The relative value of open source to open services

There’s an active debate going on in the activeCollab community stemming from the announcement that the formerly exclusively community-backed open source project will lose much of its open source trappings to go commercial and focus a closed platform providing open web services.

For those who aren’t aware, activeCollab was created as a free, open source and downloadable response to Basecamp, the project management web app. In June of last year, the project founder and lead developer, Ilija Studen, offered his rationale for creating activeCollab:

First version of activeCollab was written somewhere about May 2005 for personal use. I wanted Basecamp but didn’t want to pay for it. Being a student with few freelance jobs I just couldn’t guaranty that I’ll have money for it every month. So I made one for myself. It’s running on my localhost even today.

Emphasis original.

Ilija offered many of the usual personal reasons for making his project free and open:

  • Learning.
  • Control.
  • Establishing community.
  • Earning money.

Now, the last one is significant, for a couple reasons, as was pointed out at the time of the first release: Ilija wanted to make money by offering commercial support and customization on a product imitating someone else’s established commercial product.

But competition is good, especially for my friends in Chicago, and they’ve said as much.

But, Ilija made one fatal mistake in his introductory post that I think he’s come to regret nearly a year later: I find it normal to expect something in return for your work. activeCollab will always be free.

And so a community of Basecamp-haters and open source freeloaders gathered around the project and around Ilija, eager to build something to rival the smug success of Basecamp, something sprung from the head of the gods of open source and of necessity, to retrace the steps of Phoenix before it (later redubbed Firefox), to fight the evils of capitalism, the injustice of proprietary code, and to stave off the economic realities of trying to make a living creating open source software.

For a little under a year, the project slogged on, a happy alternative to Basecamp, perfect for small groups without the ability to afford its shiny cousin, perfect for those who refuse to pay for software, and perfect for those who need such collaboration tools, but live sheltered behind a firewall.

A funny thing happened on the way to the bank, though, and Ilija realized that simply offering the code for people to download, modify and run on their own servers wasn’t earning him nearly enough to live on. And without an active ecosystem built around activeCollab (as WordPress and Drupal have), it was hard to keep developing the core when he literally was not able to afford continuing to doing so.

Thus to decision to break from his previous promise and close up the code and offer instead an open API on which others could build plugins and services — morphing activeCollab from a commodity download to a pay-for web service:

Perhaps I am naive, and this was the business model all along. i.e. Build a community for the free software during early development and testing, then close it up just as the project matures.

That was not original plan. Original plan was to build a software and make money from support and customization services. After a while we agreed that that would not be the best way to go. We will let other teams do custom development while we keep our focus solely on activeCollab.

But, the way in which he went about announcing this change put the project and the health of his community at risk, as Jason pointed out:


I’m a professional brand strategist, and while nothing is ever certain, I also feel that this is a bad move.

Essentially you’ve divided your following into three camps. For, against and don’t care. A terrible decision.

What you should have done (or should do… its not too late)__

—> Start a completely seperate, differently branded commercial service that offers professional services

—> Leave your existing open-source model the same and continue to develop the project in concert with the community


Sugar is not a great model to follow. It’s not.

A better example would Bryyght[dot]com, a commercial company hosting Drupal CMS. The people there are still very actively involved in the original open-source project.

Overall, you should choose your steps wisely. While you’re the driving source behind the project – NOBODY fully owns their own brand.

A brand is owned by the community that are a part of it. Without customers, a brand is nothing.


A brand is owned by the community that are a part of it. Without customers, a brand is nothing. (Hmm, sounds like the theory behind the Community Mark).

I think JH has a point, and with regards to open source, one that Ilija would do well to consider. On the one hand, Ilija has every right to change the course of the project — he started it after all and has done the lion’s share of work. He also needs to figure out a way to make a living, and now, having tried one model, is ready to try another. On the other, closing up the core means that he has to work extra hard to counter the perception that activeCollab is not an open source project, when indeed, parts of it still will be, and likely, won’t be the worse for it.

That many of the original Basecamp haters who supported Ilija’s work have now turned their anger towards him suggests that he’s both pioneering a tribrid open business/open service/open source model and doing something right. At least people care enough to express themselves…

And yet, that’s not to say that the path will be easy or clear. As with most projects, the test is now how he manages this transition that will make the difference, not that he made the decision.

All the same, it does suggest that the open source community is going through an evolution where the question of what to be open about and with whom to share is becoming a lot harder to answer than it once was. Or at least how to sustain open source efforts that play into facile operation as web services.

With the Honest Public License coming in advance of the GPL v3 to cover the use of open source software in powering web applications and services, there are obvious issues with releasing code that once you could count on being tied to the personal desktop… now with the hybridization of the desktop/internet environments and the democratization of scripting knowledge, it’s a lot harder to make a living simply through customization and support services for packaged source code when you’re competing against everyone and their aunt, not to mention Yahoo, Google and the rest.

Steve Ivy asked a poignant question in his recent post on Open Source v. Open Services: If the service is open enough, what’s the value of the source?

Truly, that is a question that I think a lot of us, including folks like Ilija, are going to have to consider for some time to come. And as we do consider it, we must also consider what the sustainable models for open source and open services look like in the future, for we are now living finally living web service-based economy, where the quality of your execution and uptime matter nearly as much, if not more, than the quality of your source code.

Author: Chris Messina

Head of West Coast Business Development at Republic. Ever-curious product designer and technologist. Hashtag inventor. Previously: (YC W18), Uber, Google.

14 thoughts on “The relative value of open source to open services”

  1. Regarding “If the service is open enough, what’s the value of the source?”…

    In theory, if the service is “open enough” (i.e., there are ways to export all your data) *and* people are diligent about backing up their data *and* if there’s a close-enough substitute for the service that can directly use the exported data, then the service should be “open enough” for everyone.

    Having the software backing the service available as open source dramatically increases the odds that there will be a close-enough substitute. After all, anyone could grab the code and start hosting a clone of the service. Not all users of the original service may have the skills to launch a clone, or even run a private copy, but they still have the option of hiring somebody to do the same. So, if you use a Drupal-powered service, knowing that you could re-host somewhere else running Drupal and have a decent shot of getting your templates and content over in one piece is a good thing.

    Sans the source, it may be much more difficult to do anything with the backed-up data. Just because such-and-so service gives Aunt Millie the ability to dump her data as a .tar.gz blob of XML files by writing her own script to invoke a bunch of RESTful URLs doesn’t mean that she will somehow be able to do anything about it. Heck, it may be tedious for somebody *with* the requisite tech skills to do anything meaningful with the data.

    Certainly fewer people will care about open source given an open service, particularly those for whom “open source” means “Who-hoo! I don’t have to pay!”. For those who look to open source for technology continuity, though, open services beat closed services but still aren’t necessarily “open enough”.

  2. Good points Mark, especially from the community side of things.

    How does this situation though (effective data lock in through lack of effectual data portability) affect the developers trying to make a living? i.e. is it okay for them to create this kinda of de facto lock in (your data can be exported, but it’s hard to use it) given their need to find a way to make a living?

  3. Hmm, under which license was ActiveCollab released? Couldn’t the community simply fork and continure development on a Open Source version. Similar to what happened with Mambo/Joomla.

    I also wonder of Ilja can even legally close the source. If it was licensed under an Open Source license like the GPL and he accepted code contributions from others, these contributions were surely done under the GPL as well. He could have a hard time to remove all code he doesn’t wrote him self.

  4. There were folks who suggested forking a la Mambo/Joomla, but that gets to my point: you’re suggesting that, instead of Ilija being able to make this into a business and leverage the value of his work, people should take his code, fork the project, and leave him in the lurch, having already invested a great deal of time and effort and having no income to show for it.

    I’m not bitter about Ilija changing his model — I support him.

    In terms of closing the project, Ilija’s said that very little outside contributions ever made it into the project, so he can either gut those pieces and clean it of outside IP or have the copyright owners assign their work to him or his new company.

    What I find curious is that you’re not suggested ways for Ilija to make money off his work, but instead suggesting ways that I can leverage the value he’s created without paying for it. How is he supposed to sustain himself if we can’t figure out a model that works for both open source advocates and the developers writing the code?

  5. I did not suggest anything in that direction because that was covered already in the article and the other comments.

    I’m an Open Source software author my self, and no I can’t live off it. You’re right about that we should find a way to find a business model that works for open source authors. But closing the source is for sure not the way to go (well at least for me).

  6. I think it’s important to note that Illija never let anyone actually submit patches and no one else ever had commit privileges. He seemed to always present patches in a way that made it seem like he’d changed them before committing. There wasnt’ a lack of interest or community, he just didnt’ let anyone do the work they were asking him to let them do.

    Really, this was just a community driven closed-source application (except insofar as it’s a script and thus user-viewable) from the beginning, and Illija just made it official after. The problem is that it did and still does say “open source” on the front page, which is why it got so popular and HAD that community (Dreamhost started offering it among it’s 1-click open-source installs, which is an example of a serious advertising benefit derived from aC’s open source status).

    I think it’s also important to note , on the subject of potential forks, that Illija literally said that he was dissatisfied with the project as it was headed (up to the 0.7.x mark) and was re-rewriting a lot of it to have a better plugin api etc for the 1.0 release, which would be the next one. Thus, by closing the source at 0.7.x and leaving only the old version under HPL he effectively says “yeah, sure, fork, but it already sucks and you’ll just be cleaning up my mess again”. A lot of people feel that this is a bit of a cop out because they’ve been submitting bugs and working with him to improve the sketchy pre .7.x versions only to have the actual fruit be unnavailable to them without a license.

    Also, is it just me, or is aC really slow? I actually hope that someone just rewrites the whole thing but better.

  7. Good article Chris.

    As Jason stated, Ilija should take his commercialization to a project called “ClosedCollab”, not closing the gates ActiveCollab could and would open.

    Ilija is trying to seize easy money. Matt (Photo) did not. The projects are quite similar and both have a potentially large user base. Buuuut, I am guessing Matt will be the one diving in the Money Bin.

  8. Hi Chris,

    what a nice article and interesting comments.

    I just want to let you know that there is another fork of activeCollab 0.71 called ProjectPier. I am one of currently 5 people working on that.

    Right now we are collecting all 0.71 themes and translations and rebrand aC 0.71 to PP to PP 0.8 which will be our first version. Additionally we are working on a Drupal based website at to replace the (ugly) existing website at

    In the next version we like to enhance the tool and simplify it on it’s rough edges 🙂
    If you like to join, please contact us through the Drupal web site. You will be very welcome.


  9. I also forked ActiveCollab into RailsCollab. In a way i can sympathise with Ilija, as it took me quite a while to port everything across to Ruby on Rails. However i still think he greatly mis-represented ActiveCollab by claiming it was going to always be free.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: