Slashdot Mirror
Ask Slashdot: What Defines Success In an Open Source Project?
rbowen writes "Nine years ago, Slashdot readers discussed what makes an Open Source project successful. The answers were varied, of course. An academic paper summarized the results, agreeing (albeit with more precision) that motivations for Open Source projects are varied. Has anything changed since then? In the era of mobile apps, social media, and Google Ad revenue, have the definitions of Open Source project success changed at all? Have your reasons changed for being involved in Open Source?"
Your project is a success when a corporation embeds it in their product and violates the GPL.
I think widespread usage is a good metric and not just gloating over profit like the Apple fans like to do. "Apple derived the most profit from the cell phone industry." they say, to put down Android's usage gains. By that metric, IIS is totally killing Apache and Nginx in the web server space, but most folks consider Apache beats IIS. Which of this is true?
This space for rent.
If RMS takes credit for the project and insists that everyone put "GNU/" in front of the name.
If so, you're not done yet. If not, find another itch to scratch.
Give me Classic Slashdot or give me death!
Nothing will make an Open Source project successful the way an endorsement from Alexander Peter Kowalski will.
On a personal level, your open-source project is successful when it accomplishes everything you set out to do with it. On a non-personal level, widespread usage is probably the best metric.
My blog
When respected authorities begin to compare you directly to the commercial alternative, even if you're still found somewhat wanting, you have arrived.
What political party do you join when you don't like Bible-thumpers *or* hippies?
I believe RMS said it best when he declared the following metrics required for FOSS project success:
1) To crush your enemies
2) To see them driven before you
3) To hear the lamentation of their women
For a good example of this, check out how Android has dominated Window Phone 7 and how their womenfolk continually spam Slashdot with first posts about their crushed dreams.
Mod me down, my New Earth Global Warmingist friends!
I see what you did there. Troll better.
It's a success once you can profit from all the labor that was invested into it. Which is hard to do with a project that "never ends." Corporations seem to get the profit right away because they don't actually invest their labor into the problem.
More people eat McDonald's Double Cheeseburgers (or whatever they're calling them now) each day than eat a nice medallion of filet mignon.
I know which beef product is superior, and it isn't the salty pink slime-infested crap off the dollar menu.
(Not that I'm defending IIS; hell no. Apache might suck, but IIS's vacuum will one day cause the universe to implode upon itself.)
1. Did it solve the original problem it was intended to solve?
2. How many other people had their problem solved by it? (usage stats, as much as possible)
3. How many other people were motivated to improve it? (got involved as developers, testers, documenters, etc)
4. Did it reach a point where it was so darned useful and bug-free that nobody really needed to think seriously about the problem ever again? (e.g. GNU's "bc" utility, which hasn't changed since 2000, and does its job beautifully)
The ultimately successful open source project goes through a lifecycle of something like:
1. solve an immediate problem
2. get developers, testers, documenters involved solving the problem in a wider context
3. solve the problem for a whole lot of users
4. nobody thinks any more work is needed
I am officially gone from
When a project evolves into that state where developers and users get along an coexist peacefully, then you have an environment that benefits both groups. It seems like a simple social skill, but actually this is rather rare. I have been in a couple of projects, one where the users and developers have something of an acid relationship and have a confrontational nature. Little gets done, and nobody is happy. But in the other one, users, developers, and other contributors (I18N, addons, builds, examples, etc) all get along harmoniously and produce a wonderful product. The producer/consumer model does not work in open source projects. Mutual respect and courtesy are the key to getting the job done. This also includes upstream library developers, distro managers, etc.
The exact same thing that defines success in non-open source software: It does what you wanted.
Doesn't matter whether it's a log rotation script, a web app, a POS system or firmware for electronics on the next spaceship. Software success is determined by only one metric. Open source doesn't enter into it.
I'd say, when your open-source project is better because you made it open-source, it's successful. My own minor projects have been improved by testing, bug reports, bug fixes, and new feature contributions through the years. Certainly they are more useful to me now than they would have been if I was the only one working on them.
PHEM - party like it's 1997-2003!
Success is when you have reached the goal you have set, nothing more, nothing less.
An open-source project can be success even if it has NO users whatsoever outside the developers, and equally well it can be a failure even if it had 200,000 users. An outsider cannot really say whether a project is a success or not, it's the developers who has that say.
Because Redhat makes no money from their nearly $1,000,000,000 in revenue per year and you can't pay them to support their OSS projects.
Easy: A release v1.0
You know you've arrived when you've been sued for patent infringement.
Never shake hands with a man you meet in a fertility clinic.
... describing the software you created.
That would be a Dr Hook moment.
Meanwhile, Red Hat has now a yearly revenue of one billion dollars (source). Clearly OSS doesn't work.
Dilbert RSS feed
You're a troll, but both responses about RedHat completely missed your point.
This.
Knowledge is how to play a game, intelligence is how to win, wisdom is knowing what game to play.
When your corporate competitor is scared enough to threaten you. That's how you know.
They're using their grammar skills there.
When the entire industry goes to your product first to fill the need, then you have succeeded.
....start acting crazy.
I think that continued development is hugely important.
So many projects are started and then quickly become stagnant.
Reach your goal, and continue to refine it.
When a how-to book about your project can be found in the computer book section of Barnes and Noble. Bonus points for making it to the "For Dummies" series.
...and what a wasteland of failure lies before us... :)
Even better is when they see a presentations by your users at a conference and their marketeers show up in force at those same users' offices first thing the next week. :)
Being compared is one step. Winning many of the comparisons is another. The marketing staff recognizing your project as a threat is yet another.
I would say under that metric there is Linux, Firefox, VLC, Bsnes, Dolphin (Wii Emu), Pidgin, SumatraPDF, Filezilla, Blender.... The rest of the FOSS world has some way to go.
I'd say there are a few that are getting close: Gimp, Ardour, Libre Office, OpenShot.....
The bummer is how many FOSS games are just not good enough. It's the year 2012, and there's STILL not a better FOSS Civ 2 than the original Civ 2. Almost all of the best emulators are FOSS though, so.....
Also disappointing is the audio apps. Winamp 2.81 is still the best on Windows. Audacious is good on Linux, but there's no Windows version, so.....
I hold very few opinions. I hold information based on observation and fact. If you wish to disagree, please use facts.
Also disappointing is the windows> audio apps. Winamp 2.81 is still the best on Windows. Audacious is good on Linux, but there's no Windows version, so.....
FTFW. Windows appears to be the *only* operating system that is still stuck with horrible music software.
When there is not a superior commercial product.
xeyes?
SJW n. One who posts facts.
Have gnu, will travel.
you know exactly how popular it is based on how many licenses have sold.
In related news, gonorrhea judged a favorite based on penicillin sales.
Have gnu, will travel.
Another key feature of most of the successful FOSS projects GP mentioned is that they are all cross platform. If you don't have a Windows port, you're not going to be successful (this is considering only desktop apps). If only KDE had taken the time to port KOffice to Windows. They'd be miles ahead of LibreOffice and associated application now. But because they kept it exclusive to *NIX until quite recently, and even now have an awkward installation process, noticeable foreign looking window theming and tons of bugs, they'll never get traction against project with larger install bases. The more widely distributed your application is, the more users are using your app and promoting the features they like, more bugs are experienced and reported, more developers are likely to notice your project and want to help out, and more people are likely to hear about it, eventually the momentum snowballs into a Mozilla Firefox, or Apache Server or some other popular FOSS project. Refusing to port to Windows is a sign of either bad initial design decisions, or a dev team that doesn't realize the benefits of a huge userbase.
A lot of FOSS projects have Windows ports, they just don't get a lot of love (like the Audacious Music Player mentions previously). Whatever your feelings about Windows as a platform or as a product of Microsoft, the huge userbase gains should be reason enough to put a lot of love into the Windows port.
Try the Foobar2000 audio player. It's almost open source; freeware with OS components.
It pisses on WinAmp for memory footprint (I still use it on a 14 year old P3, 384mb, while running a host of other programs), plays everything I chuck at it, and it feels like a FOSS application (no gimicks, no fancy graphics, no bloat).
This post contains benzene, nitrosamines, formaldehyde and hydrogen cyanide.
A name that no one can pronounce!
- How's your FOSS project doing? :)
- xcxcczgfhkklngs! Is a huge success!
Defining Statistics and Social Research
I'd say there are a few that are getting close: Gimp....
Not close. Not close at all.
Certainly not for pro or semi-pro use.
Sigs are so 1990s. No way would I be seen dead with one.
I write a project that addresses the needs of a very small population (I won't link to it, because I don't want to slashdot my server).
I only have a few dozen users, and we all agree that it is a rousing success.
If it meets the needs of its intended audience, in its intended scope, then it is a success.
"For every complex problem there is an answer that is clear, simple, and wrong."
-H. L. Mencken
The problem is that most people do it for fun, and it's just not a lot of fun to work on something you're never going to use and that requires you working in an environment you'd rather not work in.
Firefox's goal is to have an open web, not to be first in the browser market or even have a single user. Open Java's goal wasn't to have any usage of their product, but to threaten to enough to open the main (then Sun's) Java implementation.
If corporations were legally responsible to their charter (rather than maniacal about profit), the world may see successes that are equally as good for society.
Science & open-source build trust from peer review. Learn systems you can trust.
Success is measured by the number of forks the project has on GitHub.
That's not entirely true. Sure windows may have 50-100 times as many users, but they sure as hell don't have 50-100 times as many developers willing to donate their time for free, otherwise windows would have a decent audio player by now!
I use Gimp professionally daily. For all kinds of web work, fliers, biz cards, posters... The only thing I miss from photoshop is cmyk and the heal brush.
I hold very few opinions. I hold information based on observation and fact. If you wish to disagree, please use facts.