Enterprise readiness of Jython?
Written on December 10, 2005 by Conor O'Neill
In our first Python post we mentioned that there can be push-back on the usage of Python in solutions for customers, particularly if the entire system is to be built in Python.
An interesting alternative which can gain more customer acceptance is Jython. This is Python reimplemented in Java and it provides features which are compelling both from a technical perspective and customer acceptance perspective. To the customer, Jython can be viewed as simply another part of a Java based solution where it is providing some scripting, automation and high-level glue. For the developer, it opens up the facilites of Java to the Python interpreter and all the facilites of Python to the JVM.
Tim Bray (Mr XML, now in SUN) has a particularly good blog posting on it.
So, in theory, we like it a lot, and we are particularly looking at it for integrating together existing Java sub-sytems from other vendors and from our customers. However, we are concerned about the history of its development which has been very stop-start and also the lag between it and the mainline Python interpreter. The alpha status of the current release would obviously have to change before we could consider doing a major customer development in it.
Are there any developers using this for Enterprise-level software with actual live, stable deployments in the field? If you are wondering what we mean by “Enterprise-level”, an example would be the system we delivered to a customer last December which has had 100% uptime over the past 12 months.
Tim Bray mentions that Propylon have deployed Jython systems in the field, which is very interesting.
Additionally, as with the previous post, we are interested in feedback on development environments for Jython. We are currently playing with both PyDev JyDT and along with Editor+CLI and Coyote on Netbeans.
Technorati Tags: Python, Jython, Enterprise, Enterprise-level, Eclipse, PyDev, JyDT, Netbeans, Coyote
If you enjoyed this post Subscribe to our feed
I’m in a similar position to yourself, constantly evaluating the alternatives to Java to see if any are ‘enterprise ready’.
From a technical point of view, I think Jython is about as ready as any of the alteratives. Given that the next release of Java will have Rhinoscript (aka Javascript) embedded in it, the buzz around Ruby on rails and familiarity of Groovy’s syntax to Java programmers, the competition in this space is quite large!
Compared to these, my choice of the ‘next big thing’ is quite boring. It’s already in use in a lot of sites and just gets the job done. Yep, it’s PHP. While it’s not attractive to Java people in a lot of ways (it’s easy to use, which means that rates are not as high), IBM and millions of websites can’t be wrong!
We are always of the mind-set of “right tool for the job”. PHP is perfect in certain situations (specifically web-related ones) and we wouldn’t be sitting on top of Wordpress if we were not happy with PHP. However we are looking at more general non-web applications (such as database ETL) where we like the idea of stitching together and automating lots of java sub-systems with an overall glue like Jython (or Groovy or whatever).
Everything we have evaluated to date seems functional and capable but if we are to go beyond building small systems and prototypes in Jython, we’d really like to get more field feedback of success stories and horror stories.
Appreciate your comments!