Ask YC: To open source or not, that is the question

I may get down-modded for this, but I suggest not releasing it given the scenario you described. I love open source, but most open source contributions are made for joy of creation, pride, fame, sometimes naivete (by assuming you'll get the benefits described) and finally, money. Many projects are created because they're ancillary to some business need--like SUN paying Charles Nutter to work on Jruby. These motives don't really apply to your case, and the benefits you list don't materialize for most projects. It costs you time to open source things properly and limits your options. Rightly or wrongly, some companies will be reluctant to buy you with your code released.

Statistically, most open source projects do not gain traction, significant users, free features, or free QA. They sit there on SourceForge, unloved. I wish I had a citation here. Try browsing SourceForge projects. Publicity-wise, consider how many you've heard of. It seems likely you'll get little in return for open sourcing. It's like a dance, and there's a decent chance that Baby will be put in a corner.

I would second this. We currently run an open-source project (distributed storage) and commercial venture based upon it, but the open-source was the driver done because we thought that it was an important problem to solve and that it would be worth more as open-source than as a proprietary product. Our commercial venture arose by choosing one specific application of the code and building a business to address it. Note that our open-source project is something that we (and others) have worked on independently of the commercial venture. The community is separate (though aware) and drives the open-source project in directions that we did not expect.

Hard choice. One other factor you didn't mention about is community. In releasing your code to a wider audience, you are trying to create a community around your code. Community requires support; a wiki/bug-tracker to post problems and solutions, a mailing-list to support conversations and questions, a robust source control server to host your code, and a understandable release process to push out new versions.

Also, you will need to write a lot of documentation about your code, how to install, how to extend, etc. You will probably want to write a FAQ first to fend off first-time newbie questions about installation, management, etc.

You will also have to make conscientious decisions about what versions your software will run on, and which requests you will support -- we run a stack on Tomcat 5.5 with Java 5, for example, but we have gotten X amount of questions about support for Java 6, JBoss, JRun, etc etc etc.

My two cents: it's rough to get other people involved. The good karma for open-sourcing can be more trouble if you are not organized about how to handle and attract new members to your project.

First off - don't let me put you off - you are managers of your own destiny but...

I am not sure that SMEs do a lot of formal project management - they might just possibly buy a copy of MS Project but that's as far as it goes in the main. I am sure there is a generic ERP market among SMEs but those folks are careful with their money - spending it in areas that are more central to the areas where they make money. One problem area is going to be the sheer diversity of the market place - large corporations are remarkably similar (hence the wide adoption of SAP perhaps) - but SMEs tend to run leaner and making more use of individual skills - thus shaping their operational procedures around people rather than corporate "this is how it's done". I would look for a vertical market to tackle first while striving to make each ERP component as generic as possible. Good luck anyway.

They do not right now, but if we provide enough value above what can be achieved with MS project, why would they not be willing to part with their money? (Ah, or so I wish :) )

Let me see if I've got this right: you want to open source your code so you can get a bunch of free bug fixes, QA testing, and positive press, and what you're offering in exchange is a crippled, deployment-size-limited version of your package for small teams and other open source projects?

Thanks, but no thanks.

It sounds to me like you simply don't have the resources you need to develop your system, and want to get a bunch of free labor from the OSS community without really opening your code.

If you're going to go open source, you're still going to need folks to do documentation, QA, and marketing. You're still going to have to plan and manage releases. Hell, you're going to have to fight an uphill battle convincing some managers they should pay you for anything when the base product is "free."

Plus, you're going to have a community just itching to fork your product and put you out of business at the first sign of you putting your bottom line ahead of their interests.

Do yourself a favor, and either scale back the scope of your project so it's something you can handle, or find the resources to staff up and finish the damn thing. Then decide if you want to open source the code base. Just don't expect something for nothing.

How's it different/better than something like OFBiz or one of the other open source systems? What's your edge over all the other project management things out there?

With open source, you're essentially foregoing any revenue from licensing in the hope that you'll make money consulting, unless you do something like have a GPL library where people will pay to be able to link to it in proprietary apps.

I'm not entirely convinced that the world needs yet another project management thing. I think it had either be really good and far better than what's out there now, in which case you can charge for it, or... perhaps another direction is best?

PS: YC's formatting isn't very advanced.

> I think it had either be really good and far better than

>what's out there now, in which case you can charge for it

That is what we plan to do. The idea is that over time, we will have a full service, not just a project management. A hrms system, a payroll system, a project management system, which all tie together. That is what makes big ERPs be able to charge such high licensing fee. This is what we plan to bring to SMEs

To be very blunt: I don't know, I'm just not convinced that two guys are going to go up against even existing open source systems... Not impossible, but I don't see it as easy unless you really have something up your sleeves.

Open source it for non commercial use. Don't expect much benefit from open sourcing at first. It takes patience to build a decent community around an open source project.

While most open source projects never gain any traction, plenty of companies have thrived with an open source model. (Not all end up like MySQL though)

If it's open source, you can't dictate what people do with it. That's part of the definition. You could use a license like the GPL that means they're not going to take it proprietary, but it's simply not open source if you can't use it for what you want.

GPL only covers source code. It doesn't cover other copyrighted works (like help manuals, logos and other text) or allow users to user your trademarks. So while the source code is open you only license use of all the other material for noncommercial use.