Mushrooms
feedvalidator.org feeds.archive.org feedvalidator.sourceforge.net atomenabled.org rss.scripting.com
And, coming soon, there will be one on localfeeds.com
Apparently, there is some controversy over scripting.com hosting the feedvalidator. From my perspective, the license clearly permits this usage. Futhermore, I encourage it. And if Dave or Andrew want to submit patches to provide an option to disable validation of Atom and/or RSS 1.0 on their installation, then I will see what I can do to integrate such patches into the system.
Meanwhile, the validator got its first bug report on sourceforge. The validator currently issues errors when it encounters relative urls is RSS feeds. Unless I hear any objections, I'll remove all such checks, messages, and test cases pending resolution of this issue.
I've submitted requests to create a CVS repository and mailing list on sourceforge. I'm told that these should be available shortly.
Although allowed by the license, I can not think of a single reason why this is a good thing. I can not tell you how many validator bug reports I've received that start out "your validator doesn't like my feed, but Userland's validator says it's fine". The situation was so ongoing and annoying that I wrote up a FAQ on it and I just point people there.
The next time you deploy new code on feedvalidator.org that is more strict than previous iterations, you're going to get bug reports "your validator doesn't like my feed, but three other validators say it's fine". How is this an improvement? Whom does this help?
Then there's the question of feed type and version support. Dave says his validator is "primarily to promote RSS, not other formats". Does this mean he is going to drop Atom support (his validator currently claims that Atom feeds are "valid RSS", and has alternately flamed and ignored everyone who has reported this bug to him). His validator currently validates RSS 1.0 feeds; will he drop support for those? His validator currently validates "funky RSS" feeds (defined by him as feeds that use Dublin Core); will he drop support for those? Is he actually adding anything to this marketplace (other than confusion)?
The same questions should be asked of the validator on atomenabled.org. It says "enter an URL of your Atom feed", but it validates RSS just like feedvalidator.org. (At least it labels them correctly.) Will atomenabled.org's validator drop support for RSS altogether?
Posted by Mark atSam, It would be nice if feed validators supported compression or at least detect and report otherwise.
Posted by Don Park at
Mark, these are all questions for you and Sam to answer. Why did you decide to turn the RSS Validator into a Feed Validator? These were your decisions, not anyone else's. Perhaps you should publish some guidance on this.
Sam, I hope your policy of crossing out ad hominems still applies here.
Posted by Dave Winer atMore validity checking has to be a good thing, but Mark's got a mighty good point. Perhaps the validator could indicate its own version?
fyi, the scripting.com install now gives an Atom button when that kind of feed is checked.
It seems there's been a change of strategy :
"Now at least [Bloggers] totally incompatible feed has a different, not confusing name."
"Some people felt that calling it an RSS feed was confusing. I'm pretty sure users will end up thinking of Atom as a flavor of RSS."
Gentlemen, we have a new Borg.
Posted by Danny atre: "Why did you decide to turn the RSS Validator into a Feed Validator?"
Um, because it validates feed formats that aren't called "RSS".
Posted by Mark atTo be clear, the (real) Feed Validator still validates feed formats that are called "RSS". All 7 versions of RSS, as a matter of fact.
You still haven't answered my questions. Should I try another pseudonym?
Posted by Mark atThis "different validator problem" is no problem at all. People have multiple choices of validators for HTML and they are quite able to differentiate between them and for example take the w3c validator as a reference, even if other validators validate differently. All is needed is to make yourself known.
Of course Userland will allways be seen as a reference by Radio users and others, as Userland is referenced in the RSS world all over the place. If you want your validator to take the default role, make it known and make your self known.
There are different strategies to do this. One would be showing competence. This is usually shown by giving feedback to problems, answering questions and documenting problems, building up resources to link to.
Another would be flaming people to death for nothing. This usually makes one known, too, but usually people don't take any competence for granted for those people. It might work out fine (at least it did so for Theo de Raadt), but it might backfire and destroy any competence and credebility people grant you.
The third way is by building some institution around a technical theme. This usually helps by hooking the competence up with the institution and hooking up the flaming and social incompetence with the persons. Or to make it more clear: the institution is what provides standards and documentation and reference, the people themselves can happily flame away.
Maybe it's time to build a Syndication Institution that is based on competent people both of RSS and Atom fame, so the advocates can continue with their stupid little pissing contest and developers can go on to use whatever technology suits their needs ...
Mark, you haven't been willing to accept my answers so far, that's why you think I haven't answered them.
Basically, for the nth time, we're just running your software. Nothing more to say. I don't know what the future holds in re this validator, you probably have a better idea than I do.
Sam, I think you're letting some flames through, so until this is moderated, I'll see y'all later.
Posted by Dave Winer atAlthough allowed by the license, I can not think of a single reason why this is a good thing.
First, I see this a form of bringing the validator to the users. The first bug report is an example of another user becoming aware of validation. I see this as a success.
And, yes, having this validator deployed on other sites does create versioning issues. As does having it installed with client programs like newsreaders. This is clearly something that we have to deal with.
A second benefit is in increased quality of patches. This is a lesson I have learned in open source - my first contributions to PHP were in creating a set of Windows build instructions; early on in Jakarta, I became the first release manager for Ant. The lesson I learned is this: until people can build and deploy a software product for themselves, they can only guess as to what a correct fix might be. Giving them the opportunity to directly make fixes on an environment that they control creates the potential for growing a development community.
I may be biased, but I do believe that the feed validator that we wrote is the best one out there. This being said, it still is not as good as it could be, primarily because you and I have other demands on our time.
Posted by Sam Ruby atThe example of NetNewsWire embedding the feed validator is an excellent example, because it clearly illustrates the difference between an constructive use of our code and a competitive and malicious use of our code. What Brent is a completely new venue for validation, and integrating it into a client-side product will surely have many benefits which my article enumerates. Doing will also surely create versioning issues, but it is very clear to NetNewsWire customers that the validator in NetNewsWire is part of NetNewsWire. If NetNewsWire customers have issues with it, they will go to Brent first, and if necessary he will come to us.
Dave-Andrew's web-based validator fork, on the other hand, is not a new venue. It has no visible contact information, and no acknowledgement of the original project other than a misleading link on the results page which makes it appear that their fork is somehow official, and that all bug reports for their web-based service should be directed first and foremost to the bug reporting system on Sourceforge (which is wrong). As the leader of the RSS Advisory Board, Dave will undoubtedly link to his validator fork from the RSS spec, and as the founder and current chairman of Userland, Dave will undoubtedly pressure Userland to add support for his validator in Radio, Manila, Frontier, and other Userland products.
Meanwhile, people who use none of these products and don't read specs will search Google and stumble across all of the validators, and have no idea which one to use. It has been my experience that people use all the web-based validators they find. I say this based on the steady stream of bug reports that I receive of the form "your validator says my feed is broken, but aggregator.userland.com/validator says it's fine". I see this has finally been taken offline and now links to Dave-Andrew's validator. Thanks for that, but since Dave has publicly stated that he has no intention of supporting or maintaining his validator, the next time we rev the real Feed Validator, I expect to start getting new bug reports of the form "your validator says my feed is broken, but rss.scripting.com says it's fine". This is not an improvement for anyone, least of all the users who just want to do the right thing.
Dave, your validator is a fork that benefits no one. You are hurting the very people you claim RSS is all about: end users. Please shut it down before you cause any more confusion.
Posted by Mark at[...] steady stream of bug reports that I receive of the form "your validator says my feed is broken, but aggregator.userland.com/validator says it's fine". I see this has finally been taken offline and now links to Dave-Andrew's validator.
That just made my day.
Posted by Sam Ruby at"feedvalidator.org feeds.archive.org feedvalidator.sourceforge.net atomenabled.org rss.scripting.com"
Which one of these is not like the other ones? Which one of these just doesn't belong?
Posted by Aaron Swartz atMark, you had the chance to stop this... when you wrote the license. If you really don't want other people to use the validator, you should have said so then, rather than complaining about it now that people are taking you up on your terms. You could have stipulated that no one was allowed to create a web front end for the validator. You could have included a clause stating that any site embedding the validator had to link to feedvalidator.org and mention that feedvalidator.org is the definitive version. You could have done lots of things, but you didn't. Sure, the validator wouldn't be so Free, but your posts here suggest that you only want it to be free subject to your approval anyway.
Posted by jgraham at
jgraham: to be fair, I talked Mark into this particular license. That being said, what is currently deployed on rss.scripting.com apparently is a Frontier front end to the FeedValidator backend, with perhaps one file changed. The backend was always intended to be open source, and given that, the most that could have been reasonably requested is that the modifications be released back... and only then if the license didn't have a distribution loophole.
Posted by Sam Ruby at
jgraham: it is not true that I want the Feed Validator to be "free subject to my approval". It was very important to me that the validator backend be open source, and Sam convinced me that a Python license was better than the GPL (so that, for instance, closed-source client-side programs like NetNewsWire could embed it without negotiating a special license with us). This is just now beginning to bear fruit, as the next version of NetNewsWire will embed the validator, which is great news.
The problem here is that Dave-Andrew have forked the validator service, taking our backend code and slapping their own interface on top of it. Despite what they're trying to sell, they're not "using our code" in any constructive way; they're "using our code" in order to clone our site in order to directly compete with us (badly, I might add). We knew this was possible, given the license we chose for the backend code, but we didn't actually believe that anyone would be rude enough to actually do it. After all, what good could it possibly do to have two identical web-based validators?
We also tried to mitigate the risk by adding things to the real Feed Validator website which were not open source, such as the main CGI script which did feed autodiscovery and SOAP, the easy-to-use stylesheets and layout, and the extensive integrated documentation. We figured that even if someone were malicious enough to try to fork the validator, they would have to spend a lot of time playing catch-up, recreating those portions of the service that sat on top of the open source backend code. As it turned out, when Dave-Andrew forked the validator, they simply ignored all the functionality they couldn't steal, slapped a minimal UI on our code, and spent the rest of the afternoon patting themselves on the back for creating a "brand new validator" using nothing but Ctrl-C, Ctrl-V, and the Backspace key.
The problem here is that Dave-Andrew have forked the validator service
Forked in what sense? I was under the impression that there was already a validator service at scripting.com? Forked in the sense that, instead of using their own code, they're using your code? And might change it? It seems to me that that's just a risk of releasing software as Open Source - and the reason so many companies are loath to do so. If some BSD contributer complained at Apple for using his code because "they forked the Operating System", they wouldn't have a leg to stand on. The reverse is true if you used the Open parts of OSX to build a competing operating system. It's just the nature of the game - if you decide to open the code under a liberal license so people can do what they want with it, you can't reasonably expect that they will only do things that benefit you.
Having said that, I think, for this kind of software closing the distribution loophole would be a good idea. However, that would, as you said, make it more difficult to incorporate the validator in closed source software. Maybe you could dual license it; something GPL-like for open source products (with modifications to include network access in the definition of redistribution), and a license that forbids redistribution of changes for closed source products. Obviously, at this stage, that change may be just as harmful as it is good.
Posted by jgraham atForked in what sense?
Forked in the sense that there are now 2 web-based validators run by 2 different groups, based on what started out as 1 code base but are now 2 different code bases being maintained by 2 different groups. Is there some other definition of "fork" that I am unaware of?
if you decide to open the code under a liberal license so people can do what they want with it, you can't reasonably expect that they will only do things that benefit you.
Well, we can hope. But yes, you are correct that I can't claim that they are doing anything contrary to our license, which is why I have never stated that they are doing anything contrary to our license. That doesn't make it a good idea.
for this kind of software closing the distribution loophole would be a good idea.
A conservative is a liberal who has been mugged. A GPL advocate is a BSD advocate who has had their work used against them. As I stated earlier in this thread, Sam convinced me to go with a liberal license over the GPL for the backend code, and we tried to mitigate the risk by adding extras in the web-based frontend which were not open source. Next time we'll know better than to think that usability would be a deterrent against paranoid competitors with big egos.
There is one nagging question that neither Dave nor Andrew has addressed: what purpose does it serve to have 2 web-based validators? (This question is also addressed to atomenabled.org's validator, which I strongly opposed when it was brought to my attention and which I strongly oppose now.) If there were some insanely cool new feature that they wanted to implement on top of the validator, that they had discussed with us and that we had rejected outright and that they felt was going to change the world enough to justify a fork of this magnitude, then that would be one thing. But Dave-Andrew have no such features planned; Dave has publicly stated that Andrew may or may not even keep the backend code up to date as specifications evolve, and that he certainly has no plans for additional features, or maintenance, or even basic end-user support. WHAT IS THE POINT OF THAT? No one has coherently answered that question.
Posted by Mark atRE: Mushrooms
Mark wrote
>There is one nagging question that neither Dave nor Andrew has addressed: what purpose does it serve to have 2 web-based validators?
Interesting. Can I also ask what purpose it serves to have more than one operating system, web browser, or C++ compiler? What exactly is wrong with choice? I thought that was the whole point of Open Source.
So far I haven't seen Dave and Andrew do anything malicious yet. They are well within their rights to take your code, change some of the validation rules and slap it on the Web since it is Open Source. So far all they've done is change the UI.
It seems you are basically arguing that your validator should be the only one that exists. Your way or the highway...
Is there some other definition of "fork" that I am unaware of?
My personal definition of forked requires the forking party to add features to the product which are incompatible, or not contributed back to, the original product. If they haven't actually changed anything, it's not a fork, it's just a distribution. If they are using an outdated version, it's not a fork, it's just an outdated version. That may not, I suppose, correspond with the "official" definition of fork.
2 web-based validators run by 2 different groups
Am I right or wrong in believing that this was the situation before they started using your code? As far as I am aware, there have been two web based validators for a while now, and the fact that they are, at present, using very similar code is an improvement over the situation as it was a few days ago when the validators had entirely different behavior.
There are certainly multiple validators for HTML, and the world doesn't seem to have ended yet.
Next time we'll know better than to think that usability would be a deterrent
So? If your validator is better, one might hope that it becomes the defacto standard. At least, that's the theory, although I've never been too convinced by that theory. In any case, if you're more concerned with Atom than RSS, your validator has an excellent chance of becoming a defacto standard. You can even specify it as the official Atom validator, if such a thing takes your fancy.
WHAT IS THE POINT OF THAT?
Maybe they wanted to replace their old validator with a newer, better one, whilst still providing the same service to their customers? Maybe they don't like your "Valid Feed" icons? Maybe they just wanted to annoy you personally?
Next time we'll know better
Please don't speak for me. I still think that the Python license was the right license for this code.
Posted by Sam Ruby atTo be clear, there was a point to my post about web browsers (I wasn't being entirely facetious) - validators are certainly not alone in the class of software where different implementations leads to headaches for users and authors. If Mark gets email asking why his validator says a page is invalid per the feedvalidator.org validator but validates in the scripting.com validator, that provides a clear parallel to the situation where a web page is functional in one browser but not in another and the web master tells people who complain that their tests show that the page is fine. However, there are very few people asking what the point of having multiple web browsers is.
If you'd like to extend the analogy further, there's a bunch of Gecko code in Safari. So Apple forked parts of Mozilla, diluted them into a bucket of KHTML code, and released a new browser when they could have just gone with a gecko based browser and reduced the compatibility problems a new rendering engine creates.
Posted by jgraham atUsually people are happy when others provide mirrors and share the bandwidth bill. I can see that labeling Atom feeds as valid RSS is considered insulting, though.
Posted by Henri Sivonen at
So Apple forked parts of Mozilla, diluted them into a bucket of KHTML code, and released a new browser when they could have just gone with a gecko based browser and reduced the compatibility problems a new rendering engine creates.
Yes, and a lot of people made exactly that point when Safari was first released. But Dave Hyatt has spent the past year working incredibly hard at improving Safari's rendering engine and releasing the changes back to the core KHTML team. And Dave's coworkers have worked hard adding an interface on top of that rendering engine that my father can use (and does -- he even manages his bookmarks without calling me).
Back to Dave-Andrew's validator fork, they have done none of this, and they have stated that they have no plans to do anything like this. So my original point stands, which I will repeat since none of you seemed to have been paying attention the first time:
Posted by Mark atIf there were some insanely cool new feature that they wanted to implement on top of the validator, that they had discussed with us and that we had rejected outright and that they felt was going to change the world enough to justify a fork of this magnitude, then that would be one thing. But Dave-Andrew have no such features planned; Dave has publicly stated that Andrew may or may not even keep the backend code up to date as specifications evolve, and that he certainly has no plans for additional features, or maintenance, or even basic end-user support. WHAT IS THE POINT OF THAT? No one has coherently answered that question.
I think that mirroring the validator at some UserLand site would have been a great idea - with identical functionality and proper credit to the developers as well as the company providing the hosting. I can't believe Sam or Mark would have rejected that offer. It might even have been a good way to demonstrate that both "parties" care about standards.
Dave's assertion, though, is that the feedvalidator's Atom validation feature somehow uses RSS to promote Atom. Even if that were true (and I don't think so), an acceptable way for Dave and Andrew would have been to develop their own validator as an alternative. Licensing aside, I believe it is obviously wrong to use somebody else's work without even giving proper credit.
Posted by Stefan Tilkov atBut Dave Hyatt has spent the past year working incredibly hard at improving Safari's rendering engine and releasing the changes back to the core KHTML team.
But he didn't fork KHTML, he 'forked' some of the Mozilla code (unless there have been a bunch of checkins from Apple into Mozilla CVS which I haven't noticed - this seems unlikely because of licensing issues). Regardless of whether Safari has become a good browser or not (it has) the point is that Apple forking Mozilla:
Doesn't benefit Mozilla or its users (because no changes are returned to Mozilla)
Doesn't benefit web authors (because they have a different set of CSS/DOM quirks to work around and need a proprierty OS running on proprietry hardware to do so effectively - unless Konqurer is keeping up with the changes faster that I think it is).
Apple acted entirely out of self interest. Everyone here is acting out of self interest. People, in general, act out of self interest.
Posted by jgraham atBut he didn't fork KHTML, he 'forked' some of the Mozilla code
This is the first time I've heard of Mozilla code going into Safari. What code was it?
unless Konqurer is keeping up with the changes faster that I think it is).
Konqueror is reasonably up-to-date, 3.2 was released a day or two ago, and CVS is always available.
Posted by Jim Dabell atStefan, if you think we wouldn't have gotten roasted for developing our own validator you don't understand how it works. People would have said we just don't "get" open source, or we're trying to undermine Sam and Mark, or whatever.
Also, I sent a heads-up to Sam, privately, a few days before the public announcement, explaining what we were doing, and giving him time to comment, or ask for clarification. I didn't get a response. I don't know how we could have played more fairly.
BTW, Mark's conjecture is incorrect, I've got a new service to announce shortly, built on the validator. I have to take a lot of bashing from him, it's not fair, but I've come to expect that from him. C'est la vie.
C'est la vie shouldn't be flamebait, should it?
One point of having another: if I was stuck in an environment where I couldn't install my own copy (I'm not, I did), but wanted to automatically validate my feed when I added an entry, I could screen-scrape Dave's validator's output. At least, unless and until he adds TOS saying I can't.
Posted by Phil Ringnalda atThis is the first time I've heard of Mozilla code going into Safari. What code was it?
I'm not exactly sure. grepping the Safari source gives the files:
- khtml/rendering/render_layer.h
- khtml/rendering/render_layer.cpp
- khtml/rendering/render_arena.h
- khtml/rendering/render_arena.cpp
- khtml/misc/arena.h
- khtml/misc/arena.cpp
- khtml/html/html_documentimpl.cpp
As being under the MPL. I'm not sure these are the only files because the MPL might allow you to used MPL'd files under the (L)GPL and release modifications under one of these licenses (which would prevent the modification going into the Mozilla tree).
I'm certainly not claiming it's a lot of code (although I vaguely remember someone saying it is quite nasty, hard to get right code). The point is, Apple took a look at Mozilla, decided they didn't want to use most of it, but still took the bits they did want to use. The only people who benefited from this were Apple.
Posted by jgraham atPhil, or we might add an XML-RPC or REST interface (we use a REST interface internally).
Or we might implement periodic validation with reports via email. (Tease tease.)
If you look at the "Feed Validator" it's tilting in favor of Atom. We didn't complain about that, instead we did something positive, providing the validator we felt was fair to RSS, which we have a lot invested in and in which we want to invest more.
I think where this leads is some working-together, and maybe some fun. What would be so terrible if that happened??
Posted by Dave Winer atI'm not asking for Mark's respect or praise or even approval. He's written that what we did was "rude" and "amateur", and while he could have been more diplomatic about it, in most of the comments I saw a positive effort to point out what he objected to and how he thought the service could be improved. I support that.
Speaking for myself I never intended to be rude or to fork his work. I wanted to promote RSS and maybe contribute to an interesting open source project. I have not forked his and Sam's code. I've studied it and submitted patch that I hope will improve the core codebase. I intend to do more of the same as the needs of the new service demand it.
Getting back to the comment that launched this post...What kind of way is this to treat potential collaborators? Back at ArsDigita we used to develop an open source web toolkit called the ArsDigita Community system. Hundreds (at least) of Web applications were launched on it. Thousands of people downloaded it, including our competitors, competitors that sometimes were bidding on the same projects we were!
All the while we continued to code away furiously to build the toolkit and launch customer sites. Philip Greenspun would give these free marathon weekend classes to teach people why and how to build Web applications. I remember one time when Janine and Mike Sisk from Furfly, a Web application development company just like ArsDigita, showed up at one of the classes. Philip asked them to stand up. Did he chew them out for using our tools, deride their work, and ask everyone to shout "tsk! tsk!" as they marched from the room in shame? Of course not. He said, "Say 'hi!' to Janine and Mike, who downloaded our toolkit and built a successful business on it."
Posted by Andrew Grumet atSome stuff I left out bears noting. The free classes and code were not acts of charity. The free classes were unbelievably good marketing for the company. Each weekend landed at least one million-dollar customer. It still boggles my mind that the post-VC management team didn't recognize this. The open source code, well, I probably don't need to champion its virtues to the people hanging out here. But being open source meant having competition. We welcomed that competition because we wanted to spread our ideas and be seen as leaders. And because we recognized that the world didn't owe us its attention.
Posted by Andrew Grumet at
Dave, you may be right that - no flame intended - there would have been some issues between Mark and you (though maybe not with Sam. Who knows.) But the main point I can't get over, and that I find really annoying, is that even the new 'scheduled validator' fails to give credit to the people who wrote 99.9 percent of it (as it is only a very thin layer on top of the original validator).
I think you would agree that acknowledging other people's contributions is a simple matter of good manners, and I seem to recall your pointing this out on the Atom list. Maybe it might be a good idea to follow one's own principles.
Wow. Guys. Couple of steps back, couple deep breaths right about now would be good for all.
I think it was a very cool thing for Mark and Sam to open up their source like that. I also think the license used was a good one, because ultimately the feeds being valid should be the important aspect of all of this.
With this type of license, yes people can fork a copy off and do their own thing. The gracious thing to do in this case is give credit to those who provided the open source code, but if the license doesn't demand it, well, such is life.
But this 'choosing friends' talk -- wow. Doesn't help RSS, Atom, or open source. Doesn't seem like its helping anyone here either.
Too bad. Well, darn. I need to go find cute baby or puppy or kitten pictures to look at now.
Posted by Shelley atBut this 'choosing friends' talk -- wow. Doesn't help RSS, Atom, or open source. Doesn't seem like its helping anyone here either.
Agreed.
It was clear that people weren't getting the message with my strike outs. I have now marked a number of recent comments have been as spam. This means that any such messages will enter into my throttle calculations for 72 hours.
I have noticed a tendency of people to say things on other people's weblogs that they wouldn't dare to put on their own weblogs.
Please don't respond on my weblog to comments that I have marked as deleted. Your feelings may be honest, valid, and you may have a genuine need to express then. If so, please do feel free to express them - on your own weblog. Feel free to use trackback or pingback to draw attention to any such blog entries.
Posted by Sam Ruby atBabies and kittens
This is my daughter Talia. She's nearly two and a half, and quite a handful. These are my cats The orange, tiger-striped one is Ludo. His name comes from Ma Vie En Rose, which we watched shortly before adopting the...... [more]Trackback from Andrew Grumet's Weblog at
Huh, that trackback came out looking stark and aggressive. Not what I intended at all. I thought Shelley's suggestion was an excellent one. Looking at these pictures helps to put things in perspective for me. That's what I was trying to contribute with the trackback.
Posted by Andrew Grumet at
Atom Validators
Online Atom feed validators : feedvalidator.org feeds.archive.org feedvalidator.sourceforge.net atomenabled.org rss.scripting.com localfeeds.com These are all built around the open source Python code from Mark Pilgrim and Sam Ruby. (Via Sam)......Excerpt from Finally Atom at
Not Again
I suspected that a non-issue like this won't turn into a problem, but there's nothing that a few skilled trolls can't achieve. Keep your eyes on this thread to watch a bunch of software developers and other tech-inclined people turn into...Excerpt from Firas at
Aftermath
I don't ask much. I only ask that you see the world exactly as I see it. (1233 words)...Excerpt from dive into mark at
The aftermath is better than the thing itself
I really liked Marky Mark's recent entry "Aftermath". I especially liked this: Certainly Apple + BSD is a success story, but like the guy on stage at an Amway convention, every success story hides a thousand failures. How many times has a company...Excerpt from Hacking Log 4.0: Phase II at
let a thousand validators bloom
No, I'm not creating a new validator, see below... Dave Winer and Andrew Grumet released rss.scripting.com yesterday (announcement), a new web front end to the Mark Pilgrim/Sam Ruby open source RSS & Atom Feed Validator, that claims to be purely...... [more]Trackback from Ross Notes at