Open Source Obligations(red-sweater.com)
red-sweater.com
Open Source Obligations
http://www.red-sweater.com/blog/1078/open-source-obligations
8 comments
It strikes me that perhaps Mr. Gruber can't win here--either he's too lax as a project parent, or he rules the project with an iron fist, and gets complaints because he's not flexible enough.
Being involved in an Open Source project is expected for anyone that wants other people to use the code - implicit in publishing it. It doesn't imply working for free when asked, but it does require responding to email even if it's just to say 'sorry, I'm really busy right now'.
If he rules the project with an iron fist, that's still much better. People know if the project is alive or not, and can make the decision to fork the project easily.
If he rules the project with an iron fist, that's still much better. People know if the project is alive or not, and can make the decision to fork the project easily.
Cute. But it misses an important point in Jeff Atwood's original post, which was that John Gruber got actively hostile when someone else extended his Markdown code in a way he didn't approve of. If nothing else that was in poor taste.
When you put that quote in it's proper context, it doesn't sound 'hostile' at all: http://news.ycombinator.com/item?id=1020864
That's not completely true. Atwood misrepresented Gruber. Read this part of the email thread:
http://six.pairlist.net/pipermail/markdown-discuss/2008-Marc...
Gruber states that he doesn't have time update Markdown and forks should be clearly marked as such.
Gruber states that he doesn't have time update Markdown and forks should be clearly marked as such.
I think the problem with markdown is that it's too complex to write a good parser in a few lines of code, otherwise there would be no "open source obligations" issues at all, because what is really needed in a markup system is that it is widely accepted, all the rest will follow, unless the format is so complex to implement well that most of the implementations will be hacks.
I also don't like how it looks from the human point of view, but this is just my opinion and I guess it's not very common as my feeling is that most people love markdown.
Given that this is the standard de facto at this point I guess that what's really needed is to write an high quality, fast, simple, reference implementation written in an algol-like language (guess what? The good Haskell implementation does not qualify), and share it under a BSD license so that it will be both the specification and the reference implementation everybody will use to derivate their Python / Ruby / C / ... solid implementations of markdown.
This is the perfect project to emerge in the coding community with little efforts if you are smart. It's a few days of work and it is very useful to many people, so what are you waiting for? :) Get some coffee and start coding now instead to waste all your time on hacker news! Then setup a page for your code and write a blog article explaining what are the hard part and tradeoffs about implementing markdown, and possibly how markdown could be enhanced. Don't worry, most of us understand that what you produced is FREE and will just praise you, without obligations ;)
I also don't like how it looks from the human point of view, but this is just my opinion and I guess it's not very common as my feeling is that most people love markdown.
Given that this is the standard de facto at this point I guess that what's really needed is to write an high quality, fast, simple, reference implementation written in an algol-like language (guess what? The good Haskell implementation does not qualify), and share it under a BSD license so that it will be both the specification and the reference implementation everybody will use to derivate their Python / Ruby / C / ... solid implementations of markdown.
This is the perfect project to emerge in the coding community with little efforts if you are smart. It's a few days of work and it is very useful to many people, so what are you waiting for? :) Get some coffee and start coding now instead to waste all your time on hacker news! Then setup a page for your code and write a blog article explaining what are the hard part and tradeoffs about implementing markdown, and possibly how markdown could be enhanced. Don't worry, most of us understand that what you produced is FREE and will just praise you, without obligations ;)
Actually, Discount (http://www.pell.portland.or.us/~orc/Code/discount/) has been around for a while, and is rather high quality. It's a C library, but has a very simple API and should be easy to wrap in any language with a halfway decent FFI.
Oh, and BSD license? Yup.
Oh, and BSD license? Yup.
Seems to me as an ignoramus to these goings on that part of the impediment to progress is ambiguously named forks of the spec. For instance MarkdownSharp on the face of it isn't just a reimplementation of Markdown in C#, it's a fork of it. As someone who doesn't know C# were I to do an implementation of Markdown I'm not likely to look for improvements in something called MarkdownSharp.
Has Jeff Atwood contributed any open source code? Given that he's criticized Gruber for not giving enough, I'm curious as to his own contributions (on Github, his codinghorror account contains no repos or activities: most likely a placeholder to prevent name squatting).
Completely aside from this flamequest: I am sure at some point OS authors will start putting some promises about how they'll conduct their projects. Not because it's their duty - but because it can make the code more useful.
The problem with forking is that you can end up with a situation similar to the UNIX Wars of the eighties and early nineties with multiple, not-quite-interfunctional versions. That almost killed UNIX and could more easily do it to a minor program like Markdown.
I think the moral of the story is, "Don't feed the troll". I'm not sure if Jeff is deliberately provocative in his posts, but many of them do turn out that way, and you can bet he gets a lot of ad revenue from the publicity.