Been having a discussion on another board regarding the “open source” movement, and whether it has been a force for good or bad. And I have to share my thoughts on this, that it has been resoundingly terrible – the open source movement has been the worst thing to happen to computer science since the invention of the bug.
I can think of at least a half dozen companies that had truly innovative and competitive products, companies which died as a result of the plethora of half-baked “free” versions of similar software which never lived up to promise, and which invariably failed in critical ways. Some of those companies were gearing up to offer competitive desktop OS software that could have forced Microsoft to the table and could have generated honest competition with MS.
Instead, we get a constant barrage of 80%-finished products which have no coherent product management behind them, which are almost never backwards-compatible (or often forwards-compatible). Often including “poison pill” GPL licensing that prevent any sane company making a finished version.
Oh, and did I forget to mention that all these “god’s gift” open-source products are rarely given anything more than a cursory testing regimen? And when no one is QC’ing this beatch, anything could and will go wrong with them. I recall all the numpty Linux heads back in ’99 and ’00 trumpeting about how having open source was more secure than proprietary software…don’t hear much out of them any more, not since it was discovered that SSL had a major breach which had existed for twelve years being exploited with no one noticing.
Nope, open-source is a hobbyist infection that has caused immense and crippling damage to the entire industry of computer science. It reminds me of the obsession with Communism as a utopia back in the ’50s. Lots of idealists with no real plan for how to actually make it work as intended.
Look at what’s going on in advancements in computer science just this year. AI developments are going bananas, with all manner of cool opportunities coming up. The drivers behind these? Proprietary software companies competing with one another. Not one single open-source project is in the news as a source of innovation, among what, like six new releases this week alone? I will grant freely that TensorFlow and PyTorch are popular tools within certain AI circles, but I will also suggest that these tools succeed not because they are open-source, but in spite of that. They are single-purpose tools, not AI products themselves.
Open source is a failure. Worse, it is actively retarding the advancement of computer science and engineering. It traps good minds into thinking they are accomplishing something, when in reality they could have been contributing to a better future for everyone – including your grandma, who still doesn’t know how to use her computer. I submit we’d have had “Cortana” or “Siri” or (insert name here) being a genuinely useful version of “Clippy” driven by something like GPT3+, twenty years ago, if we’d not had this massive and worse-than-useless distraction. If all those good intentions had decided to put themselves to work and formed a collective association of ethical coders instead of the dippy “protest coding” it turned out to be, the world would be a very different, and very much better place. We – as software professionals – would also have had an opportunity to form a power base that could influence the companies that shape the future, and drive it towards better practices.
But we didn’t. And we have been paying for it for decades.
I will also step down from this pulpit for a few moments to point to places where “free” options have been extremely successful – highly-focused, simple applications. In these spaces, my argument falls short. PuTTY, FileZilla, Apache, Chromium, again TensorFlow, Docker, these spring immediately to mind (though in the case of Docker, I think it was just some script-kiddies who couldn’t figure out what they needed to build a clean installation script, so they just decided to clone the development desktop environment instead and call it a “dock”). They are very narrow in scope, and have very little “wiggle room” for interpretation of their purpose. A talented developer could whip something like that up over the course of a summer and call it done. They also don’t present themselves as highly-marketable apps – I can’t think of the last time someone in my orbit who actually paid for web server software or an FTP client. I will openly admit that in cases such as these, the open-source community has stepped up and provided apps which probably wouldn’t have found a simple commercial solution.
But all that said, I stand by my original point – we are far worse off having the open-source “movement” around, than had we let the commercial proprietary companies have at each other. We still would have had “freemium” software and the variety of “private” projects by hobbyists acting under non-profit status, and we also would have had a lot more success as an industry.