~singpolyma/notes

Software freedom advocates have always said "if you can't write the code, at least you could hire someone to do it". But is that a realistic option? Most changes a small number of users might want are too small for someone to do out of interest (unless it is self-interest where the developer also wants to do it), too small for someone to do for money (since the money would not be much), or if they would do it for money the amount of money that would be attractive is too much for the users to reasonably spend.

There are a lot of factors at play here, but at least one is that good coders are expensive, and if you make 6 figures by day you're more likely to do interesting work for free by night than side-hustle for comparably smaller sums.

However, if a lot of desired bug fixes and changes are in fact fairly simple, perhaps it does not take good, experienced, productive coders to do them. Someone just needs time and motivation (and, probably, mentorship) to accomplish the task.

One idea might be to run a "code bootcamp that pays you" / "low-wage software consultancy" that hires people of little-to-no experience and pays them a living wage that is well below the market for Big Tech Coders. Somewhere in the $30-40k CAD range, perhaps. These workers are getting on-the-job training, but also real jobs with valuable work that does pay their bills.

Of course, initially there might be a lack of clients who want to pay the lower amounts for modifications. But we have no shortage of users who want free software modifications made! There are millions of tracked tickets on every project of any size that represent changes at least one user has wanted made. We could garden these and provide endless, useful work, hoping to eventually make some of it also bring in revenue.

Claiming bug bounties might be one way to get (small) amounts of revenue, though these are often posted for much more complex features instead of small changes, and you tend to need to get code merged upstream in order to claim them.

Speaking of upstream, while it would be useful to try in many cases (and good experience also) to get changes merged upstream, there is no reason this always has to happen. If only a small number of users want a change, it may be very valuable to them, but there is no reason to burden upstream with maintenance of this niche modification. Creating a Guix channel (or similar) for a variety of small-time forks that may be used by users who find them useful would still be very helpful.

(By analogy: just because you want your chair upholstered so, does not mean the factory will make a line of chairs like that. That's why you have re-upholsterers.)

Hacker-Upholsterers would keep their copyrights but be contractually bound to release them according to the license of the project being modified, or if doing greenfield work to release under (AL)GPLv3. They would all be independent contractors so they may work from any location and handle local taxes themselves.

About this wiki

commit d91398fe83ec706fb4492f4b1efb58d1a51a5f5e
Author: Stephen Paul Weber <singpolyma@singpolyma.net>
Date:   2020-12-14T19:58:17-05:00

Convert from org to md
Clone this wiki
https://git.sr.ht/~singpolyma/notes (read-only)
git@git.sr.ht:~singpolyma/notes (read/write)