Stop publishing in RSS

People who generate syndication feeds have a choice of feed formats. The major candidates are Atom1.0 , RSS2.0, RSS1.0.The RSS 2.0 specification is copyrighted by Harvard University and is frozen. No significant changes can be made (although the specification is under a Creative Commons license) and it is intended that future work be done under a different name; Atom is one example of such work.Atom 1.0 is specified in RFC 4287 (HTML Version); it represents the consensus of the Atompub Working Group within the IETF, as reviewed and approved by the IETF community and the Internet Engineering Steering Group. The specification is structured in such a way that the IETF could conceivably issue further versions or revisions of this specification without breaking existing deployments. Its high time that everyone in the Feeds community get together and decide to leave their differences in the past, and create one standard format. As for me, I have worked on both Atom and RSS, and personally feel that the Atom community is far more stable, better accepted, has more support, and more developers working on that. So, I strongly feel that RSS format should be scraped.

The actual problem with multiple formats is that, different readers (like bloglines) consume different formats(in some cases the consume more than one format). So all the sites/ blog platforms that support Atom are publishing multiple feeds in the autodiscovery section of the blog. So when a user goes to subscribe to the blog the browser presents a completely confusing choice of multiple feeds to subscribe to. That choice is meaningless to blog readers and just causes unnecessary anxiety. Even blogger advertises both RSS and Atom :( . Ironically, many of the RSS links are published through Feedburner which has some automatic feed rewriting that ends up serving Atom anyway.

Although RSS was the first format, it lost its way/value somewhere in the middle. It was a great idea which was started in Netscape. But clearly Atom is way more matured compared to RSS. Its time for RSS to be retired.

Cafe Babe!

I have been working on Java for 2 years now. But I got to know about this kinky side of java couple of days back. Till then, whenever i felt the need to look at a compiled java class i used to use either a Java Decompiler or use javap (disassembler). But 2 days back I used a Hex Editor to view it. And to my shock found the first 32 bits to be 0xcafe 0xbabe. What a genius who wrote java compiler. Relating everything to Coffee, Coffee bars and the beautiful gals who come there. It definitely has to be a man. To my surprise I found that a java interpreter determines whether it is a java class file or not, by looking at first 2 Bytes. So all newbie hackers out there(like me), who want to deal in the world of 0's and 1's can think of this as a humble start!

Experience with Retroweaver!

Develop With 1.5 And Deploy With 1.4!

For a couple of months now I am using a open source project called Retroweaver. The objective of the Retroweaver is to weave JDK 5.0 compiled java class files to run on a JVM1.4.x. It means that the Retroweaver is a bytecode weaver, that enables us to take advantage of the new Java 1.5 language features, while still retaining total binary compatability with 1.4 virtual machines. Retroweaver operates by transforming Java class files compiled by a 1.5 compiler into version 1.4 class files which can then be run on any 1.4 virtual machine. All the generics, annotation support, Enum, varargs, extended 'for' loops, reflect API's etc. which are all features on Java 1.5+ can be now used in Java 1.4.

Very impressive till now. It promises a lot of thing, and most of them do work wonderfully well. Although the concept they use to weave is pretty simple (they mirror all the java class files that are specific to 1.5 and produce 2 jar files. First one is used to weave the 1.5 class files. This runs through all the class files, finds all the instances where 1.5 features/API's are used and replaces by its own API's. Ex: When it finds the statement "import java.lang.annotation.Annotation; ", it replaces it with "import net.sourceforge.retroweaver.java.lang.annotation.Annotation;". The second jar file is the runtime jar file. This contains all the "replica" 1.5 rt.jar files which are not there/modified in 1.4. Pretty simple ah?

Now coming to the other side of the story. Till now I have mentioned all the good features about this product. So let me mention all the negatives about it as well. It is very depressing to see a dormant open source community. I am not sure if many people are actually using this product. Because I myself have encountered many bugs in it, and when I raise these defects, only ONE person always replies (that too delayed response). There are 2 full time developers, no committers that i know of, and no one who keeps a close eye on the forums/mailing list/bug reports. So if you have a bug, you better solve it yourself than wait for them. I have done this a couple of times now, after painfully waiting for 1 month for a reply from them. The other sad part about this product is the code, which is not up to the mark, in the sense that, a 3rd party user can never understand their code( no comments, no java docs).

Overall I am actually not very impressed by this product. Not a product I would recommend any one. There are similar other open source project like the RetroTranslator, which unfortunately faces the same problems. So 2/5 rating for Retroweaver. Don't use it unless you don't have any other alternative.

New era of Petaflop!

Yesterday, the IBM developed "RoadRunner" supercomputer for the U.S. Department of Energy’s National Nuclear Security Administration, broke a historic barrier: one thousand trillion calculations per second -or one “petaflop.” I was really excited to hear this news, and to know that my company is constantly pushing the barriers of innovation, to make the world a better place to live in.