Open-source and the demise of proprietary software (2018)
blog.timescale.com
blog.timescale.com
Open source is enabling proprietary software to be more profitable, because it makes free all the commodity pieces that are needed to make proprietary software work.
Customers would otherwise have to pay for these things (in an ongoing manner) leaving less of their $$$ to spend on your payload.
Open source powers SaaS which is so closed and proprietary that the user doesn't even get to execute the binary code on their own hardware.
Of course, if your business is selling operating systems, development tools, databases and other infrastructure stuff, then, yup, open source has eaten your business.
It seems like the rule is that open source is eating all-except-niche software.
The market just naturally gravitates to the biggest player.
I agree with what you said. With the caveat that some lucky few companies do seem to do well in this space too. Recently learned JetBrains (makers of IntelliJ) had $270M revenue/year and $100M net income in 2018.
Even with sectors like developer tools, there is a lot of room for creating value-added solutions that are proprietary, and that customers will happily pay good money for.
In 1990s, you turned a blind eye to small-time piracy to get engagement. Now you can offer a much better deal.
Open source powers google and amazon yet the authors get little return from it.
Open source has killed indie development.
And facebook has already been used for genocide: https://www.nytimes.com/2018/10/15/technology/myanmar-facebo...
A different license could fix that but that goes against the philosophy; it would fix it though. People are trying that and a lot of other people are dead against it for good reasons, but how else can it be fixed?
Even if you don't make it viral towards revenue, then GPL works the best imho; at least the big corps must give back when they make changes. The problem is that many use non viral licenses and then it's just free for all. I personally do that as well at times, but it does what you say.
> Open source has killed indie development.
In what way? I don't understand the remark.
> In what way? I don't understand the remark.
Open source has reduced scarcity by such a high degree that writing quality code is valued next to nothing in monetary terms. Also, communities around closed source software have better ways to monetize, since there is usually no expectation from clients to get it for free as is from clients who are oss centric. Sure one can get the "opportunity" to work at FAANG just like their forefathers had the opportunity to work at Ford, Chrysler and other googles and amazons of the last generation, but that is not freedom. In other words, OSS reduced scarcity, shifted the balance towards a few centralised software powerhouses and has forced the vast majority of great engineers into 9 to 5 jobs.
EDIT: I am hugely in favor of open source as a means to share knowledge, but i am not in favor of OSS as free code for companies to monetise. Companies are entities that generate revenue, and for them, our effort to better the world is just a reduced investment cost, and nothing more.
This is a big assumption. How many open source developers do so for altruistic reasons? How many open source developers are funded by companies looking for good will.
There’s a simplistic view that developers only want money and that if they don’t get rich, that’s bad. That may be true for some, but those people would need to be stupid to desire wealth from software and still make lots of contributions to open source projects.
There are many benefits other than just payment.
I’ll also add that open source is really diverse and there are many companies profitable and happy chugging along or getting bought out (red hat, canonical, jetbrains, etc etc).
If people want to get wealthy fro software there are many companies doing so. Either through selling licenses (Microsoft) or just making secret sauce that doesn’t get sold (google, Amazon, Wall Street).
I would agree with you if there was no path to riches or there was some weird open source draft that forces people to code without pay.
But as it is now, developers can choose, and do, to donate their time and brainpower writing free/libre software.
The desktop OS market has basically two players: Windows and Mac OS. Desktop Linux is so small it's no relevant force at all.
> development tools
What about the intellij tools? They cost quite a large amount, and it's a monthly cost. MS Visual Studio is still around as well.
> databases
SAP is still a major force on the German market. The company behind it is amongst the 30 biggest public German companies.
We all use these infrastructure open source software products because they serve needs that practically anyone doing technology work will find themselves to have.
The larger universe of commercial software is HUGE and most of it is proprietary and will remain that way.
It is also a PR piece by TimescaleDB which has all interest in convincing people/investors that an open-source company is a good deal:
> (Disclaimer: Our company TimescaleDB is an open-source time-series database startup and shares investors with several of these companies, including Elastic, MongoDB, Hortonworks, Confluent, and Databricks.)
There were more paragraphs, but I only picked 2 to comment:
> If you are an individual, your project (or your contributions to someone else’s open-source project) will get you visibility and help your long-term career. If you want, you’ll travel the world, deliver talks on your work, and meet like-minded people along the way.
I doubt this is still the case. With GitHub/Universities promoting open source contributions all over the place, it offers much less leverage. Maybe a very popular project helps the career but otherwise I think it's just a work out of passion. Might even hinder your career.
>Some think that open-source foundations should fund open-source projects directly through stipends or fellowships. But these foundations are not money-making machines: e.g., the Apache Foundation barely made $500,000 in 2011. Perhaps these foundations can fund individual developers, but are extremely unlikely to be able to support open-source projects at scale.
True. But some foundations do fund specific areas in need, eg. FreeBSD Foundation. Apache is a bad example as they don't fund anything. What do they need $500,000 for is unclear to me since infrastructure for projects can't cost so much (and nowadays it's offloaded to GitHub, etc).
Btw for others who are reading - contributing to open-source is still very much a way to build your career, travel the world, etc. I personally know people who are living their best lives because of this. It doesn't happen for everyone, but if you are early in your career then I very much recommend it.
I would like to see some data point towards the success related to open source involvement. I'm sure outliers exist just as there are people doing really well with their solo AppStore app or with their niche website.
And I'm saying this as somebody involved in open source: better to see it as a hobby done out of passion. Any other benefits are random/luck and except if you get to feel some passion for a highly sought out niche (hello Machine Learning) you shouldn't expect anything out of it. Entering into it with the expectation of building your career, traveling the world, etc is going to lead to disappointment.
I suspect the biggest change to affect proprietary software is the shift to SaaS rather than open source.
I'm not saying this is actually happening, and experience with companies who build on FreeBSD show that upstreaming their changes helps keep the diffs small when pulling down updates around their proprietary code.
I think I still agree with most of the points in the article, but if I were to write this today I'd emphasize the role of Cloud and SaaS even more.
I think it's pretty clear that Cloud has become the dominant way to consume open-source for new projects.
It's also true that SaaS is already replacing traditional proprietary "on premise" software.
But I think there's a wrinkle here: open-source software can still build larger communities (and as a result, reduce reliance on a single vendor) than a SaaS company ever can.
For example, look at Elasticsearch. Even if you don't want to run Elasticsearch yourself, you can choose to run it on Elastic Cloud or on AWS. So now you have at least two different companies who are invested in making Elasticsearch better for you. Compare this to Sumologic or Splunk, where you are clearly beholden to that vendor.
In other words, SaaS is the new proprietary software. And I still believe that open-source is poised to "eat" it.
More important in my mind is that open source software is strictly higher quality from a technical perspective. Because it's not prostitution most of the time, it's passion and wider perspectives such as sharing knowledge. And a higher quality foundation means better software, because whatever commercial cruft someone adds on top it will still work better.
I recently reinstalled a laptop that was running Windows 10 with Kubuntu for a non-technical friend, which is not an excellent example but still applies since Canonical is semi-commercial. And it's a total win, because Windows is simply crap from the bottom up.
Android is another example, as is macOS.
People don't certainly exercised their right to modify the source code as often as their usage. It's certainly an important right, but less often than the right to use the software as the user see fit.
Arguably freedom zero is much more important in a world in which every business and companies are trying to nickel and dime you as an eternal income stream.
I tried to compile Android from source this year, and spent weeks reading about it, finding distributions and tools to help, downloading packages, verifying signatures, and eventually running an enormous build command. I was ultimately unsuccessful. No matter what I fixed, there was something else that didn't work. Everything had the classic open-source excuses: it's being fixed soon, that's the wrong branch, you need to upgrade your tools, the docs are out of date, ...
Android today is not anything like Andy Rubin's "the definition of open" from 2010. As far as I can tell, it's basically SAAS for smartphones. A corporation gets the benefits of open-source, but users don't.
Some types of software lend themselves to Open Source, others don't. I once saw a disparaging comment in a forum attempting to create an open-source accounting package, "who would want to work on that boring stuff?"
See http://progenygenealogy.blogspot.com/2018/01/the-open-source...
Proprietary software reinvents itself as corporate-driven Open Source with SaaS Lock-In
Look at Microsoft. Used to charge an arm and leg for software. Then ran up against competition from Apple and Google.
2000s were not good for Microsoft.
They’re getting their act together now. But for a while they were very much old tech.
In many other fields this collaboration is enabled through government. Companies pay taxes, taxes are used to fund infrastructure such as roads and railways.
If I remember right for example Apache Software Foundation had a requirement that core team for project should not come from just single company.
One way to categorize open source projects is to check how many companies really collaborate on the project. Is it a true joint effort or is it in practise dependent on one entity.
These three categories cover the spectrum of motives for doing software.
What's key is to respect these motives and protect the legal regime that supports their creative tension.
One of the many papercuts was actually more serious, to the point it was funny: I remember how in their stable versions, they had a humorous bug: if you did \copy instead of insert to a timescaledb, the timestamps looked right when you eyeballed them with a select, but acted wrong when you used them in a query.
Theefore it passed tests, but took down production whenever than happened. Ouch.
To their credit, they preserved this bug very well for several stable versions - for far longer than my patience with broken software.
That's quite bad for a database that intends to do timeseries - something so basic you won't ever see in proprietary software.
Then before there was pipelinedb which also had its fair share of bugs before going under. Some of them took a production database offline. But it was such a CPU hog that it had been anticipated.
Still after that timescaledb bug, I banned the use of third party postgres extensions.
This problem you mentioned is new to me - have not seen it before. Did you file a Github issue for this? Happy to chat more there (or feel free to ping me directly - ajay (at) timescale.com).
Just try to do a massive '\copy from' into a table that is already a timescaledb, using a tsv source with at least a few million records, make sure the dates are well spread out. Then use windowing functions to count the records per time unit. Compare date_trunc to time_bucket and select. Unless the issue was silently fixed, I'm sure you will be laughting too - but I recommend you wait until after xmas to have that good hearthy laugh, as it may cause you to want to fix the issue ASAP :)
Once of the field was defaulting to NOW(), maybe that's what caused the funny bug but I didn't investigate much after that. It was the straw that broke the camel back. Thousand papercuts. Not worth the time investment.
But to a programmer, I guess it may look like open source is replacing everything else. That's because we programmers have gradually have replaced everything we use with open source - compilers, editors, web servers, libraries for just about everything. Even stuff we use play with rather than get paid for is replaced - like photo and video editors, 3D printers, model plane piloting software.
It's easy for us, and cheaper in the long run, and given enough time the end result is better quality than the proprietary stuff because we all pitch in to knock off the paper cuts.
But it really is a case of "scratching itches", so if a programmer doesn't get itchy about something then it usually won't have an open source version. It's not hard to find examples. There are some usable PCB design tools out there - but 3D mechanic drafting tools weren't a thing until 3D printers tickled our interest in that area. Anything that requires specialised hardware is going to be out of reach - so you won't see open source MRI scanners, but I have no doubt you will see devices using 2 cheap web cams to do SLAM in the not too distant future.
This has had the accidental effect of making us more efficient. Any tool we can conceive of that allows us to do our job done better is free. The tools we already constantly get better - for free. In any other industry this might be a problem as supply overwhelms demand, but we seem to be in the odd position of the world demanding what appears to be a near infinite amount of software.
I am constantly amazed by how the demand for good programmers never seems to tail off. Even where I live (Australia) which has very high wages by international standards, and in the face of huge numbers of programmers coming out of low wage areas, _and_ in the face of jobs being some of the most transportable on earth, anyone competent programmer who wants a job gets one within weeks. I think we partially have open source to thank for that.
But step outside of what programmers want, and we are as mercenary as the rest of them. We are more than happy to develop nice new shinies we use and give them away for free, but we know who butters our bread. We don't write code for them for free. And since we are all getting paid, that must mean most of the code we write isn't available for free.