5 ms·

"I might make myself look a bit messy by releasing this. I also imagine that some other people might have to put in some amount of work (much less than the work

by gaze 13y ago

"I might make myself look a bit messy by releasing this. I also imagine that some other people might have to put in some amount of work (much less than the work I put into writing this in the first place) to get this to work, so I'll make them rewrite it all from scratch if they are in need of this functionality." I don't believe this person understands empathy. I realize if you want your project to flourish as an open source project, you need to make it palatable. However, a tarball of shit is better than NOTHING. This person isn't even deluding themselves by saying they'll "get around to open sourcing it." You won't do the world a disservice by releasing working code. Seriously, just come out and be honest and say "I don't want to open source it because I don't wanna."

pnathan 13y ago
> However, a tarball of shit is better than NOTHING. bzzzt! Wrong! Releasing stuff that is "close enough" as a product and have it succeed is wasting other people's lives as they struggle to fix your ball of mud and make it work better over time. Better to wait and not waste other people's time.
- daleharvey 13y ago
  Just in case anyone who is thinking about releasing open source softare reads this, it is entirely wrong. Software (open or otherwise) is never done, its always too ugly, missing some features, slow and / or buggy. If you are thinking about releasing your stuff as open source, do it now, it probably is too early and nobody will use it, someone may pick up it or your next library, share some knowledge report some bugs or evern fix stuff. It is very easy for software developers to verify the stability of the libraries they are using, you have absolutely no responsibility to them by releasing software that has bugs, it all does.
  - pnathan 13y ago
    Baloney! Releasing dysfunctional crap is giving your users crap that they have to do the grunt work to maintain. It's entirely one thing to release functional and ugly code, it's entirely another to simply code dump broken crap and think that's a good thing.
    - daleharvey 13y ago
      People spending all their time fixing broken libraries that they would have been quicker writing themselves is a problem that pretty much never happens, there are multiple ways in which to determine if a library is likely stable or mature enough to be used, and testing whether it works or not is usually a fraction of the time that it would take to write the thing from scratch. If you open source some code without making any grandiose promises and someone somehow ends up in the position where they are wasting their time fixing it, its entirely their own decision / fault.
bradleyland 13y ago
I think reality is somewhere in between. Spend some time browsing the issue tracker on a good OSS project. Then, if you're feeling brave, head over to a bad one. I don't think any of this negates your argument, but I understand the apprehension involved in open sourcing a project one feels is sub-par.
otterley 13y ago
> a tarball of shit is better than NOTHING Not necessarily. A tarball of shit can become the de facto standard for some utility, and end up haunting us for ages after it gains critical mass, even long after a superior replacement arrives. See, e.g., rvm.
thaumaturgy 13y ago
I have to disagree here unfortunately. Like rachel, I've developed my own methods and approaches and toolchains for projects. The upside is that it all works exactly the way I like it to; the downside is that nobody else wants to touch any of it, and anytime somebody else gets a whiff of it, I expect a lot of "why aren't you just using X, Y, or Z?", "this is really shit", "you're a terrible programmer", "lol can you even code?", "look at this dude suffering NIH", and so on. And ... I've got a pretty thick skin after years and years of net time, but I already struggle with staying motivated on projects, and that kind of feedback really doesn't help. So, for the most part, the whole "social programming" thing is just not for me.

by gaze 13y ago

pnathan 13y ago

> However, a tarball of shit is better than NOTHING. bzzzt! Wrong! Releasing stuff that is "close enough" as a product and have it succeed is wasting other people's lives as they struggle to fix your ball of mud and make it work better over time. Better to wait and not waste other people's time.

daleharvey 13y ago
Just in case anyone who is thinking about releasing open source softare reads this, it is entirely wrong. Software (open or otherwise) is never done, its always too ugly, missing some features, slow and / or buggy. If you are thinking about releasing your stuff as open source, do it now, it probably is too early and nobody will use it, someone may pick up it or your next library, share some knowledge report some bugs or evern fix stuff. It is very easy for software developers to verify the stability of the libraries they are using, you have absolutely no responsibility to them by releasing software that has bugs, it all does.
- pnathan 13y ago
  Baloney! Releasing dysfunctional crap is giving your users crap that they have to do the grunt work to maintain. It's entirely one thing to release functional and ugly code, it's entirely another to simply code dump broken crap and think that's a good thing.
  - daleharvey 13y ago
    People spending all their time fixing broken libraries that they would have been quicker writing themselves is a problem that pretty much never happens, there are multiple ways in which to determine if a library is likely stable or mature enough to be used, and testing whether it works or not is usually a fraction of the time that it would take to write the thing from scratch. If you open source some code without making any grandiose promises and someone somehow ends up in the position where they are wasting their time fixing it, its entirely their own decision / fault.

bradleyland 13y ago

I think reality is somewhere in between. Spend some time browsing the issue tracker on a good OSS project. Then, if you're feeling brave, head over to a bad one. I don't think any of this negates your argument, but I understand the apprehension involved in open sourcing a project one feels is sub-par.

otterley 13y ago

> a tarball of shit is better than NOTHING Not necessarily. A tarball of shit can become the de facto standard for some utility, and end up haunting us for ages after it gains critical mass, even long after a superior replacement arrives. See, e.g., rvm.

thaumaturgy 13y ago

I have to disagree here unfortunately. Like rachel, I've developed my own methods and approaches and toolchains for projects. The upside is that it all works exactly the way I like it to; the downside is that nobody else wants to touch any of it, and anytime somebody else gets a whiff of it, I expect a lot of "why aren't you just using X, Y, or Z?", "this is really shit", "you're a terrible programmer", "lol can you even code?", "look at this dude suffering NIH", and so on. And ... I've got a pretty thick skin after years and years of net time, but I already struggle with staying motivated on projects, and that kind of feedback really doesn't help. So, for the most part, the whole "social programming" thing is just not for me.