4 ms·
A few thoughts: (1) Most open source projects do not provide the outsized value impact that their creators seem to think they do. Most projects exist around th
by 013a 7y ago
A few thoughts:
(1) Most open source projects do not provide the outsized value impact that their creators seem to think they do. Most projects exist around the realm of "its 20% easier to install this than build it ourselves, so lets grab the dependency". That's nice, but lets be clear here: Standard.JS is not building Rails, or NodeJS, or Express, or React, or another seriously valuable project that saves companies thousands of dollars.
(2) Open source maintainers are deathly afraid of restrictive licensing. That could mean GPL, or it could mean a straight-up enterprise "pay for a key" licensing. I think these are fantastic ways to level the field between individual OSS devs and billion dollar companies, because most billion dollar companies take licensing very seriously. Where does this fear come from? I think a big part is that these OSS devs "want their cake and to eat it"; they want the ego and community of an altruist, but the income of a Capitalist. Another component is point (1); if you really tried to place a number on the value of their project, they'd be disappointed with the number customers would quote. They don't want to find out what that number is; they'd rather just ask for donations (of which they get very little, because, again, this software by-and-large does not provide outsized value).
(3) Its crazy to me that companies would pay for advertising space in an npm install script. I'd bet developers behind projects like StandardJS went to these companies and said "look we get millions of installs every month, that's millions of impressions", and I hope the companies are smart enough to know that (A) the vast majority of those installs are in unmonitored CI scripts, and (B) even if they are local, the text flies by so quick, and (C) even if they aren't quick, no one watches that progress anyway. Even if there's no deception today, one can easily see that monitoring "impressions" here is very difficult, and thus its very difficult to place a quantitative value on that ad space.
(4) The inability to sustain yourself on open-source development is about equivalent to "I can't sustain myself on volunteering time at a food kitchen for the homeless." If you're actually having funding problems to live your life, its time to reassess. Again, this sounds like some of these projects have some egotists behind them, or maybe they just feel like they can't leave (hint: you can. you're human. your priority is you. not the billion dollar companies who use your software without paying). Find help. Take a break. See if your full-time company can "sponsor" development by letting you work on it 5% of the time.
- 19ylram49 7y ago> That's nice, but lets be clear here: Standard.JS is not building […] [a] seriously valuable project that saves companies thousands of dollars. This is quite debatable and hard to measure, given the problem that Standard solves. As an anecdote, I just started a new TypeScript-based project a few days ago and I’ve spent the past few days mostly dealing with configuration. That’s perfectly fine to me for a personal project where I’d like to decide every aspect of the configuration, but let’s say this was for a fast-moving startup, it’s probably a smarter idea, for most projects, to simply adopt Standard et al. and call it a day. Practically speaking, if you consider a software engineer’s salary and other important factors, I don’t think it’s fair, or accurate, to claim that Standard doesn’t save companies thousands of dollars; it almost certainly does