8 ms·

The author makes a good case against open source businesses, but he indulges himself in a false dichotomy between a traditional proprietary model and open sourc

The author makes a good case against open source businesses, but he indulges himself in a false dichotomy between a traditional proprietary model and open source. Increasingly often, there is no market for your proprietary software. It is then not a choice between open source and proprietary models, but rather between doing open source or doing nothing. A lot of markets are just gone, and they're not coming back. Odds are, you're never going to sell another proprietary UNIX clone, or a proprietary database, or a proprietary application server. There are no punters. Those niches have been extinguished.

mooreds 6y ago
This is a really good point. However, there are still successes that I'd never have expected. Jetbrains is a good example. Who would have thought that an IDE could be sold when there are adequate ones for free? There are also many many smaller niches where software can be sold because there isn't adequate developer interest to create OSS. One example is this: https://querix.com/products/lycia https://querix.com/products/lycia (not a customer, just ran across their product).
- hapless 6y ago
  It's worth noting that Querix was founded back in 1994 -- when Informix and proprietary databases were more popular than today.
- InvOfSmallC 6y ago
  I personally have been being using Intellij for everything for the last five years. I tried to migrate to VSCode for golang. I mean, don't get me wrong, the situation is not the worst possible, but the only reason you would want to use that is because you never tried refactoring with Goland. Are you kidding me? I had to go look for what is the current needed extensions for go that are still maintained, configure them etc...vs just download and use Goland. If you can't or don't want to spend money OK, but otherwise is a non sense to use anything else.
  - linkdd 6y ago
    The best IDE is in my head. I don't need to type anything and it can even run buggy code without any error.
    - erik_seaberg 6y ago
      I want the computer to do as much work as it possibly can. My brain only runs at 0.0000001 GHz because it’s made of meat. But I want to use a concise, powerful language that requires more work from the compiler, rather than an impoverished language that assumes an IDE will generate a lot of crud that isn’t worth reviewing.
    - ignoramous 6y ago
      > I don't need to type anything and it can even run buggy code without any error. Sounds a lot like the free IDE I use for HTML and JavaScript... Firefox.
  - zaphirplane 6y ago
    So pay for it in dollars or in time ;) but goland isn’t free last time I checked
- tinco 6y ago
  I did, there's never been really good community driven open source IDE's. For every community driven IDE I can name you an enterprise driven IDE that's popular and arguably better. Code editors are a product where polish has high value, and users are willing to pay for that polish. I don't particularly care if my editor is open source as long as it's got strong community and great polish.
  - thu2111 6y ago
    IntelliJ is open source. It was once proprietary but these days is "open core". The Ultimate edition is the open source version of IJ with more plugins. However the rebranded non-Java versions like Goland aren't open source, unless you consider the IJ base. They are the core language plugin(s) + integration and rebranding code.
  - CRConrad 6y ago
    From what little I've tried it, the open-source Object Pascal IDE Lazarus doesn't really feel all that much worse than Bo– ah, In– eh, Co– uh, Id– erm, Embarcadero Delphi.
- keyle 6y ago
  Yep IntelliJ is great. I remember back in the days, changing from Microsoft visual studio with resharper to eclipse. Hurrrr what a pain. Then you'd try something else, have to reconfigure everything, never happy. Then I discovered the makers of resharper made an IDE which supports tons of languages and never looked back. As long as they don't fuck it up like Adobe fucked photoshop (and every product they own), they'll have my money.
hodgesrm 6y ago
> The author makes a good case against open source businesses, but he indulges himself in a false dichotomy between a traditional proprietary model and open source. More generally people seem to get tunnel vision about open source models. The big distinction in business software is not open vs. proprietary but SaaS vs. non-SaaS, with SaaS increasingly dominant. OSS companies therefore need to have a SaaS strategy. You build your own on community OSS projects, supply other SaaS businesses, or use OSS as an on-ramp to your SaaS product.
- dgudkov 6y ago
  What viable SaaS strategy can you have if Amazon can come at any moment and do the same?
  - nradov 6y ago
    Amazon only offers horizontal platform services. There's still huge opportunity in the verticals.
    - dgudkov 6y ago
      Can't think of any vertical FOSS. Can you?
  - hodgesrm 6y ago
    Well you could start with a SaaS-first business model. People are obsessing about Amazon, but the fact is if you publish open source software and don't build an outstanding SaaS for it, somebody else will do it for you. It can't be an afterthought. Second, add IP on top that makes the service easy to develop applications for, secure, easy to integrate with other SaaS services, etc. Amazon gives you a service but for the most part they don't build ecosystems. Third, make it work on multiple platforms, not just AWS. Fourth, integrate with AWS Marketplace and get their reps working for you.
tinco 6y ago
Exactly, and open source companies make great money too. ElasticSearch being a prime example, yes AWS eats a large share of their pie, but who cares, the pie is absolutely huge. Another example is GitHub and Gitlab. Both got big by leveraging the open source community for marketing. Imagine all the resources GitHub spends to enable open source, that's a cost center that will never go away. If they hadn't done it though, they simply would never have become big. And then Gitlab one ups them, sacrificing an even larger share of the pie, allowing free private repos and free on site deployments. And still they're carving up a big chunk of pie for themselves. In my opinion it's the only way the future is going to make any sense. It's not a race to the bottom, it's a race to more balance between the rewards of beating the network effect. If you want to build a new company in a space that another company is already offering significant value, then you are going to have to offer even more value. And if there's no more value go squeeze out of your functionality, and you're already not charging your users, that means you're going to have to give part of your company to the users. It's a mathematical inevitability.
- carlhjerpe 6y ago
  You should know, Github offers free private repositories. I believe because Microsoft doesn't need Github revenue, but to use it as a funnel towards Azure. https://github.blog/2019-01-07-new-year-new-github/ https://github.blog/2019-01-07-new-year-new-github/
  - tinco 6y ago
    Oh I know, but they started doing that years after Gitlab did, and only when Gitlab became a significant competitor.
- thu2111 6y ago
  Is GitLab profitable? They seem to have $120M in 2019 revenue but I'm not sure how big they are. GitHub got big but then sold to MS. I'm not sure it was ever big in terms of profits. If the argument for investing in open source companies is that they can get a big exit by selling to a proprietary software company then I'm not sure the argument is strong.
crazypython 6y ago
> A lot of markets are just gone, and they're not coming back. Odds are, you're never going to sell another proprietary UNIX clone, or a proprietary database, or a proprietary application server. I would argue this encourages innovation in freedom-respecting/open-source software business models.
fphhotchips 6y ago
>proprietary database Oracle, Microsoft and SAP all do very well out of their proprietary databases for OLTP. In analytics, Snowflake was the largest tech IPO ever last year. For whatever reason, the database market seems to be entirely ok with proprietary software. Disclaimer: I sell Snowflake for a living.
passivate 6y ago
Well, how many companies are there that are selling some sort of proprietary software and still making money? I'd wager there are hundreds of thousands of them around the world. Also, software can now be sold as a service, so the market is just transforming, no?
Barrin92 6y ago
>A lot of markets are just gone for tooling or infrastructure or languages this seems true, but most plain SAAS is proprietary, almost all end-user applications are and so on. I'm not entirely sure what percentage of the market that is but it's probably the largest one?
jandrewrogers 6y ago
The market for new proprietary databases is both large and lucrative -- it is printing money right now. What has changed is the structure of the market and the product economics. The classic marketing-driven retail database product model, the kind that puts billboards on the 101, stopped working as a financial model a decade ago -- development costs are too high for the marginal return. That has been replaced by models that produce much higher margins with much lower development cost. Unlike database companies of yore such as Oracle, they use little overt marketing because they don't sell through the public. It is a completely different value proposition for the customer too. And this market is growing. Proprietary databases have found their niche with qualitatively higher performance and throughput -- often 10-100x -- than the best open source has to offer on the same hardware footprint. As data continues to grow exponentially and cloud costs increase, operational efficiency has a much bigger impact on the bottom line than licensing cost. This gap has only been increasing over time. For most open source databases this is a fundamental architectural issue, and you can't fix architecture, so they are defenseless against highly optimized competitors. I was on a project a year or two ago where a particular ElasticSearch workload was replaced by a proprietary solution that demonstrably offered 100x higher throughput. It was apparently worth tens of millions of dollars to the customer for that improvement. The reason big retail databases went away is that 90% of the development cost is in the "big retail" part and not the "database" part. At the same time, development costs for a state-of-the-art database engine have become extremely high due to wage costs (worse than FAANG), so the industry needed to find a business model that would allow those costs to be contained. It is still around but much more invisible. It would be nearly impossible to properly finance a database company as a startup, so a different model is used.
- totoglazer 6y ago
  This is fascinating! Could you share some names of these databases? I’d love to learn more.
  - sitkack 6y ago
    BigQuery and Snowflake are my top two.

mooreds 6y ago

This is a really good point. However, there are still successes that I'd never have expected. Jetbrains is a good example. Who would have thought that an IDE could be sold when there are adequate ones for free? There are also many many smaller niches where software can be sold because there isn't adequate developer interest to create OSS. One example is this: https://querix.com/products/lycia https://querix.com/products/lycia (not a customer, just ran across their product).

hapless 6y ago
It's worth noting that Querix was founded back in 1994 -- when Informix and proprietary databases were more popular than today.
InvOfSmallC 6y ago
I personally have been being using Intellij for everything for the last five years. I tried to migrate to VSCode for golang. I mean, don't get me wrong, the situation is not the worst possible, but the only reason you would want to use that is because you never tried refactoring with Goland. Are you kidding me? I had to go look for what is the current needed extensions for go that are still maintained, configure them etc...vs just download and use Goland. If you can't or don't want to spend money OK, but otherwise is a non sense to use anything else.
- linkdd 6y ago
  The best IDE is in my head. I don't need to type anything and it can even run buggy code without any error.
  - erik_seaberg 6y ago
    I want the computer to do as much work as it possibly can. My brain only runs at 0.0000001 GHz because it’s made of meat. But I want to use a concise, powerful language that requires more work from the compiler, rather than an impoverished language that assumes an IDE will generate a lot of crud that isn’t worth reviewing.
  - ignoramous 6y ago
    > I don't need to type anything and it can even run buggy code without any error. Sounds a lot like the free IDE I use for HTML and JavaScript... Firefox.
- zaphirplane 6y ago
  So pay for it in dollars or in time ;) but goland isn’t free last time I checked
tinco 6y ago
I did, there's never been really good community driven open source IDE's. For every community driven IDE I can name you an enterprise driven IDE that's popular and arguably better. Code editors are a product where polish has high value, and users are willing to pay for that polish. I don't particularly care if my editor is open source as long as it's got strong community and great polish.
- thu2111 6y ago
  IntelliJ is open source. It was once proprietary but these days is "open core". The Ultimate edition is the open source version of IJ with more plugins. However the rebranded non-Java versions like Goland aren't open source, unless you consider the IJ base. They are the core language plugin(s) + integration and rebranding code.
- CRConrad 6y ago
  From what little I've tried it, the open-source Object Pascal IDE Lazarus doesn't really feel all that much worse than Bo– ah, In– eh, Co– uh, Id– erm, Embarcadero Delphi.
keyle 6y ago
Yep IntelliJ is great. I remember back in the days, changing from Microsoft visual studio with resharper to eclipse. Hurrrr what a pain. Then you'd try something else, have to reconfigure everything, never happy. Then I discovered the makers of resharper made an IDE which supports tons of languages and never looked back. As long as they don't fuck it up like Adobe fucked photoshop (and every product they own), they'll have my money.

hodgesrm 6y ago

> The author makes a good case against open source businesses, but he indulges himself in a false dichotomy between a traditional proprietary model and open source. More generally people seem to get tunnel vision about open source models. The big distinction in business software is not open vs. proprietary but SaaS vs. non-SaaS, with SaaS increasingly dominant. OSS companies therefore need to have a SaaS strategy. You build your own on community OSS projects, supply other SaaS businesses, or use OSS as an on-ramp to your SaaS product.

dgudkov 6y ago
What viable SaaS strategy can you have if Amazon can come at any moment and do the same?
- nradov 6y ago
  Amazon only offers horizontal platform services. There's still huge opportunity in the verticals.
  - dgudkov 6y ago
    Can't think of any vertical FOSS. Can you?
- hodgesrm 6y ago
  Well you could start with a SaaS-first business model. People are obsessing about Amazon, but the fact is if you publish open source software and don't build an outstanding SaaS for it, somebody else will do it for you. It can't be an afterthought. Second, add IP on top that makes the service easy to develop applications for, secure, easy to integrate with other SaaS services, etc. Amazon gives you a service but for the most part they don't build ecosystems. Third, make it work on multiple platforms, not just AWS. Fourth, integrate with AWS Marketplace and get their reps working for you.

tinco 6y ago

Exactly, and open source companies make great money too. ElasticSearch being a prime example, yes AWS eats a large share of their pie, but who cares, the pie is absolutely huge. Another example is GitHub and Gitlab. Both got big by leveraging the open source community for marketing. Imagine all the resources GitHub spends to enable open source, that's a cost center that will never go away. If they hadn't done it though, they simply would never have become big. And then Gitlab one ups them, sacrificing an even larger share of the pie, allowing free private repos and free on site deployments. And still they're carving up a big chunk of pie for themselves. In my opinion it's the only way the future is going to make any sense. It's not a race to the bottom, it's a race to more balance between the rewards of beating the network effect. If you want to build a new company in a space that another company is already offering significant value, then you are going to have to offer even more value. And if there's no more value go squeeze out of your functionality, and you're already not charging your users, that means you're going to have to give part of your company to the users. It's a mathematical inevitability.

carlhjerpe 6y ago
You should know, Github offers free private repositories. I believe because Microsoft doesn't need Github revenue, but to use it as a funnel towards Azure. https://github.blog/2019-01-07-new-year-new-github/ https://github.blog/2019-01-07-new-year-new-github/
- tinco 6y ago
  Oh I know, but they started doing that years after Gitlab did, and only when Gitlab became a significant competitor.
thu2111 6y ago
Is GitLab profitable? They seem to have $120M in 2019 revenue but I'm not sure how big they are. GitHub got big but then sold to MS. I'm not sure it was ever big in terms of profits. If the argument for investing in open source companies is that they can get a big exit by selling to a proprietary software company then I'm not sure the argument is strong.

crazypython 6y ago

> A lot of markets are just gone, and they're not coming back. Odds are, you're never going to sell another proprietary UNIX clone, or a proprietary database, or a proprietary application server. I would argue this encourages innovation in freedom-respecting/open-source software business models.

fphhotchips 6y ago

>proprietary database Oracle, Microsoft and SAP all do very well out of their proprietary databases for OLTP. In analytics, Snowflake was the largest tech IPO ever last year. For whatever reason, the database market seems to be entirely ok with proprietary software. Disclaimer: I sell Snowflake for a living.

passivate 6y ago

Well, how many companies are there that are selling some sort of proprietary software and still making money? I'd wager there are hundreds of thousands of them around the world. Also, software can now be sold as a service, so the market is just transforming, no?

Barrin92 6y ago

>A lot of markets are just gone for tooling or infrastructure or languages this seems true, but most plain SAAS is proprietary, almost all end-user applications are and so on. I'm not entirely sure what percentage of the market that is but it's probably the largest one?

jandrewrogers 6y ago

The market for new proprietary databases is both large and lucrative -- it is printing money right now. What has changed is the structure of the market and the product economics. The classic marketing-driven retail database product model, the kind that puts billboards on the 101, stopped working as a financial model a decade ago -- development costs are too high for the marginal return. That has been replaced by models that produce much higher margins with much lower development cost. Unlike database companies of yore such as Oracle, they use little overt marketing because they don't sell through the public. It is a completely different value proposition for the customer too. And this market is growing. Proprietary databases have found their niche with qualitatively higher performance and throughput -- often 10-100x -- than the best open source has to offer on the same hardware footprint. As data continues to grow exponentially and cloud costs increase, operational efficiency has a much bigger impact on the bottom line than licensing cost. This gap has only been increasing over time. For most open source databases this is a fundamental architectural issue, and you can't fix architecture, so they are defenseless against highly optimized competitors. I was on a project a year or two ago where a particular ElasticSearch workload was replaced by a proprietary solution that demonstrably offered 100x higher throughput. It was apparently worth tens of millions of dollars to the customer for that improvement. The reason big retail databases went away is that 90% of the development cost is in the "big retail" part and not the "database" part. At the same time, development costs for a state-of-the-art database engine have become extremely high due to wage costs (worse than FAANG), so the industry needed to find a business model that would allow those costs to be contained. It is still around but much more invisible. It would be nearly impossible to properly finance a database company as a startup, so a different model is used.

totoglazer 6y ago
This is fascinating! Could you share some names of these databases? I’d love to learn more.
- sitkack 6y ago
  BigQuery and Snowflake are my top two.