Re:Open Source is not dead: On the recent demise of RethinkDB
by pluma on Monday October 17, 2016 @12:21PM
(#12726482)
My feeling is that there are really just three forms of open source beyond the small-scale "single developer hobby project":
1. Standalone open source companies like RethinkDB. These typically rely on secondary sources of income (e.g. support contracts, SaaS, etc) or eventually adopt an open core model (deriving a commercial product from the free open source project).
2. Major software companies that open source part of their work to push their commercial services (e.g. Microsoft, Apple and Google), or as an elaborate recruitment tool (e.g. Facebook).
3. "Independent" open source projects where maintainers are directly sponsored by outside companies or indirectly via foundations (e.g. Linux, Apache, Mozilla).
I think #2 is only workable at a certain scale (because open source work basically becomes a major marketing expense) and #3 typically requires several significant backers to guarantee long-term survival.
So looking at startups the only workable solution for open source seems to be #1 but because open source lies at the core of what the business does it's vital to generate a source of income other than the open source work itself (unless the open source work is de facto unusable for most users without a commercial license, e.g. Sencha ExtJS and Touch).
RethinkDB seems to have failed at becoming profitable. My understanding is that they tried very hard to maintain a "pure" open source ideology and therefore fully relied on support and training contracts -- which is directly at odds with their primary goal of making RethinkDB easy to learn and use.
I think RethinkDB's biggest failure was that as a product it didn't really find a market where it could become profitable. Based on their marketing, the main audience seems to have been startups. They were likely hoping that startups would give them a reputation and that reputation would translate to traction which in turn would translate into paid support, training and commercial licenses.
But even though they could have made it a lot easier to give them money (the support and training is hidden behind the "Services" section on their website and I only heard about the availability of a commercial license from RethinkDB users) I think that alone wouldn't have helped them because there just isn't much money they could have extracted from that particular market.
RethinkDB is unattractive to most companies outside the startup sphere that aren't tracking the new hotness. Streaming is a great selling point but there are very few companies who would be willing to trade for that benefit when it means a ton of drawbacks in other aspects. Horizon was acting as a competitor/successor to Firebase but even Firebase doesn't have much of a market.
People repeatedly point out SaaS as a possible source of income but creating and maintaining a SaaS product is an entirely separate can of worms from "merely" working on an open source database. Plus it brings with it a large set of financial risks and overhead that can be a major drag on resources -- all the while hinging directly on the success of the open source main product.
And SaaS also suffers from the same problem as offering support and training: SaaS is most useful when deploying and maintaining the same services on your own hardware (or cloud) would be hard or troublesome, but RethinkDB tried to position itself as easy to use and easy to learn.
For databases the only plan that seems to work for #1 is the open core model. True, it's not ideologically "pure" open source but as long as the community edition is a decent product on its own, there's no harm in providing enterprise services on top of it at a premium.
1. Standalone open source companies like RethinkDB. These typically rely on secondary sources of income (e.g. support contracts, SaaS, etc) or eventually adopt an open core model (deriving a commercial product from the free open source project).
2. Major software companies that open source part of their work to push their commercial services (e.g. Microsoft, Apple and Google), or as an elaborate recruitment tool (e.g. Facebook).
3. "Independent" open source projects where maintainers are directly sponsored by outside companies or indirectly via foundations (e.g. Linux, Apache, Mozilla).
I think #2 is only workable at a certain scale (because open source work basically becomes a major marketing expense) and #3 typically requires several significant backers to guarantee long-term survival.
So looking at startups the only workable solution for open source seems to be #1 but because open source lies at the core of what the business does it's vital to generate a source of income other than the open source work itself (unless the open source work is de facto unusable for most users without a commercial license, e.g. Sencha ExtJS and Touch).
RethinkDB seems to have failed at becoming profitable. My understanding is that they tried very hard to maintain a "pure" open source ideology and therefore fully relied on support and training contracts -- which is directly at odds with their primary goal of making RethinkDB easy to learn and use.
I think RethinkDB's biggest failure was that as a product it didn't really find a market where it could become profitable. Based on their marketing, the main audience seems to have been startups. They were likely hoping that startups would give them a reputation and that reputation would translate to traction which in turn would translate into paid support, training and commercial licenses.
But even though they could have made it a lot easier to give them money (the support and training is hidden behind the "Services" section on their website and I only heard about the availability of a commercial license from RethinkDB users) I think that alone wouldn't have helped them because there just isn't much money they could have extracted from that particular market.
RethinkDB is unattractive to most companies outside the startup sphere that aren't tracking the new hotness. Streaming is a great selling point but there are very few companies who would be willing to trade for that benefit when it means a ton of drawbacks in other aspects. Horizon was acting as a competitor/successor to Firebase but even Firebase doesn't have much of a market.
People repeatedly point out SaaS as a possible source of income but creating and maintaining a SaaS product is an entirely separate can of worms from "merely" working on an open source database. Plus it brings with it a large set of financial risks and overhead that can be a major drag on resources -- all the while hinging directly on the success of the open source main product.
And SaaS also suffers from the same problem as offering support and training: SaaS is most useful when deploying and maintaining the same services on your own hardware (or cloud) would be hard or troublesome, but RethinkDB tried to position itself as easy to use and easy to learn.
For databases the only plan that seems to work for #1 is the open core model. True, it's not ideologically "pure" open source but as long as the community edition is a decent product on its own, there's no harm in providing enterprise services on top of it at a premium.