Learn from engineers, designers, academics, startups, and indie developers who are putting local-first into practice and reaping the benefits of a cloud-optional architecture.
I had a similar hunch and am glad to see it resonate with people I admire so much! To be more specific my vision was that with the introduction of LLMs that can code many subscription-model and freemium businesses can be put to the sword. This will have the effect of thinning the forest of the trees in the middle - problems that are solved poorly for free, that entrepreneurs offer to solve well for a modest amount of money. If these AI-enhanced barefoot developers do their job right many profitable businesses will face competition from truly free alternatives. This was not possible before because the effort involved and the opportunity for profit available drove people to make professional apps and charge users. These services proliferated a lot and now they are ripe for that forest fire. But the thing is, what will be left will be a fat and growing number of free barefoot software on one end, and an entrenched monolithic empire of Big Tech software on the other end. The middle will also be under constant pressure from people's ever-shrinking disposable income (see the Surplus Energy Economics blog). But this middle is the target of many people's aspirations. Earning a living as an entrepreneur on the Internet, or RU-vidr, etc. As these opportunities taper off the barefoot developer challenge will be replacing the wealth derived by spending money on a handful of subscriptions and paid apps with their free alternatives, built using AI LLMs. Sort of like in the before-times when a lot of the economy was not monetary, but services provided free of charge in the household. For example meal planning apps. Or an even clearer case, plain-text accounting - there are many shiny subscription-model finance apps out there, but beyond the command-line wall lies hledger, or beancount. These work, they're local-first, and can be gotten to do almost anything you want with your finance data. But it's effortful and frustrating to string together commands to get text reports. And visual reports involve even more effort. But eventually this can be solved. And then more and more people will no longer have to pay that subscription.
Would like to propose that we do already have a giant sub-industry within tech that provides charges based on the value it provides rather than the threat to delete data, that we could port over business models from - gaming. Be it f2p, microtransactions, or “battlepass” type models, i think there’s a lot we may be able to learn and apply over to traditional b2b / b2c software as well
Tuomas is the hero we don’t deserve. It’s still a puzzle to me that I can sit and watch all these pioneers share their insights into how they build state-of-the-art applications, for free. What a time to be alive. Thank you for another great talk.
Very interesting. But when I am thinking of it I only can imagine some apps where part of functionality is local first. I can't imagine case where I don't need some centralized data and logic at all.
As I understand it, this is a good usecase for work that involve a couple of people, like in a company (Notion, Excel, Google Docs). Not your social media or youtube stuff.
Well done: informative, very clear and easy-to-grasp graphics that punctuate the narrative, and confidently presented. Overall, it shows that you've put a lot of thought into this topic. Thanks for your effort.
This is great. But I would like to know what happens when you want to enforce certain rules (that you can not express in a wchema) or what if you need a different type than the one for editing (where everything is optional)
Nah, indie hackers (xxi century version) are greedy for MRR and stuff like that 😅 it’s more about the classic software hackers of 80s-90s which did build crazy stuff they envisioned just for the heck of it… but still not the same… I like the idea anyways 👍
Oh my goodness, I have thought about this for some time. I can see a substantial rise in SaaS development catering to niche workflow tasks, and they are genuinely gaining traction. Furthermore, a speech delivered by the founder of Patreon at SXSW resonates with the concept of the "barefoot developer" idea. The talk name was, "Death of the followers".
Cool talk. Gets to an interesting point. I've solved both of the hard problems, but it's taken me a custom language with efficient closure serialization to do it. (eyg.run) from the otherside sharing DB's across apps is awesome.
Very interesting talk. The "hypothetical future sync service" sounds a lot like a Jakarta Messaging to me. JMS already offers a standardised means of asynchronous communication whereby changes made on one device could be published to the service and picked up by another device which is subscribed to the same service.
The speaking slots were only 25 minutes and there were some technical difficulties that took time away from my slot which were edited out of the video.
@@AaronBoodman Would love to know about the Zero cache to database mapping, and whether this is closer to Replicache's more BYODB or Reflect's opinionated storage.
@@judahdsouza9196 it’s closer to Replicache. However you don’t have to write your own pull endpoint. Zero-cache consumes your upstream database replication stream and exposes it directly to the client (applying auth of course).
I've been following this movement from a bit of a remove but I’ve been and remain extremely psyched about it. It seems this was a highly impactful conference. I wish I could have been there, but in lieu of that being able to watch the talks via the channel has been wonderful. Thank you to all the organizers and speakers!
@Jack McCloy, any thoughts about the challenges of commercialising local first software? It seems that's a big hurdle to overcome, since a lot of developers like building SaaS because it gives them an easier way of monetising their software.
