world. In that scenario, one author instance is used to drive 70 public instances - on 70 different servers.

Another example is having 20 different authoring servers publishing to a single public server to present the content -- this is a typical university setup, where each department wants control over its own server, but the university wants a single public access point, preferably using a unified look and feel.

Monday, March 20, 2006

Simply Magnolia

In a press release from yet another venture burning open-source company we read:
Alfresco Announces Simple Open Source at BrainShare(R) 2006; Simple to Install, Simple to Use, Simple to Scale Out
Now this is so similar in its intent to what we announced for Magnolia more than 2 years ago that I am almost flattered. In its current incarnation, its worded as follows:
Magnolia makes content management as easy, fast and flexible as possible. Its streamlined features ensure easy deployment, easy templating and easy editing of your websites.
Magnolia's tag line once was
Everything Easier
And believe me, thats the way it will always be, no matter what the tag line says.

Sunday, March 19, 2006

Feels like 1999

This week I have hired a new employee (Daniel) on the spot, Pascal was in Spain and sold several Magnolia licenses to the Spanish Government to upgrade their current Magnolia site.

He came back Wednesday at 4PM - I left at 5PM for England, where I gave a 2 day advanced Magnolia user training in Ditcot for Diamond, who are building and running Great Britan's most advanced scientific research lab. The lab (and especially the synchrotron) will most likely become the largest Magnolia deployment in the world with a potential 70+ Magnolia servers - what a supreme reference. They are big Magnolia fans and they ordered licenses.

Friday evening an online order came in for an enterprise license (including CRX). This must have been the most successful week we had in the last couple of years.

Saturday Pascal and I finally had a chance to talk about last week's happenings and future plans. Magnolia 3.0 is going to rock the world pretty seriously.

By the way, we are still hiring.

Tuesday, March 14, 2006

Magnolia podcasting

Tom has continued his quest into RSS and now has added podcasting capabilities to the upcoming Magnolia templates. All you need to do is upload a couple of audio files, categorize them (so that the RSS generator has a chance to generate a feed that matches a given category) and Magnolia generates an audio RSS feed (podcast) out of it. Pretty slick.


I am in the process of organizing a class reunion, and a tool that provides a forum, some voting mechanism, a wiki, a foto upload option and an adress book would be very useful. So I'd expect to simply select the tools that I'd like to see, configure them and off we go. Add a sprinkle of RSS and mailinglist support, and the reunion is already half way organized.

So I stumbled across Ning, apparently one of Marc Andreesen's startups.

Its a pretty cool idea - a social webapp builder that is available for free as a hosted service. The elevator pitch is just what the doctor ordered - I have been looking for tools that match my mounting ad-hoc requirements (s.o.), and ning sounds like it could just pull it off.

But no sir, that would just be too simple.

I must say I am uttely unimpressed. Not only was it dead slow (who builds such a thing on PHP?), it crashed on ever second click (thats ok, its beta). The thing that is really annoying is that I could not figure out how to do anything except clone an already existing app, or delve into the dungeons of PHP (a browser that lets you manage all properties of the system is available).

Maybe I missed something obvious, but for the moment I stick to email and phone to organize my reunion.

Tuesday, March 07, 2006

snow record

I spend last Saturday with John in the office (see his blog). While we were running a 12 hour coding marathon to solve some pressing issues, it was snowing outside.

It was still snowing when we had some excellent pasta (delivered by Adriano on foot through the mounting snow outside).

When I finally went home, it took 3 cooks and well over 30 minutes (instead of the usual 7 minutes) to get home.

Within 24 hours, half a meter of snow came down on Basel - more then ever recorded. The sight was quite spectacular. So here are two impressions taken from my home.

We had some great fun that Sunday and build a huge half pipe in the garden, which is still mostly covered in deep snow today.

Next time when I meet John I better be prepared for anything ;-)

Wednesday, March 01, 2006

Linuxtag with Magnolia

At least one of my proposals for Linuxtag has been accepted - I will be presenting (the business benefits of) Magnolia at Europe's biggest open source conference in Wiesbaden. Looking forward to meet you there!

Monday, February 27, 2006

Amgen Tour of California

While I am busy trying to help one of the biggest outsourcing companies in the world to sell Magnolia to one of the greatest airlines of the world for their new intranet, here is what current users say:

Thanks, you guys made my day Magnolia! We just finished up the Amgen Tour of California,, and everyone loved it!

We were very impressed with how Magnolia handled under the load, even before I got caching working. We averaged about 580 hits a second, although that was an overall average and we probably had about 4 times that during lunchtime. This was all running on a single server per magnolia instance, an intel xeon 3.0 ghz processor with 2 Gb of ram.

Again, thanks for a steller product that really worked out! Feel free to contact me (...) if you want more info to maybe put on the magnolia website.

Is it worth 80h weeks? Maybe not, but it sure feels great to know that we make somebody's day out there.

Sunday, February 26, 2006

The Jackrabbit vibe

Too funny what other things are named Jackrabbit:

The Jack Rabbit Vibrator is an all time classic vibe. This wonderful and fun vibrator is loved by women around the world!
Luckily they don't carry a model named "Magnolia".

The JCR shell

Did I mention we need a JCR shell? John just wrote one. Can't wait to see it announced on the Jackrabbit list...

Magnolia training rocks - and so does Magnolia

Tom Jensen of Boise Cascade has writen a very positive review for our Magnolia trainings:
I just finished up yesterday at the training for Magnolia down in Laguna Beach, CA. I was glad I went and felt it well worth the time and money to go there. It was so nice to have a knowledgeable instructor (Giancarlo) to help understand Magnolia and how it relates to the JCR. In the past I've thought it a bit odd to have training for an open source project. I thought that I was a wimp if I couldn't just understand the complete concept of the software from reading any docs on it and looking through the code. The training not only will have saved me many hours of head scratching it has helped me to get the vision of what was intended with Magnolia and how to better use it along with the underlying repository.

I am very happy that Tom choose Magnolia. During the evaluation process it was openCMS vs Magnolia in the finals, and when the decision was made, Tom told me that it was pretty clear: they had the users test the two systems for a day, and while users were having a hard time figuring out how openCMS was supposed to work, the opposite was true when it came to Magnolia - they simply could use it without asking questions. The difference must have been striking. Tom sent me he following note:

This is a great product and I'm looking forward to continuing using it. It is not only a simple, intuitive program for end users (content creators) to use but is pretty straight forward in template creation.

Thats exactly what we intended to deliver from the start - looks like we pretty much succeeded.

Friday, February 10, 2006

Magnolia in the press

The first part of my article about open source community evaluation - and why it matters for your enterprise - has been published as a cover story in the Enterprise Open Source Journal this week. (See page 60 for my article). The theme of this edition is "Demystifying open source" - a great subject and about time bring to our attention, too.

What I like most is that the same issue mentions Magnolia as one of the 10 popular open source CMS systems - even if Ric Shreves has gotten some things wrong in the article:
  1. the licensing is not dual mode but we have several products, some commercial, some open source
  2. the community is quite extraordinary, with 200 people subscribed to the developer list and some outstanding external (to obinary) core members like Fabrizio (who is also committer for displaytag and Maven) and Alexandru (who has been contributing to a number of open source projects)
  3. there are no modules because the way magnolia works, functional modules are rarely needed. Since the data management is completely dynamic (no database tables to create and rows to add) you can simply define a data unit (paragraph) and are read to rock.

Monday, January 30, 2006

The bigger picture

I recently commented on "Should BPM be a part of ECM":

One of the core ideas behind Magnolia 3.0's integration of openWFE is exactly that - making Magnolia part of a potential enterprise wide process management system. This allows to create content as part of a wider process, possibly involving completely different systems and technologies (openWFE is the only open-source distributed cross-platform BPM I know of and includes options to talk to it via REST, .net, Java, perl, ruby and python (I probably missed a couple).

As an example, look at the processes for an e-commerce based community plattform. This might include creating events and allowing people to register to these. The creation of an event is more than adding a web page - it needs to be organized first, and publishing and announcing it is part of the bigger picture. After that, the registrations need to be managed, additional registration information might be sent out, participants might have access to specific information on the page etc. pp.

You quickly see the power - and necessity - to provide business process integration as part of your (content management) solution once you realize that content is there for a purpose, and does not exist in a vacuum.

Wednesday, January 18, 2006

Sameer found the first feature list for Magnolia. This must have been nearly 3 years ago now (I think we started coding in March 2003). As the chinese proverb goes "A journey of thousand miles starts with the first step" - enjoy!
Obinary CMS OpenSource

Feature list :

Version :1 (very basic)

User Administration
  • None
  • System treats all users as Anonymous
Site/Content Administration
  • Page creating with selected template (user interface)
  • Content editing within page (basic text and image editing tools)
  • Possibility to create 'n' dynamic container lists (multiple components / multiple paragraphs).
  • Image upload (page related or shared images which could be uploaded to the image db)
  • File upload (same as image)
  • Easy editing for multiple sites, multi lingual sites.
  • Easy sharing of content and JSP's within different templates and content pages
  • Site structure is not dependent on language or multiple sites , its entirely up to the authors
Page generation (text based)
  • HTML
  • WML
  • XML
  • Standard API to create and retrieve content, specified structure.. but not restricted to or dependent on any specific implementation (API handling content storage and retrieval in a heirarchical structure)
  • Servlets or JSP templates
  • Templates would represent only a way to store and retrieve/display data
Simple caching
  • Content caching (text / image)
  • Independent to any implementation, authors will be able to push/activate content to various locations
  • Not limited to any number of stages
  • Admin and live sites could be 2 separate stages

Sunday, January 15, 2006

Red Hat: The mother of all open source business models

Matt Asay in AC/OS: Red Hat: The mother of all open source business models provides some insight why he thinks Red Hat has such a great open source business model. With all of us struggling to find the real McCoy in open source business models, examples of things that actually work are rare and welcome.

Simply put, the difficulty lies in making money when people can download your software for free. Apparently one way to avoid the freeloaders (well, at least significantly increase the number of people that pay for your software) seems to be to provide the source but not the binary for free. This was a suggestion I made when we started out with Magnolia - simply said, I proposed to sell the installer.

The one thing I never figured out is how to make sure others do not simply take the code, compile it, and then provide it for free. In the case of Magnolia, anybody could provide an installer. The same is true for all other open source code - unless the license forbids you to actually resell a compiled binary, anybody could.

Of course, the question remains: who actually would? Well, it has been done before (remember how vTiger took the sugarCRM code base, removed the copyright, and rebranded it as vTiger?). These cases are rare, however. At the same time they are legal (except, removing the copyright notice was not).

Lets have a close look at what Matt says about Red Hat and why it works. If there is a fireproof way to create license revenue while providing the source code, I'd love to make all our products open source.

LinuxTag - my proposals

I have had a number of ideas for the LinuxTag call-for-papers - but as so often, time only allows to flesh out two of them - which I did.

My first proposal will be a magnolia product presentation for decision makers. This talk will focus on the things that make Magnolia interesting for the enterprise:

We will present an introduction to the business benefits of Magnolia, aimed at decision makers, not developers (but of course everybody is welcome).

The talk will include a product demo to give an impression of the usability benefits of Magnolia, but focus on providing examples of how Magnolia's enterprise aware design provide significant benefits for medium and large enterprises.

We will shortly introduce JSR-170, the new standard for content storage, that has been the basis of Magnolia development. Magnolia currently is the only open source content management system that has been built specifically on JSR-170, but several other projects are starting to see the benefits of a standard repository and are on its way to adopt it.

It will further present the new Magnolia 3.0 architecture which combines Magnolia with openWFE, resulting in the most sophisticated workflow and enterprise application integration possibilities of any open source CMS available today.

We will then present the major benefits Magnolia brings to your enterprise, and talk about some existing implementations to give an overview of where Magnolia is useful.

An overview of coming attractions will wrap up the talk.

My second proposal recylcles my previous investment into researching the evaluation of open source communities, which I did for my series of articles for the Enterprise Open Source Journal (released any day now):

This talk looks at "The Community" from an enterprise perspective. From the premise that "Open Source is more than access to code" we define what "The Community" is, why it is important for your enterprise, and how to gauge the liveliness of any open source project. We provide some real world numbers for content management system communities. It ends with an overview how your enterprise can benefit by becoming part of a "The Community" itself.

Lets hope this is the stuff the LinuxTag guys are interested in ;-)

Thursday, January 12, 2006

LinuxTag 2006

I am on the team preparing the Information Web track of LinuxTag 2006

LinuxTag is Europe's most important event for free and open-source software. In 2005, over 12,000 visitors experienced this unique combination of conferences, tutorials and trade fair. The twelfth incarnation of this great arrangement will provide up-to-date information to power users, decision makers and researchers, as well as to software developers, newcomers and the open-source community.

Seems this is going to be a great event. Only three days left for the call-for-papers. I hope I will be there (and talk about something Magnolia)