Slashdot Mirror
Ask Slashdot: Can SaaS Be Both Open Source and Economically Viable?
An anonymous reader writes: The CTO behind Lucidchart, an online diagramming app, recently cited the open source rbush project as an invaluable tool for helping implement an "in-memory spatial index" that "increased spatial search performance by a factor of over 1,000 for large documents." My question is this: what risks does a SaaS company like Lucidchart face in making most of their own code public, like Google's recent move with Chrome for Android, and what benefits might be gained by doing so? Wouldn't sharing the code just generate more users and interest? Even if competitors did copy it, they'd always be a step behind the latest developments.
Have a question for Slashdot's readers? Take a look at other recent questions first to see if someone else has had a similar question. And if not, ask away! The more details and context you include, the more likely your question will be selected.
The GNU Affero General Public License version 3 was designed to preserve user freedom and flexibility even when software runs on a leased server. It ensures that users can obtain and improve the software that they are using even when they are currently running it on someone else's computer. That way, if a particular service goes out of business, its customers can spin up an instance on their own servers with little interruption.
Google's Chrome would be a good example. Google's business is not selling browsers. Their business is selling advertising. Many of the services they offer to attract eyeballs (and data) for their business require a good browser. So they don't lose any revenue by giving their browser away and letting other people build browsers based on the code, in fact the more modern browsers out there that're all compatible the better for Google. In that situation it makes sense to open-source their Chrome code. For any business, if the code's utility code that's necessary for the business but not a significant part of the parts that separate your offering from everyone else's it'd make sense to open-source it. You don't lose anything, you gain brownie points, and you may be able to use the bug fixes and enhancements others make without having to spend your own resources on them.
You don't, however, see Google open-sourcing the details of their analytics algorithms, or the exact code that drives PageRank, or the other things that set them apart from other search engines. Those things they need to keep secret because if they got out Google would lose a competitive advantage. Open-sourcing code like that would cost a business revenue, so it shouldn't be open-sourced.
Its simple, you use somthing like the Affero license so no one can make changes that you can't access. Since anyone can try, whatever good ideas other people have you can re-incorporate, so you can potentially have a far bigger unpaid developer base. If your product is known to attract hacker types as customers, they can act as force multipliers, easily. As compared to a closed program, you'll have more eyes on the code. Its also your code, and you know it better than anyone else.
Then you simply focus on having the best quality of service. You can copy software, you can't copy quality of service.
Combine these two, its not as easy as you think to compete against someone else with their own software.
You also have your brand name and reputation. which is built on that quality of service. Despite the fact that CentOS is given away for free, people pay good money for RHEL subscriptions, and RH is an economicly viable company.
The support is where the money is. The actual product is a loss leader.
Once the novelty of free and open source has worn off the only advantages I can see would be user goodwill and perhaps some modest community contributions. Disadvantages include enabling competitors, devaluing software production, and risk of community backlash should one change models or they disapprove of the license choice. My guess is such companies hope to overcome any competitors by executing well and retaining a first-mover position in the market. They are probably unaware of the macro effects of FOSS on labor, and oblivious to potential community issues.
Software has zero intrinsic value. It doesn't generate a single cent (unless you've written a BitCoin miner, I guess).
Video games generate 20+ billion a year and downloadable ones are comprised entirely of software.
As someone who owns a web hosting business, and recently migrated all of their servers from MySQL to MariaDB. It was the easiest transition we've ever performed. On our cPanel boxes it was done in just a couple of clicks.
MariaDB really is a drop in replacement for MySQL. They have done an awesome job ensuring its a dead simple upgrade.
We are currently looking into upgrading our DNS network. We are toying with MariaDB Galera, and PowerDNS. Our initial testing has been very positive.