Open source is the way out of a tech downturn(plural.sh)
plural.sh
Open source is the way out of a tech downturn
https://www.plural.sh/blog/open-source-is-the-way-out-of-a-tech-downturn/
40 comments
It depends how you're valuing the company. If it's speculative based on multiples of revenue, then you're right.
If you go back to a EBITA multiple, then moving away from expensive cloud providers might genuinely pump your value.
The real question is whether or not you have the tech chops or the willingness to invest in the type of people you need to do this well.
If you go back to a EBITA multiple, then moving away from expensive cloud providers might genuinely pump your value.
The real question is whether or not you have the tech chops or the willingness to invest in the type of people you need to do this well.
The best way to save tech profits is to spend less money on software, somehow?
I thought he was saying that everyone should use Typescript. . .
Thank you for saving my time thought!
I don't think the writer of this article has thought this all the way through. Sure, replacing one of the SAASes from your tech stack with some self-hosted OSS will reduce your spending (as long as you spend less on the operations than you did on the SAAS of course). The main problem for tech companies comes when all their customers have the same idea and cancel their account in favor of OSS alternatives. That would lead to a much deeper downturn.
I also wonder how the "use FOSS, it's free!" mindset will coexist with the "pay OSS devs more" movement. That will be very interesting to watch for sure.
I also wonder how the "use FOSS, it's free!" mindset will coexist with the "pay OSS devs more" movement. That will be very interesting to watch for sure.
Hey I'm Michael, the author of the article, and I did consider this but didn't go into detail for fear of bloating the read. I think there's two points here.
1. Not all software has viable OSS alternatives. The most obvious is consumer software, where some self-hosted solution is simply nonviable, but even something like Snowflake provides a ton of value by offering turnkey petabyte-scalable datawarehouses which are operationally complex enough I struggle to find an alternative to it (that said many smaller snowflake users probably can do just fine with a cheaper alternative like Clickhouse).
Where I think the switching will mostly happen is where SaaS lock-in has been exploited far too much to create exploitive pricing power (looker is the best example of this to me, it's really just an admittedly very good visualization tool but extremely overmonetized), if there are low-friction self-hosted OSS alternatives, people should swap out of those low-value, high-cost products and recycle the cash to prevent layoffs or better grow their own products.
2. At Plural we are working on improving OSS monetization mainly for self-hosted OSS which is actually very valuable in theory (fortune 500s often prefer it for security reasons), but usually poorly monetized. This is still early-stage obviously but will ultimately be us allowing oss developers to monetize for either support or open core features, and us handling the licensing and billing in a centralized, low-friction way.
This could theoretically claw back some of the cost savings, but if the overcharging is more due to excess pricing power from the realities of "cloud" infrastructure and the lock-in it creates, there should be at least a large delta remaining for users by eliminating that.
1. Not all software has viable OSS alternatives. The most obvious is consumer software, where some self-hosted solution is simply nonviable, but even something like Snowflake provides a ton of value by offering turnkey petabyte-scalable datawarehouses which are operationally complex enough I struggle to find an alternative to it (that said many smaller snowflake users probably can do just fine with a cheaper alternative like Clickhouse).
Where I think the switching will mostly happen is where SaaS lock-in has been exploited far too much to create exploitive pricing power (looker is the best example of this to me, it's really just an admittedly very good visualization tool but extremely overmonetized), if there are low-friction self-hosted OSS alternatives, people should swap out of those low-value, high-cost products and recycle the cash to prevent layoffs or better grow their own products.
2. At Plural we are working on improving OSS monetization mainly for self-hosted OSS which is actually very valuable in theory (fortune 500s often prefer it for security reasons), but usually poorly monetized. This is still early-stage obviously but will ultimately be us allowing oss developers to monetize for either support or open core features, and us handling the licensing and billing in a centralized, low-friction way.
This could theoretically claw back some of the cost savings, but if the overcharging is more due to excess pricing power from the realities of "cloud" infrastructure and the lock-in it creates, there should be at least a large delta remaining for users by eliminating that.
> The main problem for tech companies comes when all their customers have the same idea and cancel their account in favor of OSS alternatives.
That implies that all tech companies have OSS alternatives.
That implies that all tech companies have OSS alternatives.
In a utopian world, companies would “take what they need, pay what they can”.
In the real world, the tech giants will likely dominate the open source market, either by donations that allow them to prioritise, or by killing other versions.
But is it necessarily a bad thing? It’s not perfect but perhaps it’s the closest we’ll get.
In the real world, the tech giants will likely dominate the open source market, either by donations that allow them to prioritise, or by killing other versions.
But is it necessarily a bad thing? It’s not perfect but perhaps it’s the closest we’ll get.
This article didn't price in opportunity cost.
It isn't just a $150k/annum expense on an individual SRE salary. One should consider the HR time to source candidates and time to interview. It would include the time your tech team is brought into system design and security meetings, time your tech leads spend in 1:1s. It may involve time for your CTO to document the part of your systems for enterprise customers who require full architecture details for their requisition process. It may involve time for your product team, business analysts or executives to be retrained on software that is unfamiliar and non-standard to them.
Most of these activities slow down critical parts of your business which could be focused on generating revenue. A fair accounting should attempt to estimate these costs. Consider: if you didn't disrupt the business and instead spent the same effort on revenue generating product features then how much could you earn?
It isn't just a $150k/annum expense on an individual SRE salary. One should consider the HR time to source candidates and time to interview. It would include the time your tech team is brought into system design and security meetings, time your tech leads spend in 1:1s. It may involve time for your CTO to document the part of your systems for enterprise customers who require full architecture details for their requisition process. It may involve time for your product team, business analysts or executives to be retrained on software that is unfamiliar and non-standard to them.
Most of these activities slow down critical parts of your business which could be focused on generating revenue. A fair accounting should attempt to estimate these costs. Consider: if you didn't disrupt the business and instead spent the same effort on revenue generating product features then how much could you earn?
how about we go the other way and help out open source development by open sourcing codebases of failed companies in the downturn and releasing the patents/licences? this would be a step one for next generation of startups to build upon. The VCs funding current startups (including YC) could implement this tomorrow as a policy and create a huge body of work ready to be reused.
The captation of knowledge is made possible in large parts due to some FOSS developers using non-copyleft licenses. If the receiver has no obligation to share, and makes money off of a product they build on top, expect them to not share anything.
At this point non-copyleft licenses are a problem, and shouldn't be used anymore. Don't wait for a foreign group hellbent on maximization of profit and on which you have absolutely no control to do something; use (A)GPL
At this point non-copyleft licenses are a problem, and shouldn't be used anymore. Don't wait for a foreign group hellbent on maximization of profit and on which you have absolutely no control to do something; use (A)GPL
This is a one way process - companies take but never give.
If there is a "when all you have is a hammer, everything looks like a nail" example in this industry, this is it.
Open source is not a panacea, in fact it can be the opposite. I don't know the opposite of a panacea. But a maintainer can decide to fuck right off. And unless your team has the skill required, or another maintainer steps up, you are screwed. Businesses like predictable, they like contracts, they like knowing in 4 years somebody will support them. F/OSS advocates need to recognize that.
Open source is not a panacea, in fact it can be the opposite. I don't know the opposite of a panacea. But a maintainer can decide to fuck right off. And unless your team has the skill required, or another maintainer steps up, you are screwed. Businesses like predictable, they like contracts, they like knowing in 4 years somebody will support them. F/OSS advocates need to recognize that.
2022 is/was a terrible year for a lot of people in specific areas or regions of tech, but was it a terrible year across the board? The only fundamental shift I'm seeing is back to some semblance of rational thought and expectations, but I've been through a few of these so might have different goalposts. For me 2022 was a "2000-light" year.
> some semblance of rational thought and expectations
The other day I overheard someone asking when investors would stop caring about profit so their company could get back to focusing on growth.
It was really a testament to how long we've been in the bubble. There are people out there who earnestly don't understand that the raison d'etre of a publicly traded company is to generate profit and value for shareholders. Their entire careers, which may be nearly a decade long, have been in a world where revenue growth was all that mattered. I've seen people sincerely perplexed with a company's stock drops when they beat revenue at earnings while ignoring that their cost growth outpaced revenue growth.
From what I've seen we're still only at the borderlands of rational thought.
The other day I overheard someone asking when investors would stop caring about profit so their company could get back to focusing on growth.
It was really a testament to how long we've been in the bubble. There are people out there who earnestly don't understand that the raison d'etre of a publicly traded company is to generate profit and value for shareholders. Their entire careers, which may be nearly a decade long, have been in a world where revenue growth was all that mattered. I've seen people sincerely perplexed with a company's stock drops when they beat revenue at earnings while ignoring that their cost growth outpaced revenue growth.
From what I've seen we're still only at the borderlands of rational thought.
All ideas that are true generally, when treated as axioms (and then devoid of the context from which they were derived) eventually lead to irrational thought and behavior. "Modern" economic theory and investor capitalism seem to have perverted, "a rising tide lifts all boats," into becoming accustomed to surfing on tidal waves.
Large scale FOSS success is dependant on fiscal surplus from non-FOSS ventures and the ability of the US empire to globally enforce related legal aspects (licensing). The article correctly states that the former is diminishing. Like it or not, so is the latter.
large scale FOSS came first and I think it's a rather recent phenomenon that people are getting serious about fairly compensating the work of OSS maintainers that multi-[mb]illion dollar companies are based on
does anybody remember the original online Flash games? Newgrounds and Kongregate and all that? It was mostly people with hobbies. Passion projects. Rarely did the creators make any money off them even when their games got played by millions. When in-app purchases became a thing I saw a lot of these original flash games turn into mobile games with added in-app purchases littered throughout. Nowadays if you visit Miniclip or Kongregate and play a more recent game, you'll find the same. Either tons of ads or tons of in-app purchases
I think there's a strong cultural effect there once money is mixed in. Once upon a time online drug forums were filled with people sharing their "TEKs" about growing weed plants and freely trading seeds of certain varieties. Now in the aftermath of legalization you'll hardly ever find this type of stuff for free. I think the shrooms community is in a similar place as well.
OSS is now fueling multibillion dollar companies and maintainers are starting to demand compensation. Many will still contribute for "free" because it looks good on a resume, but I really think we're gonna see a strong decline simply due to the cultural impact of the transformation of Social Capital to financial capital
If you wanna get anthropological about this, I HIGHLY recommend this book[^0] about the social capital that fueled the social capital system that underpinned the system of traditional craftsmanship of Cairo, Egypt and the World Bank's attempt to "capture" that value as a market (and the lasting impacts its had)
[^0]: Markets of Dispossession: NGOs, Economic Development, and the State in Cairo
does anybody remember the original online Flash games? Newgrounds and Kongregate and all that? It was mostly people with hobbies. Passion projects. Rarely did the creators make any money off them even when their games got played by millions. When in-app purchases became a thing I saw a lot of these original flash games turn into mobile games with added in-app purchases littered throughout. Nowadays if you visit Miniclip or Kongregate and play a more recent game, you'll find the same. Either tons of ads or tons of in-app purchases
I think there's a strong cultural effect there once money is mixed in. Once upon a time online drug forums were filled with people sharing their "TEKs" about growing weed plants and freely trading seeds of certain varieties. Now in the aftermath of legalization you'll hardly ever find this type of stuff for free. I think the shrooms community is in a similar place as well.
OSS is now fueling multibillion dollar companies and maintainers are starting to demand compensation. Many will still contribute for "free" because it looks good on a resume, but I really think we're gonna see a strong decline simply due to the cultural impact of the transformation of Social Capital to financial capital
If you wanna get anthropological about this, I HIGHLY recommend this book[^0] about the social capital that fueled the social capital system that underpinned the system of traditional craftsmanship of Cairo, Egypt and the World Bank's attempt to "capture" that value as a market (and the lasting impacts its had)
[^0]: Markets of Dispossession: NGOs, Economic Development, and the State in Cairo
I don't think we really disagree, but I might be mistaken.
> large scale FOSS came first
FOSS, let alone large scale FOSS, appeared well into a long period of western imperial surplus and cheap fuel, in a unipolar world where the US had unique capabilities of projecting power.
> hobbies. Passion projects. (...) maintainers are starting to demand compensation (...) the transformation of Social Capital to financial capital
Precisely. When times are good, people will indulge in activities like that because incentives align differently. Maintaining enterprise-grade software for free in a contracting economy, on the other hand, will probably not be as tempting. Pay the rent or push upstream?
> large scale FOSS came first
FOSS, let alone large scale FOSS, appeared well into a long period of western imperial surplus and cheap fuel, in a unipolar world where the US had unique capabilities of projecting power.
> hobbies. Passion projects. (...) maintainers are starting to demand compensation (...) the transformation of Social Capital to financial capital
Precisely. When times are good, people will indulge in activities like that because incentives align differently. Maintaining enterprise-grade software for free in a contracting economy, on the other hand, will probably not be as tempting. Pay the rent or push upstream?
To be clear, I definitely think OSS maintainers deserve to have the financial stability necessary to produce their work
So... abusing free labor?
If someone gives something away for free, and you use it for free, that's not "abuse", that's just... taking exactly what was offered.
Until you start asking for free support, free consulting, etc. People are generally a-holes.
I stopped working on most of my open source projects because of it. I still open source stuff that I actively use / work on, but don't actively advertise the projects, and ignore any unreasonable request (all MIT, so anyone can fork as they please). Not worth the mental health burden, I get enough of that from my day-job :)
I stopped working on most of my open source projects because of it. I still open source stuff that I actively use / work on, but don't actively advertise the projects, and ignore any unreasonable request (all MIT, so anyone can fork as they please). Not worth the mental health burden, I get enough of that from my day-job :)
> I stopped working on most of my open-source projects because of it.
I've yet to get any traction on any of my open-source projects outside of myself, some old coworkers, and a few dozen random people. This morning while looking at someone else's semi-abandoned project I noticed there were 600 unaddressed open issues so I figured I would jump in and try to help some people. After answering the same question for the 4th time (linking them to the official documentation) I started to feel truly grateful that nothing I've released has become popular.
I've yet to get any traction on any of my open-source projects outside of myself, some old coworkers, and a few dozen random people. This morning while looking at someone else's semi-abandoned project I noticed there were 600 unaddressed open issues so I figured I would jump in and try to help some people. After answering the same question for the 4th time (linking them to the official documentation) I started to feel truly grateful that nothing I've released has become popular.
Well, you can ask people to pay for support, or you can ask them to shut the fuck up. You can also do what they ask from you. You are free to choose.
The real problem is that people were sucked into a social network when all they wanted was some easy to manage public code repositories. If that's a problem for you, fight the social network, but this is not an open source problem.
The real problem is that people were sucked into a social network when all they wanted was some easy to manage public code repositories. If that's a problem for you, fight the social network, but this is not an open source problem.
Or, "How Cheap-As-Free Software Can Help Consolidate More Wealth in the Hands of the Bourgeoisie"
Realistically, the money is going to the Bourgeoisie either way...
My memetic sibling in tech, allow me to introduce you to the world of democratic socialist policies.[0]
[0] https://en.wikipedia.org/wiki/Democratic_socialism
[0] https://en.wikipedia.org/wiki/Democratic_socialism
Could you tell me what you mean by the word “Bourgeoisie”? I thought it approximately meant “middle-class” or maybe professionals.
No “bourgeoisie” was meant to refer to capital owners (with “petit-bourgeoisie” referring to small business owners). Middle-class are the “proletariat” by Marx and co.
Incidentally, the underclass that many on the modern left focus on were pejoratively considered “lumpen-proletariat”.
Incidentally, the underclass that many on the modern left focus on were pejoratively considered “lumpen-proletariat”.
> Incidentally, the underclass that many on the modern left focus on were pejoratively considered “lumpen-proletariat”.
This is absolutely correct in terms of orthodox Marxist theory. It's interesting how "lumpenproletariat" has been whitewashed out of modern Marxist analysis. You just don't hear this term anymore.
This is absolutely correct in terms of orthodox Marxist theory. It's interesting how "lumpenproletariat" has been whitewashed out of modern Marxist analysis. You just don't hear this term anymore.
I just tried to get my head around bourgeoisie again - it is so confusing and I don’t understand why anyone would ever use the word. Whenever I see it, it is used as an insult, to mean someone that is more well off than the person saying it. The word “capitalist” seems mostly a better substitute.
For software, a programmer literally owns the means of production, so all programmers are bourgeoise? My guess is that bitwize is not proletarian, yet doesn’t see themselves as bourgeoise. I guess I struggle to understand whether proletariat and bourgeoisie convey any meaning in my country.
Disclaimer: I am definitely petite bourgeoise - that allows me to own a house like two thirds of New Zealanders do.
For software, a programmer literally owns the means of production, so all programmers are bourgeoise? My guess is that bitwize is not proletarian, yet doesn’t see themselves as bourgeoise. I guess I struggle to understand whether proletariat and bourgeoisie convey any meaning in my country.
Disclaimer: I am definitely petite bourgeoise - that allows me to own a house like two thirds of New Zealanders do.
> For software, a programmer literally owns the means of production, so all programmers are bourgeoise?
No, they are not. Marxist analysis doesn't really map to information technology in service economies in the classical way. The principles of analysis still apply in today's society, but first you have to recast class definitions and class relationships in a meaningful way. There's a nice (but idiosyncratic) book called "The Relevance of the Communist Manifesto" that explains this, but it won't make sense unless you already have a basic grip on orthodox Marxism.
No, they are not. Marxist analysis doesn't really map to information technology in service economies in the classical way. The principles of analysis still apply in today's society, but first you have to recast class definitions and class relationships in a meaningful way. There's a nice (but idiosyncratic) book called "The Relevance of the Communist Manifesto" that explains this, but it won't make sense unless you already have a basic grip on orthodox Marxism.
That's congruent with some of the eat-the-rich movements: they end up being the rich steering the poor to attack the middle class out of frustration.
High and low versus middle.
High and low versus middle.
FOSS is literally optional. Don't create it if you don't want your software used by people you dislike.
there are many faces to FOSS .. maybe one is cute, another is more like commodification of industry-core building parts, yet another is personal for the author and their realms.. etc
So free labour? No thank you. Open source devs should think long and hard about how their good will is being taken advantage of. Not just by mega corps reselling their software by saas or ais but also by individuals demanding features and support. It also simply reduced the monetary value of code to zero. Unless it cures cancer there is no need to make it available free of charge.
I honestly don't think this has much to do with any reality.
I don't think any current "tech downturn" or reduction in stock prices specifically is caused by company spending on proprietary licenses; I don't think companies saving an amount of expenses equal to spending on proprietary licenses would have much effect on stock prices even if it occured (well, it might harm the stock prices of an companies selling that proprietary software if done on a wide scale!); and I don't even think that switching from proprietary software to open source can end up saving you the money he thinks it can.