remote
     synthesis

What is an Open Source Project?

Posted on Oct 23, 2006

Well, now you all have gone and done it. I put up my open source list close to one year ago (originally it was just a blog post) with about 60-70 projects, perhaps less. Back then, there were so few projects I managed to avoid having to answer the basic question of what constitutes an open source project. However, recently there has been an explosion (of sorts) of new CF open source projects, and the launch of RIAForge has even brought several new ones out into the open that every week I have found more and more need to excercise some form of editorial control over what appears on the list and on the updates. I keep having to ask myself, "Is this an open source project?" The answer comes back feeling too arbitrary and subjective for my own comfort.Is a Custom Tag an Open Source Project?
Back only a couple years ago, the only open source available for CF was some custom tags you might find at the Exchange and other sites devoted to custom tags. Even then, many of these were not free and/or not really open source. For the most part, the scopes of the projects were small, and they usually did something you could easily build yourself, but why reinvent the wheel? However, when I first created the list (even when it was just a list of favorites for myself), what constituted an open source project was generally *not* custom tags. I felt if that were the case, the list could easily become just another custom tag library...and there wasn't much value in that, which is why, for instance, I don't even list my own sortable tables custom tag.

Is a Component an Open Source Project?
However, if a custom tag is not an open source project (which isn't so cut and dry either), is a component one? There are actually several projects on the list that are comprised of little more than a single component, and several others that I have chosen not to add that are the same. For example, AjaxCFC, JavaLoader, Mailer (not to pick on these projects and this isn't a criticism) are generally comprised of little more than a single CFC.

Is an Add-On to an Existing Project an Open Source Project?
So, a custom tag is not a project, but a component can be but based upon what criteria? Now take the example of some projects on the list that are add-ons to existing projects (several for BlogCFC specifically). These projects cannot be used independent of the project they were written for, so do they qualify as a project on their own? Well, in many cases, I have chosen yes, but not in all and I am not clear on what makes the difference.

What If We Just Said Yes in All the Cases Above?
As I have stated before, one of the things that makes the list of value is that there is some editorial control involved. However, I don't intend for that control to be employed arbitrarily. Nonetheless if we said yes to all above cases, wouldn't the list lose its value? I think so, because my gut tells me that not every one of the above cases meets a certain criteria to qualify as a legitimate project, and yet relying on my gut instinct sounds so George-W-Bush-ian...so lacking in substance...so arbitrary.

Implications for the Broader CF Community
I know I have focused on how this question pertains to my list, but I do think it has an impact on the broader ColdFusion community. For example, what projects belong on RIAForge, CFCDev, ColdFusion Cookbook or CFLib (to name only a few examples)? If that isn't clear, do we risk fragmenting our efforts, drawing resources away from equally valid and valuable projects? In the case of my list, where I set some arbitrary test for determining what is a project, do I risk overlooking a valuable asset? In the case of RIAForge, by letting everything in, do we run the same risk of overlooking a valuable asset simply by overwhelming visitors (I know its not that way yet, but take a look at SourceForge for example)? Is there a happy medium, and if so, what is it? Can I stop phrasing everything in the form of a question, Alex?

Comments

Sometimes new technology shrouds old issues.
How would you solve this type of problem if you were a magazine publisher? OK, maybe "specialty newsletter" instead of magazine.

You might be saying to yourself "Geez, I got into this with just a blog post, and now I'm making all these decisions."
You might not rue it so much if you were getting rewarded for your effort.
So you've created something out of thin air, great!
You're just going through the growing pains, that's all. It's good problem to have!
Build it into some kind of revenue stream and then sell it.
Money is a very exact unit of measure for how much something is worth.
Profit from Open Source projects? It sounds underhanded!
We had this debate last generation, Communism vs. Socialism vs. Capitalism. Guess what? Capitalism won.

Disclaimer: It's getting late. I should go home.

Posted By Phillip Senn / Posted on 10/23/2006 at 7:46 PM

Hi Brian,

I think it is good that you ask the community for feedback, but I think in the end it is up to you to exercise editorial control in the way you feel fit. If you want to smote a project just because the author looked at you funny, go get smoting!!!

Seriously, every publication has a tone set by the editor. You cannot list every CF resource of value as if there is no filtering all you're offering is a Google search.

I think you should consider what you want YOUR open source list to be, you should let us all know what the style and the flavor will be and you should go ahead and do that.

If you do something nobody else wants to look at, you shouldn't mind - it is still your list. If a bunch of people happen to enjoy and share your editorial taste, then it will be popular.

Market research is appropriate and valid, but at the end of the day, the best publications are the ones that the authors are passionate about. What problem is the list solving now that RIAForge is out there? Discoverability? Filtering? Editorial comments? I think discoverability will become less important as the number of venues and resources grow and filtering and editorial will become more urgent.

Maybe there should just be a master list people can self submit to and a "Brians picks" or a set of categories you or the author assign the content to? Perhaps a project of the month which is more indepth?!

What did you first enjoy about creating the list? What was fun and cool for you? What sucks about maintaining the list? How can you direct the list to be more focused on the former and less on the latter?

I look forward to the list however it grows as I'm pretty sure whatever you do with it will be something well worth reading!

Posted By Peter Bell / Posted on 10/23/2006 at 10:28 PM

I keep forgetting to post a comment each time you post something new about the list, but great job Brian, keep up the good work!

A few years ago I was involved with a regionally popular car club in Colorado and helped manage it and what not. I was one of the first members of the club, but it eventually grew out of control to almost 400 users/members the last time I checked. This doesn't sound like much, but as you can imagine there were many passionate people on the listserv with their own opnions, etc. There was always some drama going-on, so eventually we got to the point where we laid out several basic and clear cut rules. You could be removed or put on moderated mode on the list for violating any of those terms OR at the club leaders discretion. Basically we left it up to the leader to make the final decision while a board of about 15 of us "old timers" from varying backgrounds provided feedback to the leader about various items. Some were unhappy about decisions made, but at the end of the day I always felt our decisions were fair, and anyone could bring issues to the board without getting scorned, etc.

Our common response to trouble makers who'd done enough wrong things to get removed permanently from the club were basically along the lines of "you're free to start your own club and run it the way you want to". Unfortunately you can't cater to everyone, so don't let outside influeces sway you too much. What you have is great, and I hope it doesn't eventually turn into the Allaire Gallery because it was impossible to find anything useful on there by the time CF 4.5 was out because of all of the "AAAAAAA-EZ-CART" type crap on there filled with useless keywords since their $900 custom tag did everything.

I'm paraphrasing here, but I believe Bill Cosby said something once to the effect of "You can't make everyone happy".

Posted By Brandon Harper / Posted on 10/24/2006 at 12:00 AM

It's a tough call. It's partly why I've decided to pull all my "small" projects into one Google Code project but then list them separately under RIAForge (Concurrency for CFMX, Closures for CFMX, IRCBot). I sympathize with your quandry tho', because part of the problem with the ColdFusion Exchange in my view is that there are too many tiny one-off code fragments that were never developed, never had proper documentation and, most important of all, were never "open" in the first place. I think as long as you ensure everything on your list is genuinely "open", then inclusion is right although you need to perform constant reviews (unfortunately) to ensure that projects that don't get updated (or downloaded) are aged off...

Posted By Sean Corfield / Posted on 10/24/2006 at 12:19 AM

As someone who put up one of those small components on RiaForge just recently I think it is perfectly reasonable to not list it here and for your Open Source list to be moderated. I wouldn't really expect my little component to be here. Maybe if it was larger in scope, had better examples and some documentation it would make more sense.

I guess I always thought of this list as applications, frameworks, and utilities to help build those applications/frameworks. Right or wrong that was my impression.

Just noticed you have an RSS feed of the list, and it looks by most recently added, which is very cool in my opinion.

Anyway, keep up the good work.

Posted By Joshua Cyr / Posted on 10/24/2006 at 8:38 AM