Some time ago, I mentioned that I love MacFUSE which lets you access remote systems over secure shell as if they are local disk drives on your Mac. MacFUSE is a Google Code project under the new BSD license.

I just upgraded my VMWare Fusion to 2.0 and one of the things they bundled with it was MacFUSE (1.5, though the latest is 1.7.) MacFUSE was 1.3 in January of 2008, 1.5 in April 2008, and 1.7 in June of 2008. MacFUSE continues to rev rapidly and is now being integrated into a commercial product.

Matt Asay points out that it takes time and work to contribute to an open source software project. What suprised me was that he seems upset (or perhaps disappointed) that it's so hard for an outside person to contribute to an open source software project.

First off, it's not hard to contribute small things. Small things that can be easily verified are easy to contribute. It's harder to contribute large patches and new features. That's because the open source software is built on credibility and trust. A newbie has a trust savings account of zero. They build credibility and trust over time. When an outsider submits code, that code is not just inherently trusted. It needs to be examined, tested, and verified before it's accepted. That takes time - time from already busy committers. But when the code is accepted, that outsider has won a trust point. Over time, their contributions will just be trusted and accepted quickly, not ignored because people don't have time.

Luckily for most open source software projects and the community, trust can be partially transfered between projects. "Your reputation procedes you" and all that.

I've spent a lot of my work time making it easier for people to make contributions to open source software (via OpenLogic's Expert Community and my new job as Executive Director of the GNOME Foundation), and I believe you have to work within the model. If you want it to work faster, you have to hire the people that already have the trust (assuming they'll want to work for you), or pay people to spend the time to verify the patches you have (some what like OpenLogic's model) or you have to spend the time to develop that trust.

The OpenLogic Expert Community (OXC) is a growing community of open source committers, contributors and experts who help us in supporting over 400 open source projects.  The community is open to people who have the passion, time, and knowledge to help solve technical issues on any of the open source products in our library.  Are you interested in joining?  Would you like to know more?  Here are just 5 of the top reasons you may want to consider joining the OXC today:

1. Compensation.  Get paid to do what you love.  Do you like solving configuration issues for Apache?  We normally pay $100 for each ticket you resolve.  If it's more complex we will pay more.
2. Promote Your Project. Whether your project has one committer (you!) or a hundred, we actively sell support to large enterprises on everything in our library.  By joining our community you help us promote the use of your project to enterprise users.
3. Do What You Love.  You probably are involved in open source because you are passionate and excited about your project.  The OXC offers you yet another opportunity to use your expertise to help others better use your projects.
4. Develop New Expertise.  You don't have to be a committer or contributor on a project to help us resolve a customer issue.  You may be a PHPAdmin expert who can help us resolve a generic PHP question.  If you have the skills you can choose to take on an issue on a new project to help you develop new skills.
5. Promote Open Source.  Help us get the word out.  Open source works!  By joining our community you help us create the positive image open source deserves.

If you are interested in learning more, please visit our community page.  Or join today by filling out the OXC application: http://www.openlogic.com/community/join.php.

I was reading Simon Phipps' recent blog on the funding of open source.  In it, he talks about how open source communities work.

…there is no fiscal power that any contributor holds over others, so no-one has the right to tell the others what to do. An open source community is an example of a group of people choosing to synchronise their mutual interests, each at their own expense, for the benefit of all involved including themselves.  

It reminded me of comments I've heard from others about open source vendors who "own" the associated community.  I find this idea of "owning" a set of open source developers somewhat distasteful.  What people usually mean when they say this, is that the open source vendor employs all or most of the committers for the projects and controls all commits to the project.  In some cases the vendor chooses not to allow certain commits because it doesn't serve the needs of the vendor. 

For example, a vendor may have an enterprise and a community edition for a project.  In some cases they only sell support on the enterprise edition.  Often they put all changes into the community edition immediately, and then pull a particular release of the community edition and do additional testing and certification and call this the enterprise version.  The vendor then sells support on this enterprise edition. So far that makes sense to me.

However, I have seen a few cases where a vendor who controls all the commits will withhold fixes and patches from the community edition and only offer them to the paying enterprise edition customers.  Because they control all of the committers, the vendor can then prevent others from offering the same fixes for the community edition.  When that happens, it doesn't seem much like an open source community to me.  It seems a lot like a proprietary vendor.

Has anyone else seen this phenomenon? If so, I'd be interested to hear what you think.

Our last webinar—"A Comparison of Open Source Reporting Tools for the Enterprise"—drew a huge amount of interest, so we decided to continue the trend with another technical webinar: "A Comparison of Open Source Scripting Languages for the JVM" presented by Rod Cope, CTO and Founder of OpenLogic. This webinar will be held on September 23 at 11:00 Pacific / 2:00 Eastern, and of course the recording and slides will be available for download after the event.

In this webinar Rod will discuss scripting languages for the JVM like Groovy, JRuby, and Jython, which have become popular alternatives to programming languages like Java, C#, and C++ as well as traditional scripting languages like Ruby, Python, Perl, and PHP. Rod will provide a comparison of key attributes for the most popular scripting languages for the JVM, including:

• Which languages are easiest (and hardest) to learn?
• What types of development are best suited to each language?
• How do the top languages compare in terms of ease of use?
• What are the strengths and weaknesses of each language?
• Which open source frameworks and other packages work best with each language?

If you're interested in learning a new scripting language to enhance or replace your Java development, or if you want to know which language is best suited to a particular type of development, please be sure to join us on September 23 at 11:00 Pacific / 2:00 Eastern.

[From the Open Source Census August newsletter.]

In the previous newsletter we looked at the breakdown of Linux Distributions. Many people have since asked us how open source software compares across different operating systems like Linux, Windows, and Mac, so we decided to take a look at the average number of open source packages found by operating system. As expected, Linux (an open source operating system) had the most with an average of 87 instances of open source found per scanned system. FreeBSD (also open source) was a close second at 81, but Mac wasn't far behind with 75. Judging by the large number of Macs seen at open source conferences like OSCON and LinuxWorld, there are probably a lot of Mac users who are open source fans.

Windows, although not open source, still had a respectable amount of installed open source software, with an average of 39 instances per scanned system. The number is probably lower than those for open source operating systems simply because there are already proprietary solutions included with Windows. So, in response to questions about whether open source software is found on Windows systems, the answer is definitely yes!

I was recently asked to evaluate virtual whiteboard solutions for a client and had the opportunity to look at the state of open source whiteboards. With one exception, it wasn’t a pretty scene.

By OpenLogic certification standards, nearly every open source whiteboard would have failed due to lack of activity, lack of community support and size (single developer.) If you’re reading this and have deployed an open source virtual whiteboard solution in your enterprise, please leave a comment - I would love to hear from you.

There was one notable exception in the wasteland of open source whiteboards, Dimdim.

Dimdim carries on a rich open source tradition of extremely bad product naming, but otherwise, looks like a jewel in the rough. In July 08, Dimdim announced receiving a $6M infusion, so theoretically, they should have some runway to really improve this product. Needless to say, it’s being actively developed.

Dimdim is an open source web-based meeting solution along the lines of WebEx, GotoMeeting or Raindance and is billed as a direct challenge to WebEx. While I’m not sure WebEx (or the others) have much to worry about yet, there are some compelling features of Dimdim that make it worth a look if you’re in the market for an alternative, especially an open source alternative.

First you can use the Dimdim service to freely host meetings with up to 20 participants. They also offer paid services for meeting sizes larger than 20 in their “Pro” edition ($99) and their “Enterprise” edition ($1998). Finally, if you want to host your own instance of Dimdim, you can download the GPL’d community edition and host it with no participant limitations or license fee.

Dimdim clients are Flash based and work on Windows, Mac and Linux, though initiating desktop sharing is only available on Windows and more recently, Mac. Linux users can see shared desktops, but can’t be the one sharing. If you have a webcam and/or microphone, those are supported pretty well by Dimdim. On my iMac, it picked up the camera with no fuss and the video feed is automatically integrated in the framework when you start a meeting.

Audio, like nearly all the commercial alternatives, is available embedded, but as with all the commercial alternatives, you’re better off not using it and calling in on a standard conference calling bridge. The sound will simply be much better if you don’t use the embedded audio.

Virtual Whiteboard

My main interest was in the virtual whiteboard feature for collaboration. I also wanted to make the whiteboard as natural as possible, so drawing with a mouse was out of the question.

I settled on using a Wacom tablet (Intuos 3) as an input device for the whiteboard since regardless of the whiteboard software, it could at least look like a mouse to the machine but be a much more natural drawing tool. The Wacom tablet has driver support for Windows, Mac, and Linux. The Wacom tablet Linux install is not the best and required some manual name editing of the xorg.conf file to get it working, but after that, it’s reportedly working like a champ from users I’ve talked to.

One of the things I really like about the Dimdim whiteboard was that it is multi-paged. You definitely need it, too, because it’s not resizeable and I think it is a shade too small for how a whiteboard is typically used. The second thing I really liked about the whiteboard is that when you leave a meeting or log out as a host and then resume that same meeting later, the whiteboard contents are still intact. You don’t need to explicitly save the contents. If you’re like most tech companies, your conference room whiteboards are constantly locked down in “Do not erase” mode, so having a virtually expanding, page-able whiteboard is nice.

Now for some of the first impressions on using a virtual whiteboard. My litmus was “how natural is it compared to the real thing?”

If you’re prepared to practice a little bit, you can get decent results in short order, but it still won’t be the same as grabbing a marker and whipping out diagrams. Before you present the first time, you should practice to see what works best for you.

Here are some Dimdim whiteboard pitfalls and best practices I found which were helpful in achieving more whiteboard-like results.

Don’t use the geometric tools if you plan to annotate the shapes.

This warning comes from the issue of weak selection capability in the whiteboard editor. It’s most easily demonstrated with this example:

If you draw a small shape and then encapsulate it with a larger shape, you can no longer select the small shape without moving the larger shape away from it, selecting it and then moving the large shape back again. This is obviously a pain.

It gets worse when you try to annotate the shape with text because the text area they give you will not resize to the extent of the text and therefore can radically overlap multiple elements of your drawing as in this example:

The text in the box that’s selected is supposed to be in the upper left box, but the text area will overlap many other elements making them difficult or impossible to select without temporarily moving things out of the way.

For me, the Dimdim whiteboard came into its own when it was used with the Wacom tablet in a freehand mode (not using the geometric tools or text tools.) Despite the text tool’s limitations in annotating a shape, it can do quite well for forming tight, readable lists. Using this combination, you can quickly achieve very whiteboard-like results and it’s not clunky or frustrating to use.

Here’s an scenario. Lets say you’re whiteboarding a diagram and out of the discussion an issue comes up which needs further consideration, you can’t deal with it right now, but you don’t want to forget it. The typical method is to list it in a parking lot. Here’s a contrived example showing the start of a diagram using freehand methods with a Wacom tablet, but using the Dimdim text tool for what it’s good at, tight text - in this case the Parking Lot list:

You could even add another whiteboard page and put your parking lot items on it instead so it doesn’t clutter up your diagram, but you don’t lose the info. With the Dimdim whiteboard, it’s fast to flip back and forth between pages.

Whiteboard Limitations

Some limitations of the Dimdim whiteboard that I found were:

You can’t copy and paste between whiteboard pages
You can’t duplicate a whiteboard page

Both items are supposedly in development now according to the email I got when I submitted an enhancement request.

You can’t paste an image into a whiteboard. I would love to be able to do this because often you want to be able to collaboratively annotate an image of some sort - say a map, or existing diagram. The way you’d have to do this in Dimdim is to have a presenter share a desktop with a real application running - pull in an image and then annotate it. Again, not being able to hand control of the desktop to another participant remains a collaboration limitation.

Dimdim Limitations

Desktop sharing is still not up to par with the guys like WebEx and GotoMeeting who have been doing this gig for many years. I think this may be the #1 issue that will affect the rate of adoption of Dimdim. When I say it’s not up to snuff, I mean specifically that remote people who are looking at the shared desktop will hardly ever see any interaction, but rather slow screen paints of a screen once it has settled down on the presenter’s side. There’s significant latency. I also didn’t see a way to hand control of the desktop to another participant.

I think the best it could be used for currently is to present static slides (PPT or PDF) or very static screens. If you’re trying to present interactive pages or user interface ideas or interactions, I can’t see how Dimdim would possibly fill your needs.

Dimdim Strengths

The strengths of Dimdim are that 1) it’s very easy to start an account and host a meeting 2) the process to join a meeting could not be simpler 3) there are many hosting configurations to suit a variety of needs 4) It’s free to very affordable depending upon your meeting sizes.

Would I trust Dimdim to present a product rollout to 500 sales people? No. Would I trust it for an internal collaboration tool? Certainly. Perhaps over the next year or two, it will be a worthy competitor to the commercial web-conferencing tools.

About me

I’m an independent consultant who used to do a lot of work for OpenLogic. I greatly appreciate OpenLogic and the mission they are fulfilling as well as their willingness to let me contribute to this blog. Views expressed here are not necessarily those of OpenLogic and any mistakes are 100% attributable to me. You can contact me at: landon at 360vl dot com or visit http://sawdust.see-do.org

I uploaded some of the Open Source Census data into Many Eyes and I've been having great fun playing with the data. (You can play with it yourself.)

Here are some of the visualizations I've created:

Word map:

Bubble map:

At OpenLogic, we've recently been talking to various venture capital firms about our business model.  These discussions have been interesting because OpenLogic's pioneering business model doesn't match that of the "typical" commercial open source vendor.  As we've been thinking about how to best communicate our business model as an open source aggregator, I have noticed that "pioneering" and venture capital don't always go together.

That may seem surprising — since venture capital firms are essentially paid to place bets and take risks.  But, VCs also want to manage risk by placing bets on ideas that have the best chance of succeeding.  In many cases, VCs will place bets on business models that are the same as already proven business models.  In the case of open source, they look to the business models of startups like JBoss, MySQL and Zimbra — companies that focus on a particular open source technology or application. 

There's nothing wrong with that approach.  If I were a venture capitalist, I would also look at what has worked in the past as a way to find ideas that might work in the future.  I'm there will be many more successful companies following in the footsteps of those open source pioneers.  However, it won't help VCs find the new pioneers.  Often the companies that blaze the trail on a new business model reap significant financial benefits over the next wave of followers.  Companies like RedHat, Salesforce.com, ebay, Facebook can provide a significant premium to investors because they figure out a business model that works and are able to optimize it and capitalize on it ahead of others. 

The question then becomes, how do you pick the new pioneers? 

Although I've posed the question, I don't think there is a formula.  I think there is a good approach to finding pioneers though.  For VCs that want to invest at least part of their portfolio in pioneers, filtering business ideas and models through the lens of customers offers the best opportunity for success.  Which ideas address real customer pains or needs or desires?  Which business models provide reasonable ways to address those customer pains? 

I believe this requires VCs to temporarily remove the filter of "what has worked before" in order to determine "what will work next". 

What are your thoughts?

I burned an Ubuntu Hardy live cd with Brasero (the GNOME default burning utility) the other day and it had problems booting.  Kept getting some error about SQUASHFS over and over again and never go to the login screen.

After seeing some hints about a bad burn and some unconfirmed discussion about TAO versus DAO, I burned a new copy with K3B (the software I normally use).  This disc worked perfectly.  Not sure what the difference may be in the default burn settings, but I now have very low confidence in Brasero.

Anyone aware of issues in Brasero?