[noam87]

I've been thinking recently that there is exists a completely unxplored market in tech. Local shops based around FOSS / open hardware where designers, hardware and software people work together to build custom projects for their clients. I would call this person a "technology artisan", because the closest analogue I can think of is a tailor or a carpinter.

So, for example, a family would go to their local technology artisan to put together a smart home system tailored specifically to their needs. Or perhaps a local band would commision some custom-made Raspberry Pi-based synthesizer with fancy lights for a show (https://youtu.be/_nBK8sAl9nw)

To the consumer, it would mean not only devices tailored to their needs, but also ones that are cheap and simple to get repaired and extended, because they would be based on open standards. No lock-ins, no secret surveilance.

What makes me think this is possible is how big the hobby space is around all these hackable technologies (raspberry pi / arduino, 3D printing, electronics)... people are already building home media centres and farm bots in their back yard just for fun. Hacker spaces are already a thing -- make them a service!

[icebraining]

Custom software development is not an unexplored market. There are literally thousands of companies, many of them small & local, offering those services. It's just that we do so to other businesses, since it's way too expensive to develop custom stuff for single individuals.

I've built a coin-operated electric candle stand[1] for a local theater group. I'm just an amateur at electronic circuits, and I'm sure a professional would have spent less time on it, but even at half my hourly price, I would have blown half their production budget.

You could say some builders would be willing to take a pay cut to work within your community, but frankly, doing one-off toys to upper-middle-class people is not my idea of community service. I'd rather have a well paying job writing B2B software and then work for free for people in actual need.

[1] a copy (with extra features) of something commonly found in catholic churches around here: https://www.nextnature.net/app/uploads/2010/05/led_candles.j...

[wiz21c]

Completely off-topic but you sure don't want to work for free (hint: the problem with that has nothing to do with the money)

[icebraining]

I've heard that before, but I'm not convinced. Supposedly clients treat you better if they're paying, but in my experience, the correlation doesn't hold. At least by not charging I feel freer to walk away when I'm not being respected.

[ghaff]

Yeah. I'm not convinced either. I do little things for various small non-profits/volunteer organizations and I've basically never had anyone bug me because it took me a bit longer to get to something than I anticipated. If I'm paying someone market rates to do a task, you can be sure that I'll be doing the "when will it be done?" thing if they get behind schedule.

[wiz21c]

My specific problem was : client claiming intellectual property over my work. If I work for free, my work is my work but somehow the claim was made.

Then you have bugs, months after delivering, and you can't say "no I won't fix it". So although you don't want to work anymore, you still have to.

Then you have new functionalities. You are burned out with the project but there is this little thing that your friend/customer really needs. And you don't want to do it, and since you've never been paid, you'll do it for free again and you'll get burned out some more... So no compensation.

Now you can argue that the problem here is not the money, but the lack for a proper contract, that'd be right. But the money question automatically brings the contract question. That's why I say : don't work for free, ever.

[erikpukinskis]

Your instincts are correct, you're describing the future. But there's a problem preventing it from working today.

Your idea only really works if you can get a large number of these artisans on board, across many disciplines, and keep their "dance card" stocked.

The issue is that solving any problem in software costs about $1000. And typically if you want to solve one problem you need to solve 2-20 others first.

Your mention of open source is key here. Every pizza place in the world probably has a dozen information problems that would be worth solving, but most of them aren't going to be willing to spend $100,000 to do it, even in 10 installations.

But of those 10 problems, 9 or 10 of them are not core IP for them. So something like 95% of the budget could be split across like 1000 pizza places, driving that cost to $1-20 which is a much more manageable cost.

So you need some sort of bond to finance these projects across 1000 clients.

And second, when you split the design work across 1000 clients all paying in a different moment, you end up with 990 little gigs and 10 big ones and the 990 are too small to be worth the overhead for your artisan. To make those worthwhile you need some infrastructure for turning those contracts into 5 minute non interactive (no customer support) deliveries, and packing them into a stocked queue that can pay the artisan'a rent. I.e. If there's one new pizza joint on a Tuesday afternoon, there are 50 other 5 minute tasks that they can also do so they aren't waiting around for pizza joints and getting nothing done.

And when a pizza joint needs analytics while that artisan is on vacation, you need another artisan who can step in.

So that's a two sided market, which is a notoriously hard startup problem.

And lastly, these information problems are not all code problems. There is design and stats and customer service and other stuff mixed in there. So you need a stable of design artisans, customer service artisans, etc. which means we're talking about an n-sided market.

Which the startup world has, as far as I'm aware, no known examples of.

It'll happen. It's just messy.

[icebraining]

There are plenty of small companies doing exactly those kinds of projects for SMBs; I worked for one for five years (it's still doing fine, I just moved on).

Rather than a bond, which would be unwieldy to implement, we'd take a risk: we develop it for the first few customers for a percentage of the cost, but keeping all rights to the code so we can then reuse it. Some projects pay their bills and some more, others not so much.

As for the n-sided market, we used the typical solution: having employees.

In our case, we used an open source platform (Odoo) which comes with the generic business modules - CRM, Accounting, Invoicing, ERP, etc - and which allowed us to build integrated functionality for the specific industry. We had clients of all sorts, from one-man dental offices to international clothing manufacturers.

(Disclaimer: I still work with Odoo, but I was never an Odoo SA employee)

[erikpukinskis]

A risk is a bond you sell yourself.

I would argue the issuance of the bond and the separate market for holding risk is key.

Developing a market is key to being able to fund smaller and smaller client segments. This is what I talked about above... with a single origin of corporate ownership there is an upper bound on the number of managers. You need to separate into two markets to meet the entire demand. It's just a matter of graph traversal distances. N corporate boards cannot manage N^2 managers.

But N^2/10 agents can manage N^2 managers. And they take 25%.

[orf]

Wouldn't that be insanely expensive compared to off the shelf solutions and result in a hodge-podge of insecure, half baked solutions?

[wtallis]

> a hodge-podge of insecure, half baked solutions?

That's exactly what the commercial market of Internet of Things devices is so far. You couldn't really make it worse by putting a replaceable ESP8266 or BeagleBone at the heart of each of these gadgets.

[adrianN]

After two or three similar people have had their custom thingy build the next person could use the "off the shelf" open source solution with a little customization.

[TeMPOraL]

Off the shelf solutions are already insanely expensive, because companies are still milking people on the "smart" label.

[seppin]

yep

[karllager]

> So, for example, a family would go to their local technology artisan to put together a smart home system tailored specifically to their needs.

I am a software developer and consultant and you do not figure how insanely expensive I am - when your family comes through the doors of my artisanal soldering boutique and demands a custom built home automation system, I would suggest them to invest the money in a long, long vacation with the whole family instead.

[Veratyr]

Sure but you wouldn't pay a chainsaw sculptor to cut down a tree, you'd pay a lumberjack. You can already go to a hackerspace and pay a guy a couple hundred bucks to build you something basic so long as it's mildly ineresting.

With kids growing up playing with Raspberry Pis, in 10-20 years I don't think your comment's parent is such a farfetched idea.

[drdaeman]

Unfortunately, that's also the story of how Internet of Things (well-known by other nouns) was born.

This works perfectly well for non-so-smart stuff (thermostats, automatic cat doors, etc etc) but when you increase complexity (smart home system) you need engineers with higher qualifications and better expertise. And that's expensive (unless someone volunteers and works for pennies, but that quickly gets stressful - literally not worth it).

(Not like off-the-shelf solutions are any better. They were also born that way - by lowest spending necessary.)

[amelius]

> So, for example, a family would go to their local technology artisan to put together a smart home system tailored specifically to their needs.

I'm getting a bit tired of the smart home system example. We've seen projects for smart home systems in DIY magazines since the 70s. These projects never really took off, and turned into businesses (Nest being a notable exception). It's something people don't really need or even want.

[CaptSpify]

If people don't really want it, why do they keep paying for it? IoT products are selling like crazy.

[rsync]

"So, for example, a family would go to their local technology artisan to put together a smart home system tailored specifically to their needs."

... and the artisan would refuse, and explain how much simpler and functional their lives would be without any "smart" systems or appliances ?

[driverdan]

Hackerspaces / makerspaces are full of this already even if it's not their primary focus. For example, I know two people at hackerspaces who make large interactive art pieces for concerts and festivals.