Hacker News Reader

Comment by cortesi | Hacker News Reader

Someone has to say it: Zed shaw is a cock. He doesn't get the recognition he thinks he deserves _partly_ because of his rotten public image, which he carefully crafted himself: crass, erratic, unprofessional, needy and infantile. The last damn thing a startup would want to do is to say "Here's Zed Shaw - we use his software, isn't that great?".

I say "partly", because the other half of the equation is that he doesn't deserve the reputation he thinks he deserves. He is high-profile due to controversy, but plenty of other people have done more for the Open Source community with less kudos.

View on HN

Please, avoid the ad hominem attacks. If you have something to say about the article, say it about the substance of the article, not the tone, or the character of the author. This is hacker news. http://www.paulgraham.com/disagree.html

Zed makes some interesting points about the abuse of open source software in startups. If you read some of his other writings, and don't bite on the troll/flame bait, you can see that he's been pretty upset about the lack of contract work after releasing mongrel.

I think that's a fair topic for a good rant. I'd be interested to hear the perspective of other devs of the django/rails crowd weigh in on this topic. How does your contribution to open source contribute to your bottom line?

I don't often agree with Zed wholeheartedly, but in this instance I do. I've said pretty much the same thing repeatedly on various license threads here at HN (and I work on projects under both BSD-like and GPL licenses, because the two licenses do result in very different results in the market of ideas and money).

Regardless of whether we can name lots of people who've done more (there's always somebody better, no matter what the qualifications, and they're probably better than you and me, too), Zed is absolutely right in this case, and if fame and fortune is part of his motivation for doing this stuff, he should choose licenses that are amenable to that.

A major part of the content Zed's post is his own appraisal of himself and how wonderful he is and how he never gets any attention for what he does. When he makes it the topic himself it's a perfectly valid to discuss whether it's true or not.

IMHO his attitude and public persona absolutely may have to do with what he experiences wrt his software. He's obviously brilliant and does great code, but I know I wouldn't hire him as even a temporary contractor because his personality is just incompatible with the needs of a sane business environment.

I'm not sure that you have an authority to make such judgments. The way he is on his personal website may have nothing to do with his professional behavior. We are wired in, so we see his rants and internalize them, but that doesn't mean that they present an accurate picture of reality.

I had a heck of a time trying to get my company to use one of his excellent libraries. Why? Because they saw Zed's "professional" home page for the library and documentation and didn't want to be associated with that kind of persona. "It's all about the image," they told me. I disagreed with them, but there you have it. None of these folks were "wired in".

If it's all about the image they shouldn't be using half of the libraries out there. Most of them have scant documentation, ugly websites, hard-to-use bug tracking software, or some other problem.

By the way, who exactly is "they"? If it's management, they should know that you have to trust knowledge workers to know and do their jobs. Management != programmers.

Your company makes worse decisions for choosing one library over another for worse reasons than mine does.

I think a lot of his points have merit.

Meebo is a multimillion dollar company (est $200m), built mainly off libpurple. I really don't understand how the libpurple developers can look at that situation and not feel hard done by. Tireless work reverse engineering protocols, which is then taken and silently used without real credit, to build multimillion dollar companies :/

I had a discussion about this with some hard core free software people, and they just said they thought libpurple would feel "proud". Well I'm sorry, but if someone was making millions from work I'd done, I wouldn't feed proud in the least. I'd feel like an idiot.

(I'm not saying meebo is bad to do this at all. I'm more saying that I don't get how anyone could be motivated to work on an open source project like libpurple without some protection in place).

Meebo employs libpurple developers. The libpurple team looks to have specifically OK'd Meebo's use of libpurple. I don't think they feel "hard done by".

Maybe it was harsh of me to use meebo/libpurple as an example here in that case. Hopefully they're getting a good chunk of the VC money etc.

Here's where you point out that you compete with Meebo.

Not really, Meebo don't really have a large groupchat offering, and Mibbit don't really have much of an IM offering :)

point taken though... As I said at the start, I wasn't beating on Meebo, I think they've done a great job.

I'm sure there's other examples, but Meebo was the one that sprang to mind.

Whenever I hear about some open source volunteers who "did the hard part" and then someone else came along, "slapped a GUI on it" and made big bucks, I wonder why the original developers didn't bother to do that part. Maybe the real value isn't where a lot of people think it is.

After reading all these stories, I conclude that if you build only infrastructure you are going to get nothing back. Maybe the DHH approach is better: Build an app and then factor some infrastructure out of it later.

if someone was making millions from work I'd done, I wouldn't feed proud in the least. I'd feel like an idiot.

You should probably clarify. I bet if you were getting a kickback you wouldn't care if someone else were making money. Like if you were a writer or musician and your agent were making a healthy cut, you wouldn't care if you were getting paid too, right?

So it really comes down to being jealous other people figured out how to make money from your stuff when you couldn't.. right? That skill has value and perhaps they deserve to be making the cash.

I like how you're subtly calling the libpurple developers idiots.

Some people can live with making enough money to have a pretty good life and not worrying about maximizing their revenue streams while just being happy that their work has meaning to others.

:) That wasn't my intention. I said I would feel like an idiot. Everyone has different viewpoints, and maybe they're happy to donate their work both to end users, and to multimillion dollar companies. The issue I see is that in the case of Meebo for example, a large part of that $200m worth undoubtedly comes from their ability to work with IM protocols - which has been largely 'donated' to them, from open source folks.

(Again, I'm not saying Meebo is bad for doing this, I'd do exactly the same).

Lets suppose you reverse engineered skype. Would you open source it? Or seek to make at least a little money from it. If you chose to open source, it's likely many others would make money from it instead of you.

How on earth would I make money from reverse engineered Skype? I can't even fathom it, the amount of work it would take TO make money from a reverse engineered protocol for communicating with Skype servers. I would not even know where to start.

Here's what I think: the $200MM going to Meebo has NOTHING to do with them using the libpurple library. My reasoning: Any other web based chat company would have the same capability to exploit libpurple in their strategy. Web companies don't get funding for their technology ever. They get funding for their eyeballs and stickiness.

Finally, on the subject of backhanded insults: Imagine that you were physically speaking with a group of libpurple developers and said, "Boy, if I were in your shoes, I'd feel like a bunch of idiots!"

>> "the $200MM going to Meebo has NOTHING to do with them using the libpurple library."

Without the libpurple library, Meebo may not have existed. Or at least would have been far far harder to execute afaics. libpurple isn't just another framework, or webserver. It's very valuable reverse engineering effort.

As I said, [I] would feel like a complete idiot, if I worked my ass off to create something, only to see a company exploit+monetize it. That is not in any way the same as calling libpurple developers idiots.

>> "How on earth would I make money from reverse engineered Skype? I can't even fathom it, the amount of work it would take TO make money from a reverse engineered protocol for communicating with Skype servers. I would not even know where to start."

Seriously?? There are many many ways you could monetize the ability to communicate over the skype protocol. You could build cool plugin services. You could build something to make peoples voices sound like robots - disguizr.com or b2b stuff, voicemail, you could build your own skypeout competitor. etc etc etc you could build conferencing services ontop of it. Prank/joke call things. Voice recognition bot services. I'd say at least one of those would work. If you sat down and thought, I'm sure anyone here could come up with some killer applications.

So I take it that you haven't heard of Asterisk for Skype? You can already do some of those things (at least those that don't require rewriting your CLI)

Other people said it, but you said it better: "his rotten public image, which he carefully crafted himself: crass, erratic, unprofessional, needy and infantile."

But yeah, there are a lot of people who are little-known but who have written a lot of the software we use every day, and even successfully built communities around it. Ian Lance Taylor, Rob Savoye, tom lane, Lou Montulli (!), Jacob Kaplan-Moss, Raymond Hettinger, Margo Seltzer, D. Richard Hipp, Ben Tilly, a cast of thousands.

Thanks for the list. And for people like me who don't know much about (some of) them, here is a primer:

  Ian Lance Taylor - gcc maintainer, creator of GNU UUCP
  Rob Savoye - primary developer of Gnash
  Tom Lane - member of the core PostgreSQL development team,  standardization of JPEG
  Lou Montulli - co-authored Lynx, came up with HTTP Cookies and HTTP Proxying
  Jacob Kaplan-Moss - Django maintainer
  Raymond Hettinger - core developer for Python
  Margo Seltzer - Berkley DB architect
  D. Richard Hipp - architect and primary author of SQLite
  Ben Tilly - developed a bunch of Perl modules

Edit: Added standardization of JPEG to Tom Lane's intro.