3 ms·

> Open source is different I believe the attributes given to open source here are wrong. There is a ton of open source code that is crap where not even the au

by ketzu 6y ago

> Open source is different I believe the attributes given to open source here are wrong. There is a ton of open source code that is crap where not even the author had any interest in making it actually good (the same is true for writing). There is code that is the result of what's alluded to be "copywriting work" that's open source. For me, this reads like a strawman, or maybe super idealised version, of open source is like compared to a vague strawman of what working in a company is like. It feels like this is neither true nor are those the only options out there. > So if a user of your software is addressing you because some part of your code sucks, and is willing to work with you to do something about it, and is very demanding, don’t think they are abusing you because they are not paying you. I don't feel the users that are called "abusive" to maintainers and contributers are the ones because some of your code sucks and they want to make it better. It's the ones that want you to pander to their use-case. That have unreasonable demands. But even the ones with bug reports and telling you your code sucks in some way, should have some basic civility while communicating. No power of refusing pull requests shields you from insults and the mental burden of being harassed because you refuse pull requests (or close an issue someone opened). I personally expect other people to value my time, not necessarily in money though. This is a general rule of life, not a matter of open source.

franciscop 6y ago
> There is a ton of open source code that is crap The good thing of open source is that people pick the best available, or at least some that is "good enough" when there are many choices. The bad thing of proprietary code is that you paid for that, so you are stuck with it. Throwing it again and redoing it has a high cost, staying with it has a (high?) cost. But if you want custom code or very specific code, it's going to be one-off. So yes, in general you can say that open source has better quality, because the best rise to the top, while in proprietary code it is what it is.
- ketzu 6y ago
  I believe that you could easily turn around the argument: To stay relevant with paid software you have to make something people will be willing to buy. If you produce bad software, people will not or at least stop buying it. You can dump everything into the open source domain without any restrictions. Therefore, commercial code is of better quality. It's not that I believe this, I find neither of those convincing. There is great open source software and there is great commercial software. There is a huge amount of bad software on both seids either. Is good open source software better than good proprietary software? Depends on what are the criteria for good, but I don't think that's something we could state as general as "open source is better." Going back to the original article: The distinction the article talks about seems to be personal motivation to make the code better. While there might be a shift to be more engaged in a personal project (which is likely) the licensing model of that personal project is fairly irrelevant for this. Futher, quite a lot of people are also engaged in their job and care about the product they are working on. Are people that don't care about it on their job more or less likely to even start personal projects? To contribute to open source? I don't know, maybe there is some data on that.
  - franciscop 6y ago
    > To stay relevant with paid software you have to make something people will be willing to buy. This doesn't mean it has to be good, just that you convince people to pay you, which is IMHO "easy" with the high software developer demand right now and difficulty for non-devs to distinguish between high and low quality software. I think though, that as you are saying, open source shines with generic libraries where there's competition and selection of the best, while proprietary software shines with end-products (which are normally a multi-discipline team effort).
- mceachen 6y ago
  > The good thing of open source is that people pick the best available At least at the fringes, popularity can be anticorrelated with quality. I can't count the number of times I've looked for a library to do some task, found an extremely popular repo with thousands of stars, tried it and had it fail in an obvious way, find that the bug was reported already (and ignored or closed or a pr was opened and ignored). I then try some alternative, built by an unknown author, and find that it has reasonable tests (!) and good docs (!) and actually works. 12 GitHub stars.
  - franciscop 6y ago
    That is exactly what I mean. So you used the best available one! And so will many people. The case you are describing normally happens when the previous one has been there for a while, and the new one is, well, new. Many people will do like you, and find the new one more useful, and add their stars, and write articles, so over time it'll also become popular.
    - mceachen 6y ago
      If this were true, the better libraries would have more stars, especially when the age of the libraries are similar.
      - franciscop 6y ago
        And this is generally true; for similarly age libraries, the one with better quality (accounting for docs as well!) is normally more popular. There are exceptions for sure, as recently popular dev's libraries become more popular because of how popular their author is, but I'd say in libraries that are 3-5+ years old this generally holds true

by ketzu 6y ago

franciscop 6y ago

> There is a ton of open source code that is crap The good thing of open source is that people pick the best available, or at least some that is "good enough" when there are many choices. The bad thing of proprietary code is that you paid for that, so you are stuck with it. Throwing it again and redoing it has a high cost, staying with it has a (high?) cost. But if you want custom code or very specific code, it's going to be one-off. So yes, in general you can say that open source has better quality, because the best rise to the top, while in proprietary code it is what it is.

ketzu 6y ago
I believe that you could easily turn around the argument: To stay relevant with paid software you have to make something people will be willing to buy. If you produce bad software, people will not or at least stop buying it. You can dump everything into the open source domain without any restrictions. Therefore, commercial code is of better quality. It's not that I believe this, I find neither of those convincing. There is great open source software and there is great commercial software. There is a huge amount of bad software on both seids either. Is good open source software better than good proprietary software? Depends on what are the criteria for good, but I don't think that's something we could state as general as "open source is better." Going back to the original article: The distinction the article talks about seems to be personal motivation to make the code better. While there might be a shift to be more engaged in a personal project (which is likely) the licensing model of that personal project is fairly irrelevant for this. Futher, quite a lot of people are also engaged in their job and care about the product they are working on. Are people that don't care about it on their job more or less likely to even start personal projects? To contribute to open source? I don't know, maybe there is some data on that.
- franciscop 6y ago
  > To stay relevant with paid software you have to make something people will be willing to buy. This doesn't mean it has to be good, just that you convince people to pay you, which is IMHO "easy" with the high software developer demand right now and difficulty for non-devs to distinguish between high and low quality software. I think though, that as you are saying, open source shines with generic libraries where there's competition and selection of the best, while proprietary software shines with end-products (which are normally a multi-discipline team effort).
mceachen 6y ago
> The good thing of open source is that people pick the best available At least at the fringes, popularity can be anticorrelated with quality. I can't count the number of times I've looked for a library to do some task, found an extremely popular repo with thousands of stars, tried it and had it fail in an obvious way, find that the bug was reported already (and ignored or closed or a pr was opened and ignored). I then try some alternative, built by an unknown author, and find that it has reasonable tests (!) and good docs (!) and actually works. 12 GitHub stars.
- franciscop 6y ago
  That is exactly what I mean. So you used the best available one! And so will many people. The case you are describing normally happens when the previous one has been there for a while, and the new one is, well, new. Many people will do like you, and find the new one more useful, and add their stars, and write articles, so over time it'll also become popular.
  - mceachen 6y ago
    If this were true, the better libraries would have more stars, especially when the age of the libraries are similar.
    - franciscop 6y ago
      And this is generally true; for similarly age libraries, the one with better quality (accounting for docs as well!) is normally more popular. There are exceptions for sure, as recently popular dev's libraries become more popular because of how popular their author is, but I'd say in libraries that are 3-5+ years old this generally holds true