Why I Don't Like Open Source
Why I Don't Like Open Source
Join the DZone community and get the full member experience.Join For Free
Easily enforce open source policies in real time and reduce MTTRs from six weeks to six seconds with the Sonatype Nexus Platform. See for yourself - Free Vulnerability Scanner.
Please read it. Carefully. Get all the way through and reserve judgement (if any) for the end.
It's not the first time Adam's thoughts have inspired a blog post either! These are my thoughts on open source, contributing and the points raised in the discussion.
This post is partly edited, but quite long, so might not be 100% coherent. For that, I'm sorry!
(Not) contributing to open source isn't the problem. The pressure coming from external and internal sources telling us that we should be contributing is invalid, in that it should not exist, but it does.
How we think people see us is only a reflection of how we feel about ourselves. The project getting the most retweets today is not representative of real work. Nor is the person with the most followers the better human being.
There is distortion in the web community that you need to learn more, contribute more, work more, know everything, be on the latest technology, and so on.Unsubscribe. Unfollow. Turn off alerts and notifications (or at least make your "quiet hours 6pm-9am). Find other human beings.
I want to say that Adam's tweets are not the fault of open source, but our industry's distortion. But his tweets and the subsequent replies prompted me to also write down my thoughts about open source that I've been mulling over the last few years.
It means different things to different people. That's okay.
To me, it means:
Welcome. Welcome to this code I wrote. Help yourself to bits you like or need. Help make it better for others. Help make it yours. Help make me better with your suggestions, changes or discussion. Use the code in any way you want, in ways I never thought of. And if the projects I created lives on without me, then open source has worked.
You need to decide what it means to you, and what it looks like. Which leads me to ask: is a project really open source?
There's the big, high profile projects and then there's everything else.
When I'd hear "open source", I'd think: Linux, Apache, Firefox, bootstrap, ember, node, etc. i.e. big fucking high profile massive open source projects with a fuck tonne of code*.
* Okay, there's also a decent number of sensible codebase sized projects that are used a lot, passport for node for instance
This is Type One.
Type Two open source projects are the kind of project where the source is open, and the license is "meh"/"whatever I always use but don't really understand". For me this is MIT. But that's beside the point.
The source is available. Just like the web. View source. It's the kind of open source that I'm familiar with (obviously there's discussions around licenses, but that's for another day).
Type Two projects: the kind where the source is just somewhere on the web, is the kind that, if we're honest, hardly anyone else is going to contribute to. I'll be lucky if anyone actually uses it, let alone skims the code.
If someone sends a pull request to one of these projects, there's a few things that can happen: it's ignored/forgotten about or it's blindly merged "meh, cool" (there's other outcomes too!).
For me, putting code in github as "open source" is for two main reasons: 1) so if I lose the code locally, I can find it again (fingers crossed Github doesn't vanish too quickly), 2) I can't be arsed to pony up the pitiful amount of dollars for more private repos.
The majority of the repo owners on the type one projects are companies. It's their product. The majority of type two, are individuals. Let's remember that perspective. Contributing to the small ones isn't a big deal and potentially contributing to the big ones might be like pissing in the ocean.
I can't speak for Adam, but it sounds like he's either felt pressure to contribute to some projects either externally or internally. I could be way off based, but I doknow people who've felt this way. Heck, I feel this internal pressure sometimes.
Contributing (code) to the Type One projects isn't something you can do with a few hours. So seriously, if you feel this pressure: stop it.
It takes hours, if not days to become familiar with the codebase. Then creating a bug fix or a new feature isn't quite as simple as "hacking some code in". If it's a high profile project, it'll come with tests and coding guidelines. That's more work, and you've not even contributed a single line yet.
For context, in late-2000s, I partook in a jQuery day long sprint to fix bugs. I'd say I knew jQuery pretty damn well at the juncture. By the end of the sprint, I'd managed to confirm just one bug, and pretty much have an idea of what caused it.
"Contributing to open source" is a long term commitment, and you do have to pick and choose. There's a few people like Substack filling npm with his node modules is great, perhaps reinventing every wheel, but he's also doing it to cover his living. But these people are the exception. This isn't a lifestyle that's available to most people (and frankly, npm is running out of useful names for projects!).
Ask yourself what is open source. Ask yourself if you want to contribute to that, and if you do (want to contribute) ask yourself why.
But maybe, maybe some companies will exclusively hire you based on your open source contributions. Honestly, if that's the case, it's more likely you're not applying for the job, and in fact the company is head hunting you.
It is true that an "open source" project is seen as a reference of work. But it's also a stale reference of work. And importantly, it's not the only reference.
Take my open source inliner project for instance. The code is appalling. There's outstanding open issues and pull requests. This is an open source project that isnot representative of my current skills. Yet it's still open source. I wouldn't put it forward as my best work.
JS Bin is probably a better example of my work, but honestly, the code is not my best foot forward. In fact, I've had job offers based on the companies associating me with the project, but I know for certain that they've not looked at the code.
Hiring decisions are mostly based on some early criteria that's utterly arbitrary. Like "do they have a degree", or "do they have a github account", or "what is their personal interest". The reason: to reduce 100 CVs down to a manageable number that you can actually interview.
When I interviewed Fabien O'Carroll in late 2013 he had limited open source projects I could look at (which I would have skimmed to get an idea of scope of projects he works on), which is fine, and he couldn't really share his company's code.
His code was decent, but that's not what got him the job. During the interview, when asked if there was anything else he was interested in, he struggled then eventually mentioned he built a chess timer (in wood) that he could use with his then-girlfriend. That's what got him the job, and that same sentiment is what I saw over the following year. That detail doesn't belong on github, isn't part of our community, but is part of an interview.
Never undervalue what you do. Everyone is unique. Some people aren't right for some jobs, but you github streak says positive and negative things.
Arguing that someone should get a jobsworth position* because the work 9-5 (or whatever extra overtime) is bollocks. It's not the first time I've seen it either. This kind of sentiment is utterly selfish and narrow minded.
* It's worth pointing out that James' tweet was between friends
The people who have a high chance of working on extra curricular projects are likely to be single, not terribly social and in their 20s. Why? Because when you're on your own and younger, you have more energy, and you need less sleep. When I was in my 20s (and married) I was hacking until 2am in the morning. What was my code like between 10pm and 2am? Shit. In fact, it's taken me years to realise that when I'm not on form (i.e. past 6pm) that it takes me twice as long, if not more, to complete a simple task. A complex task is unlikely to get fixed.
Our brains fuck with us when we deprive it of rest. Tell yourself you have RSI. If you have RSI, you rest. You move away from typing at a keyboard because if you push it, it hurts. Take that approach.
Now that I've got young kids, when I leave the office, someone will ask "what are you up to this evening?". The same as every other evening: I help put the kids to sleep, I cook (or help cook), I eat, I got to bed. I love that I'm with my family, and it's enough for me. There is no time for coding in that evening.
I've got another post that I'm writing about "what I love about the web", which is sort of related: enjoy what you do. Question where the pressure comes from. If possible, focus on what you love and ignore the noise of "hey, look at my cool thing".
There is this weird pressure to get a "name" in our industry. But fuck, seriously,you are amazing. You work hard and you should be proud of your work irrespective of whether others can see it in the open.
You come first, not your code or someone else's project: you.
For me, my family comes first. Work and code isn't even a distant second. It's taken me many years of working silly hard and silly hours to work that out. Now that I understand that, life is better.
Published at DZone with permission of Remy Sharp , DZone MVB. See the original article here.
Opinions expressed by DZone contributors are their own.