This is the blog of Adam Kalsey. Unusual depth and complexity. Rich, full body with a hint of nutty earthiness.
Freshness Warning
This blog post is over 20 years old. It's possible that the information you read below isn't current and the links no longer work.
3 Jul 2003
Phillip Harrington asks…
How do you feel about this NEcho stuff? I’m kind of annoyed by the “we need a new spec and it has to be everything to all people and it’s a democracy etc” feeling of the whole thing. I feel like we’ll get 0.1 alpha some time in 2005.
This isn’t the first time something like this has happened. Someone tries to start a new spec and involve everyone else. They make sure to include everything including the kitchen sink. Eventually it gets bogged down, people lose interest, and the project dies. Then someone steps in and shows leadership. They publish a spec and some working code. Other people adopt it. It becomes a de facto standard.
That’s how the Metaweblog API came about. There were various propsals and discussions about replacing the Blogger API with something robust. No one could agree on all the features. The people who would actually be responsible for implementing this stuff lost interest in the project and left, leaving the idealists to their endless planning.
Then Dave Winer stepped in and implemented something. And it worked. And other people adopted it. Was it perfect? No, and I think Dave would be among the first to admit that.
Regardless of what you may think of Dave personally, he gets things done. They may not be done “the right way,” they may not be perfect, but they exist and they work. Which is more than you can say for most other projects.
The funny thing is, from stuff Dave has said in the past, I think he’d be all for improvements to syndication and API formats. He knows that the formats aren’t perfect and there are lot of smart people out there that could improve on his ideas. But they should be simple, and they shouldn’t use names that he’s come to think of as his own. Don’t call it RSS unless you get his blessing on the format. Fine, I can live with that. He’s done more to advance RSS as a simple way to create syndicated feeds than anyone else so respect his wishes.
This project may succeed where others have failed because it has someone working on it full-time. It’s Sam Ruby’s job to figure this out. IBM is paying him to do it. Sam’s obviously a smart guy, and smarts coupled with financial backing is a powerful thing. But my advice would be for Sam to show some leadership. He’s gotten everyone’s feedback and ideas on what the project should do. Now he should close the discussion, develop a draft spec, and put it out for comment.
There will be people that disagree with the spec. There will be those who want to add this and that to it and turn it into a behemoth. Sam should stick to his principles and only make changes where they make sense. Changes should be the result of compelling arguments, not the result of conflict avoidance.
Being a leader is hard, as both Sam and Dave know. Not everyone will like you if you stand up for your principles. But people don’t lead because they want to be popular.
First of all, it's not like I'm a Dave Winer cheerleader. I don't read Scripting News. I even agree with Bill Kearney sometimes. So don't mistake me for a member of the Winer fan club. I simply see that there's a huge difference between the way things actually get done and the way that !Echo is going. "why did he criticise Blogger when they decided it didn't meet their needs and improved their own API" If you read past the petty snipes at Evan, you'll notice that his basic reason for disagreement is that he wants compatibility. A whole new format simply breaks tools. It makes developers work too hard to implement multiple things. "He puts a massive amount of 'stop energy' into competing formats and namespaced extensions." Look at it from the user's perspective. What if HTML had competing formats every few years? Look at the nightmare that was DHTML development circa 1997-99 when Netscape 4 and IE 4 had completely different mechanisms for doing layers. Has anyone actually asked Dave what he would support? Why he closed the RSS spec at 2.0? I think that was the wrong thing to do, but it's silly to go off creating another format because you don't like the way the existing one is managed. I don't like SMTP, but it would be pointless to try to replace it. It's too engrained. "When you read Sam's writings on the matter, it's overwhelmingly clear that he is not taking on the role as leader." No kidding. And that's what this project needs. A leader. That's my point exactly. "Sam is constantly saying to people 'here's an idea, run with it'. People do. Sam chips in with his own ideas again. The loop continues." And continues, and continues, and continues. We don't need a feedback loop. We need a spec. This is nothing against Sam. Taking a leadership role is a hard, thankless job. It's a lot different from being a brilliant technical architect. It's certainly not something I want to do in a case like this, and I applaud Sam for stepping up.
Excerpt: Now I never claim to be the smartest person in the world. I never even claim to be intelligent. I do claim however to have some common sense. If you haven't been living under a shell you are well aware...
I'm sorry, I don't think you've made a very good case for why this project needs a leader. What would a leader bring that the project hasn't already got? How do explain the success of all the open source projects that thrive *without* any clear leader (all the Apache stuff, Mozilla, most of Linux etc etc). What proportion of the W3C or OASIS specs had a clear leader? Dave's 'leadership' has led to a situation where the community has to move en masse away from existing specs to try and make progress. A very good reason against having a leader, IMHO. What's needed is something the syndication community can agree on, even if it is a behemoth (it's not looking too fat right now, btw). Anyone putting themselves forward with any kind of force will immediately put off parts of that community and we'll be back to square one. The hands-off, gentle guidance that Sam's applying is as perfect as it could get. I personally think it's got a very good chance of working. We'll soon find out ;-)
If progress is so badly needed, and if it's so obvious where to go, why is Necho such a bloody mess?
A leader would bring progress and cohesion. The project is mired down in people discussing ad nauseum what the project should be and how it should work. One the the first acts of this committee was to introduce scope creep. Instead of a replacement for RSS, it's now a replacement for RSS and weblog APIs. Other large open source projects have had leaders (although that's comparing apples to oranges). Apache had Brian Behlendorf, Linux had Linus Torvalds for the kernel and others for various other aspects, and Mozilla had Jamie Zawinski (jwz). Zawinski mentioned in his resignation letter than the group couldn't figure out how to get things done. He considered himself, as the leader, to be at fault for this, " In my humble but correct opinion, we should have shipped Netscape Navigator 5.0 no later than six months after the source code was released. But we (the mozilla.org group) couldn't figure out a way to make that happen. I accept my share of responsibility for this, and consider this a personal failure." But even those comparisons are moot. !Echo/Pie/Whatever isn't a piece of software. It's a concept, a general thought about how technology should work. Look at the history of this. You bring up the W3C and OASIS. The W3C delivers specs largely after companies have innovated and presented them with fait accompli. DHTML and the DOM is nearly identical to what Microsoft created with IE4. And XHTML took how long to create? And quick, name a project by OASIS. UDDI? That was mostly IBM's work. Why did the syndication community turn against RSS? Mainly because they felt it needed more. More this, more that, more complexity. Do feed developers consumers want more complexity? It seems that it's mainly the tools developers who want these added features. The people actually using and creating feeds could care less. The status quo works fine for them, which is exactly why the spec was frozen. You may not think the spec looks to fat, but the sample feeds created by Mark Pilgrim and others are almost as large as their main page content. Mark's front page currently weighs in at 7.26kb. His RSS feed is 0.56kb. But the Echo feed is 4.05kb, more than 7200% larger than the RSS feed. As far as the average feed consumer is concerned, the echo feed has roughly the same data as the RSS feed, but is seven times the size. This is better how? And let's not forget that this feed is based on a draft spec that will change. So to keep current, Mark will have to constantly update the feed format to keep pace with a changing spec. I'm sorry, but I don't see any progress here.
>>It seems that it's mainly the tools developers who want these added features. Check again. Most of the tools developers have argued in favor of RSS 2.0 and the MetaWeblog API. None of the three leaders, Pilgrim, Ruby or Bray are tools developers. Haven't seen many posts from the MT or Blogger folk. I've posted a lot more than they have.
When I said tools developer I meant those who are building things on top of RSS and APIs. Six Apart is one. I'd imagine Sam Ruby has something cooking, otherwise IBM wouldn't be paying him to do this work. But you're right, the tools that connect to the APIs and read the RSS largely haven't been involved, other than to say "when something gets figured out, I'll probably support it." Feedster, w.Bloggar, Zempt, SharpReader, NewsMonster, and FeedDemon all have some significant stake in where this goes, but I haven't seen any real participation or interest from the developers. (Well, Zempt is me and Bill, but still...) What is really being developed is BlogML. A bi-directional markup language for weblog conversations. Except that they want it to extend beyond weblogs so banks and stuff can use it. And wouldn't it be nice if you could have the weather forecast in BlogML format. And my voicemail should show up there too. The fridge should speak blogML too, so my aggregator could tell me when I'm out of milk. And wouldn't it be cool if....
This discussion has been closed.
Jim
July 3, 2003 6:39 PM
I disagree with a couple of points. "Was [the metaweblog API] perfect? No, and I think Dave would be among the first to admit that." Then why did he criticise Blogger when they decided it didn't meet their needs and improved their own API? It's all very well paying lip service to your own inevitable shortcomings, but Dave doesn't act like he believes it. "The funny thing is, from stuff Dave has said in the past, I think he’d be all for improvements to syndication and API formats." The RSS 2.0 specification, which he authored, specifically states that it is the end of the line as far as RSS is concerned. He puts a massive amount of "stop energy" into competing formats and namespaced extensions. I just can't agree with what you are saying here. "This project may succeed where others have failed because it has someone working on it full-time. It’s Sam Ruby’s job to figure this out." I agree and disagree. Yes, Sam can get the job done. No, it's not because he's working on it full time. When you read Sam's writings on the matter, it's overwhelmingly clear that he is not taking on the role as leader. The impression he gives is that he thinks of himself as a contributor. He acts that way as well. He may end up as the largest contributor, perhaps he will be _seen_ as the leader, but he doesn't act like he *owns* !Echo. The result? He is an _enabler_. From my point of view, Sam is constantly saying to people "here's an idea, run with it". People do. Sam chips in with his own ideas again. The loop continues. I can't express how much healthier I consider this to be compared with the state RSS has gotten into. Even if Dave suddenly turns into Mr. Perfect, and a completely clear specification arises from the ashes of RSS 2.0, I'd still consider RSS to be unsalvagable, because a hell of a lot of people who matter are just plain sick of RSS. +1 to !Echo.