ITT: open-source, #FediVerse, and how it can influence private sector and create change.
Note up front: I'm going to use words like "decommodify", "consumerism", "capital", and "rentierism" a lot here, because I need the vocabulary. I am not a Marxist or some other kind of radical, nor wholesale anti-capitalist, and certainly not a revolutionary. I'm a social democrat, progressive, and reformist. Keep that in mind if you reply to me.
Something I don't think many people realize is just how much the OSS movement has altered the private sector.
Sure, there have been the SCO v. Linux suits, and the battles with old Microsoft, both of which OSS won, and it's certainly produced a huge amount of software at this point.
But look back to the software industry of the 90s, and the picture looks completely different. [1/n]
Back then, there was a whole parasitic cottage industry of companies that made a living off of selling the kinds of software packages we take for granted today.
You wanted crypto, you bought a closed-source library for 4-5 figures. Same for all kinds of protocols we take for granted today.
Beyond that, standards were frequently closed, and cost that much. ASN.1 was an example of this.
@zorinlynx as a kid in the 80s/90s who was interested in computing, this was a huge impediment.
A compiler for C back then would cost you $1000+. For a lot of us, it meant the best we had was QBASIC, which was... Not Great. Hell, an assembler was over a hundred if I recall.
The only people who got to learn with real programming languages were people whose parents were in the industry, or academics, which wasn't me.
It was more profitable to deliberately produce broken/non-compliant implementations, and make them deliberately just short of too buggy to buy.
This had a serious negative impact on a number of standards. ASN.1 and CORBA are two examples. In essence, there was zero incentive for interoperability, and lots of incentive to sabotage it. [3/n]
OSS has essentially wiped all this out by making alternatives freely available. Nobody in their right mind would pay out 5 figures for an SSL implementation when we have OpenSSL and 4 other alternatives.
Also, however many issues OpenSSL has had, closed-source libraries tend to be much worse. Sunlight is a good disinfectant.
Put in terms of political economy, this is a significant degree of decommodification. [4/n]
This is directly responsible for the productivity of the software industry over the past 2 decades. The startup economy as we know it could not have existed under the previous rentier system.
There's no way you could have launched most startups in a world where you have to pay out 4 figures a head for an OS license, then that much again a head for a compiler, then either buy or reimplement every single package you want. [5/n]
There are some very important lessons to learn about the relationship of #OSS and the private sector.
Those who were around in the early days remember how rag-tag early OSS was. Anybody remember battling to get their network card working? Remember early Gnome/KDE? It was scrappy, a headache to set up, often ugly, and the UI/UX was terrible.
By the late 2000's, Linux desktops were beating Windows Vista in UI/UX. [6/n]
The reason for this is that OSS works on completely different dynamics from for-profit software.
It is highly persistent, it tends to monotonically improve over time, and it is very hard to shut down. OSS is at its best when it embraces differentiation followed by cross-pollination. This enables it to explore alternatives and find ways around obstacles. It also makes it very hard to kill.
This is something for-profit simply cannot do. [7/n]
@emc2 High-level observation that draws from ^^: For-profits can do a “great” job (meaning successful for them) capturing, cannibalizing, and killing other for-profits… but name a solid, popular OSS that a for-profit has successfully killed off and replaced with its own closed standard. There aren’t many of these stories. If anything, it’s the inverse — OSS can and has killed many a for-profit space, OpenSSL and Sendmail being just two examples.
At this point, I actually don't think OSS would have been killed had we lost the SCO v. Linux lawsuit itself. It would have been a massive setback, but an alternative would have stepped forward: BSDs, L4, or something. It's like Hydra: even if you manage to kill one project, two more will take its place.
It also snowballs over time. The OSS ecosystem of today is massive compared to the old days. [8/n]
There was also a whole conflict over DRM. That's its own story, but the OSS world has fought and won several key battles to keep computing platforms open, and prevent a whole layer of rentierism from being set up. RIAA/MPAA were major opponents in this in the 2000s.
So bringing it back, OSS has a significant effect on the for-profit sector simply by existing. [9/n]
#FediVerse had an additional barrier, that social media has a critical mass of users. For reference, I joined #Mastodon back in 2017, when it was below critical mass.
The critical mass effect seemed to have #OSS attempts to affect social media behind the eight-ball. The #TwitterMigration and #redditMigration revealed a very powerful approach for getting around this. [11/n]
Both of these reveal a shortcoming of for-profit social platforms. Cory Doctorow's concept of "enshittification" sums it up succinctly.
In more detail, a for-profit platform is under a mandate to squeeze more and more profit out of its users. In a social media platform, this means more surveillance, data gathering, "use the app", and ultimately gouging where possible. [12/n]
Something else to note: the profitability of surveilling users, collecting their data, and selling it is steadily declining. The whole adtech world is constantly climbing uphill against a landslide.
The takeaway from all this is that for-profit social platforms will eventually create a crisis for themselves. They degrade, trying to squeeze more and more profit out of an increasingly arid source, until they eventually do something stupid and blow their foot off. [13/n]
What happened with Twitter is that #Mastodon happened to be poised, almost by accident, to scoop up enough users to rocket past the critical mass point, and became self-sustaining.
What's happening with #reddit is similar, but it's also enabling something analogous to a strike by the mods.
This is essentially Naomi Klein's notion of Shock Doctrine, except being employed against capital, not by it. [14/n]
One of the biggest strengths of the #OSS movement is that it does not hesitate to use the tools of its opponents against them. Pretty much the entire left is uniformly and vehemently opposed to that idea. OSS by contrast does so enthusiastically. I suspect it's part and parcel with the hacker mindset: pulling off political-economic zero-days is right in line with how we think. [15/n]
> One of the biggest strengths of the #OSS movement is that it does not hesitate to use the tools of its opponents against them.
I see this going too far when depending a lot on GitHub or Discord instead of promoting GitLab/Codeberg/Forgejo and Matrix for example and it seems to me that today we need a constroversy in each proprietary platform to start using a FLOSS alternative... luckily there are more radical people that cultivate those alternatives for years.
@alexl It's a process, and its been going on for some time. Back in the old days, people had to do OSS development on DOS boxes with proprietary compilers.
I think the GitHub/Discord thing is just lack of priority. They aren't a significant problem, so we use them. If that changed, you'd have a dozen alternatives pop up overnight.
I am a fan of how KDE was pragmatic when supporting users of proprietary services like Google but strict with its own infrastructure (and mostly is today).
But my point is that those alternatives don't just "pop up overnight" like you said, they are developed, often for years by provident people and I would like that people moving now to Mastodon and the Fediverse help more with the rest (GitHub and Discord are examples) 🙂
Lemmy already has a foot in the door to Codeberg/forgejo and/or their own gitea server: see the thread https://lemmy.ml/comment/769383 where people who seem to be key devs @nutomic + @PureTryOut responded rapidly;
"Didn't mean to trivialize people's work." Sure, I didn't mean to imply that. It's rather that YMMV - in some cases the ecosystem may have software packages ready to "pop up overnight" - in others they may be too immature and not ready. Biological ecosystems do not always adapt rapidly to challenges.
SCM wars = RCS wars? [1]: I found cvs cumbersome before that period; I was glad to use git afterwards and find it elegant. :)
I'm old enough to have used CVS professionally (in a non-dinosaur company). I used subversion through grad school, stuck with it through the wars. Moved to git at some point in the very early 2010s.
So now #TwitterMigration fed us enough users to get #Mastodon to critical mass, and then #redditMigration proved it can be replicated. We have a model that seems to work.
Enter Facebook. They're already in a death-spiral, and I suspect somebody over there figured out essentially what I've said here. So they want to try to jump in and get out in front of it. It's worth gaming out what they're up to. [16/n]
Two tactics that have actually succeeded in being an impediment to OSS have been carpetbagging (showing up and using organizational weight, presumed prestige, etc to shove out a project's leadership and take over), and de-inventing (my term: what Google did to XMPP, RSS, and is currently trying to do to email). Google has tended to employ both of these in its quest for hegemony.
If Facebook knows what they're doing, this is what they are planning to do. [17/n]
@mojala "De-invent" is a term I made up to refer to a specific practice. It refers to a strategy for removing a common interoperable protocol or service from the public domain and replacing it with a cloistered, proprietary, often very heavyweight one, often for the purposes of harvesting user data.
For examples, look into what Google did with XMPP and RSS.
@oceaniceternity I consider "de-inventing" to be an outcome, not a specific strategy (EEE is one way of achieving it).
Recently, OSS has taken to building projects that re-create de-invented protocols and services, specifically aiming at high levels of interop (Matrix, to some extent the Fediverse). I think that is a very effective approach.
If Facebook thinks merely being part of the Fediverse is going to somehow magically going to save them from a reddit-style catastrophe, they're sorely mistaken. That will only make it easier for people to leave when that time comes (and it will).
So by this analysis, #ActivityPub and other protocols, standards, etc are what need to be protected. More broadly, OSS would do well to develop better defenses against de-invention anyway. [18/n]
As for the question of whether to federate with Facebook, I go back to differentiation and cross-pollination. Diversity is our strength. I have complete faith in the ability of OSS/FediVerse to outmaneuver a gigantic, ailing behemoth.
Somebody will figure out the right move at every point, and the rest will follow. The protocols will grow and adapt, and it will be on Facebook to keep up. [19/n]
So in closing, this is a very exciting time. #OSS has created a lot of change for the better, and back in November I felt like we were starting to remember that.
I think this model, worked out and refined over the years is really quite powerful, and recent events have showed that. It's consistently overcome obstacles that were said to be impossible.
So I'll close with the exhortation to dare to dream big about what else we might be able to accomplish.
@emc2 That's not an analysis, that's hoping. I'm not against FB coming, but FB should be better regulated by law, and it should be split up in different compagnies. FB, Instagram, Whatsapp, Messenger, Oculus, Horizon Worlds, Mapillary, Workplace, Diem... they should all be split off as independent compagnies. Moreover: their independence should be protected by disallowing big investers of having lots of shares in several of those compagnies.
@emc2 From an article on the bbc: "The idea of the fediverse is it is like email. Someone on Gmail can exchange emails with someone using Hotmail, for example, and the fediverse could be described as that idea applied to social media.
At some point in the future Meta wants users to be able to use their Threads account to interact with other social-media platformsé
I don't have objections against that. If somebody is on FB now, I too have to have an FB account to communicate with them.
@vsaraph They're ailing badly. Their earnings statements are consistently dismal, they've done several waves of layoffs, and their last two big projects were large and expensive failures (their cryptocurrency, and more recently the metaverse). They've lost key leaders, and seem to have no real vision. They've also lost a lot of users, and the ones they've kept are skewed towards old age.
@vsaraph Facebook the product, I'm less certain about.
There might be a road to saving it, but my sense is it's somewhat outpaced at this point. I'm quite far from a product guy, though; maybe there's someone out there who could buy it, rebrand it, and make it popular again.
@promovicz I view this as one of the most significant successes of OSS. It has so fundamentally changed the rules that for-profits routinely give away not only products, but funding.
The same people who charged us 4 figures for a dev toolchain in the 90s are now giving us the source code.
@promovicz Yeah, marketing ideas is going to be a thing in any system, even if there's no money involved. You're always going to have to get people on board with your ideas. Something I've come to terms with as a researcher.
@lori it's largely rendered moot by precedent and "fair use". This was an issue with the fight over DeCSS, and it comes up every once in a while with some closed, proprietary devices or standards, but we've basically won this fight.
As for things like and OS, a compiler, or an application, there would be no grounds to sue unless you can show copyright infringement.
@lori technically able to vs. Legally allowed to are two different things. As I understand, the DMCA fair use definition allows any lawful owner to do reverse engineering in order to lawfully use their devices. This got relitigated wrt phone unlocking in the early 2010s, and is being reinforced by the current movement for "right to repair".
@emc2 great thread. I am similarly optimistic. Although figuring out the right move may also involve blocking Meta and similar organisations at some points.
@Loukas I am personally on the side of not blocking them. I think Facebook has a lot more to lose than they have to gain, and in ways they perhaps can't comprehend (e.g. users seeing a readily-available zero-ad experience and wanting it).
I am also pragmatic and outcomes-driven, so I'm not categorically opposed to blocking them, nor do I think everyone needs to make the same move.
@emc2 thank you for sharing your thinking about this. I share your pragmatic approach. The reason I think Meta server(s) will get blocked quite frequently is because if we simply maintain the same standards, of defederating servers that allow bad actors and which don't moderate to stop hate speech, then I think all or most Meta severs will fall foul of that.
@emc2 I once started an initiative on how commons should deal with profiteers, since it's a pendulum effect (once they turn all high-trust communities into low-trust ones, they roam looking for sustenance).
We definitely need a glossary so we understand the scope of what we're discussing. Digital diaspora will only grow.
Add comment