Wednesday June 16, 2004

RSS Specifications and schemas (well, not really)

This morning, in an internal Sun alias, somebody asked

Are there specifications available for the various RSS formats (not including the atom work at the IETF)? ... Dare I even hope for something like a schema?

YEAH RIGHT, I want that too. My reply to that poor soul was something like this

If you dare calling them specifications here they are: RSS 0.90, RSS 0.91, RSS 0.92, RSS 0.93, RSS 0.94, RSS 1.0 and RSS 2.0.

I'm still on the quest for DTDs or XML-schemas for all of them but the folks that wrote these *specs* apparently were too busy for such distractions.

And there is even more, you have even different versions with the same version number, isn't that great? Mark Pilgrim did a very good analysis of the different RSS versions.

I'm sure others will find useful having all these links together.

Rome v0.2 is out

We've just postedRome v0.2

We've been fighting to get everything ready for the release.

Summary of the changes:

• Removed dependencies on other components except JDOM
• Hid implementation packages/classes from public API
• API interfaces/classes renaming
• Some API signature changes

Some of them were done based on some feedback we've received. Others to try bringing consistency to the naming. And others to enable certain usage pattern in applications using Rome.

All the changes (15 of them) are documented in the Changes Log.

I hope feedback keeps coming, that will help us make Rome better and easier to use.

Enough of this, now I have go back to my day job.

To myself, for the Nth time, *never underestimate the effort behind doing a release*

And I'm talking about a small project, Rome v0.2 (It's almost there).

Getting the code changes integrated it's the easy part. Getting all the supporting materials ready is a Pharaonic task. General documentation, changeslog, tutorials, javadocs, updating the samples, testing the samples, preparing the distribution bundles and staging all this in the site. Ensuring (or at least trying to) everything has been updated.

We don't want to get into the we can do that later approach as most of the times later becomes asymptotic to never.

With Rome we are still organizing the site. Deciding where things should go etc. We are using Rome Wiki as the main (and only for now) documentation source, we have to decide what documents we want to keep alive (changing as we go) and what documents we want to freeze with every release. Maybe things will get easier after we have all this settled. I'm making a mental note (too bad I keep losing them) to comment again on this for Rome v0.3, to see if things got better.

Thursday June 10, 2004

Rome is making some noise already

It was a nice way of starting the day today, I've found out there is some blogging about Rome going on.

Some of them commenting the appearance of it. Others expressing some concerns about it. That's great, we welcome both. The latter ones will help us improve Rome. I hope they keep coming.

P@ posted a blog earlier today -Rome was not built in one day- making some clarifications.

Tuesday June 08, 2004

Trying to end Syndication feeds Hell, at least for Java developers

After a month of work or so we did it. 'We' are Patrick, Elaine and myself. 'What we did' is Rome.

You may not care who we are, but [if you are a Java developer and you are doing anything that involves RSS or Atom syndication feeds] you'll likely care about Rome.

Rome is a set of Atom/RSS Java utilities that makes it easy to work in Java with most syndication formats. It supports all flavors of RSS (0.90, 0.91, 0.92, 0.93, 0.94, 1.0 and 2.0) and Atom 0.3 feeds. It includes a set of beans, parsers and generators for the various flavors of syndication feeds, as well as converters to convert from one feed type to another. It also includes a generic normalized SyndFeed class that lets you work with any feed type without bothering about the incoming or outgoing feed type.

Rome is one of those projects I'd had preferred not to have spent time working on. It came to be out of frustration, or better said to stop frustration.

What kind of frustration I'm talking about?

First, the nonsense of syndication feeds. Second, the absence of a library that does a good job abstracting all that nonsense from the developer. And third, the quality of available documentation (this last one is in fact a general problem.

So, we decided to do something about and that something is Rome. By addressing the second we took care of the first (insert second part of the title here). And for the third -the documentation- we made an special effort from the very beginning.

Rome is just starting, there plenty of things to improve, to fix and to add.

Check it out and let us know what you think.

Monday June 07, 2004

Syndication feeds Hell, Addendum I

When writing Syndication feeds Hell I've forgotten to mention the following.

Leaving evolution aside -which it could have been done in an orderly and backwards compatible manner- I don't see the different RSS flavors and Atom having irreconcilable technical differences. It's all just an EGO thing.

Sunday June 06, 2004

The changing images on the top left corner of the page

Snapshots from some of my trips.

Making them change randomly when page loads. Some dummy javascript.

Here you have a set of nine:

To blog or not to blog

It has been a dilemma for me. I wouldn't say I'm much of a blogger. I've tried a few times before but I've given up. It takes time [trying] to do it well. We'll see how it goes this time.

Tucu WHO?

I've been with Sun since early 97 and these days I work in the Sun Portal Server engineering team on some collaboration initiatives.

Why you may have heard my name ...

Does JSR168 JavaTM Portlet Specification ring a bell? I was one of the specification co-leads. As spec-lead I should have to take much of the blame. But I'm not alone there, I have a co-lead (Stefan Hepper) and a whole Expert Group to share it with, community effort I'd say. Doing JSR168 was a great experience. I've got to work with many sharp folks from different companies. I had to learn some political skills. I had fun working on the spec and I'd say I was not alone. We even managed to have an Expert Group face to face meeting in Breckenridge (CO) in January (you connect the dots :). Or Web Services for Remote Portlets (WSRP)? I've participated on the doings of WSRP specification as well. Some more blame to take. I've met more smart people. I've learned other political skills (it's a complete different ball game when you are not the captain). We had fun there too. After a face to face in Grenoble (France) I've wandered through Pais Catar (France southwest) for a few days.

In a previous life I've worked in the iPlanet eCommerce division, in the iMM (iPlanet Market Maker) product designing and implementing a multi-dimensional spot exchange (sounds cool, doesn't it?). Before that I was in Sun IT doing reusable components in Java, many of them are part of today's Sun internal IT applications. Rewinding a little more, I was living in Buenos Aires, where I've worked for Sybar (Sybase Argentinean Distributor back then) doing technical support, presales support, consulting and some management. I'm sure there are still Sybase customers down there that have fliers with my picture offering some mony in exchange of my whereabouts. And I believe the Argentinean customs MARIA system is still using some C/C++ libraries and UI framework I did for them (through SIF America) back in 93.

Syndication feeds Hell

Before you continue reading, the topic covered in this posting has been widely discussed already. Following you'll find a rant on the nonsense of Syndication feeds, RSS0.9 ... RS2.0 and Atom. I just needed to write it to take it out of my system. You've been warned.

I've been following some of the technologies used and/or being developed around weblogging for a while now. I find very interesting how features and functionality keeps being added to the weblog world with (almost) no impact on the bloggers, their weblog systems take care of doing most (if not all) the job (ie: web pinging, trackbacks, etc). My interpretation is that [maybe] the reason for this it is that not always these features are brainchildren of technical oriented folks (you know, programmers). Or if they were thought by technical oriented folks, those folks knew the fact not everybody is techie. It's like the approach has been “It would be nice if when I post a weblog the following happens" instead of being “It would be nice if I can do the following when posting a weblog".

Granted that things get a little hairy when you get N ways of doing the same thing in a short period of time where N closely equates to the number of weblog software packages out there. In the end it's Darwinism in action and after a while things settle. Not a bad thing, even if the prevailing technology doesn't do it for its technical merits.

But there is an exception. One of the first and core features weblogs rely on, syndication feeds. Things went wild there and it is not that we are getting insects, reptiles or mammals. They are all almost the same. And *almost* is the problem. I won't go on the history, all the differences among the different types of feeds or their design problems, pretty good articles and weblogs have been written about it. One of my favorites is Mark Pilgrim's The myth of RSS compatibility. And don't forget to add Atom to hte mix.

It's just plainly insane. Their specifications shouldn't be called that. Think, if you are doing an application that will parse an XML document, what is one of the first things you want to do, ensure the XML document is valid. To do that you'd like to have a DTD or the XML-Schema handy. With the exception of Netscape's RSS 0.91, (that has DTD), none of the RSS or Atom specifications includes a DTD or XML-Schema(*). If you are a programmer that has to deal with syndication feeds, you know this is just the beginning of your nightmare.

A few of months ago I started working on some projects around collaboration and I've became a victim myself of what I call Syndication feeds Hell.

Addendum I

(*) If you find them please let me know, I've spent Google hours with no luck.