Ask HN: Why don't more web startups open-source their applications?

[te_chris]

This is something I've been thinking about a bit recently, but decided to look into it a bit more. One of the only ones I could find (apart from the obvious such as diaspora, though they don't seem to have a business model as far as I've seen/googling tells me) is SugarCRM, who seem to be doing quite well with a combo of GPL'd code and a commercial SaaS product built off that code.

I guess what I want to know is why is this such a bad idea that a lot of us, who are OSS advocates, don't follow this sort of business model (or another OSS biz model)? Also I was wondering if there are any famous casualties of companies trying this and it failing on them?

[jasonkester]

How about this as a fundamentally equivalent question: Why don't more web startups run poetry competitions?

Much like Open Sourcing your code, running a poetry competition is unlikely to add enough business value to offset the resources it will take to accomplish.

Sure, there are guys on your team who really like poetry and would go out of their way to help organize the contest. And the poetry fans that happen to be your users would think more of you for doing it.

But at the end of the day, it's just not a good business decision, so you don't do it.

[csense]

As a business, you want to look for "sustainable competitive advantage." A way to differentiate themselves from competition and make a profit.

Having closed-source code is a sustainable competitive advantage for many companies selling heavily software-dependent products like games, apps, websites, etc. "Advantage" because your competitors don't have it, "sustainable" because, with adequate investment in new features, you can keep your platform ahead of copycats.

There are a variety of strategies to make money with open source, but as a business, your first priority should be figuring out how you'll make enough money to adequately pay your owners, employees and vendors. The answer to that question is sometimes "open source," but not always.

Strategic reasons can sometimes justify open source products. For example, consider infrastructure components that are an incidental byproduct of your operations, not a core piece of your product. Twitter Bootstrap immediately comes to mind. Twitter releasing Bootstrap as open source doesn't help its competitors in the same way that open sourcing the code to its main website would, and allows the community to provide some amount of testing and improvement which may help them in the future.

But there is no one-size-fits-all answer; whether to open-source your code, and what to open-source, depends on your specific situation.

[benologist]

For a lot of sites the code is pretty much the sum of the business and the rest is just customer acquisition which plenty of people will have more experience and money and smarts to figure out better and faster than you. Maybe not directly at your expense but it's easy to imagine bidding against competitors you created or gave a head start to in AdWords.

There are exceptions, anything where the code is secondary to the community or something like reddit.

[_delirium]

One problem is that with SaaS apps, the GPL isn't "viral" in the same sense that it is with desktop apps. It doesn't provide protection against a competitor taking your code, adding proprietary extensions to it, and then running a competing product that's an improved version of your product without sharing back the improvements, because they would only be "internal" improvements that don't have to be shared. You can stop that with the Affero GPL, which considers hosting the app to be equivalent to "distribution", but enough people seem to hate the AGPL that there's not a lot of incentive to bother releasing at all.

[michaelpinto]

My guess is that sadly investors value unique intellectual property that can be protected, and they might not be too crazy about open source as a result. This is because if the company goes south the patents may be the only thing the investor can salvage.

[mechanical_fish → michaelpinto]

The problem with this hypothesis is that you never hear anyone in the startup community talk about their patents and how much they are worth. Not on HN, not in the media, not on Techcrunch. You never see an "Ask HN" about how to file a patent, or license a patent, or litigate a patent. And I don't recall ever hearing a VC talk up the awesomeness of patents, or encourage a startup to waste even 0.1% of its time filing for patents instead of doing something more valuable.

[michaelpinto → mechanical_fish]

If you're getting angel money you're right (because you're doing proof of concept) -- but if you're at the VC stage that's incorrect. And I've gotten that from VCs...