Friday, October 22, 2010

IronPython is Dead; Long Live IronPython

As expected:

http://blogs.msdn.com/b/jasonz/archive/2010/10/21/new-components-and-contributors-for-ironpython-and-ironruby.aspx

This is good news: IronPython development will now be controlled by the community, as will IronRuby.  In IronPython’s case, the coordinators will be 4 heavyweights from the Python and compiler development communities.

Miguel de Icaza: head of Mono project, likes Python, compiler expert. 

Jimmy Schementi: worked at MS on DLR team and then Iron* teams as compiler guru, now works for financial industry.

Jeff Hardy: Python guru, works in .Net.

Michael Foord: Python guru, wrote the best book on IronPython, firm he works for is built on a product running on IronPython.

Why It’s Good News

Why is it good news?  The short answer is that development should progress faster than it has, due to community involvement. 

The longer answer is that until now, only Microsoft could contribute, due to legalities involving liability for code, etc.  The 4 coordinators, on the other hand, all have track records of producing results in Python and compiler development, and in at least 2 cases, work for employers who have an interest in having IronPython continue to grow, as they have products built on IronPython.

The presence of Miguel alone should be heartening: he leads the Mono effort bringing .Net to Linux, working for Novell (who in turn have licensed Microsoft patents to protect customers of their Enterprise Linux solutions).  He has consistently delivered for many years, with Mono often outdoing .Net at its own game (e.g., a compiler that can run as a service for continuous integration), having the advantage of being second and seeing the weaknesses before, rather than in the middle of, development.  Mono exists with the full knowledge of, and in fact assistance from, Microsoft, who see interoperability with Linux as being in their own interest.

If Microsoft held on to IronPython, the question would have nagged: will MS do to IPy what it did to VFP – starve it into (hoped-for) non-existence?  Instead, there is now no question but that IronPython will continue to thrive.

In the end, the best revenge would be such wide-spread adoption of IronPython that MS wishes it had the product back.  While hoping for revenge is self-injurious, hoping for others to experience the benefits of programming in Python under .Net seems, as of today’s announcement, to be both caring and realistic.

PS: (added 12:59 pm)  What about Jim Hugunin, who created IronPython and brought it to Microsoft (aided by Jason Zanders, the author of the blog entry referenced above)?  He’s going to Google. Note that Python is a primary programming language at Google – to the point that the Google Apps API to this point has always come out in Python first, and Java second.  Also note: Jim created Jython, Python in Java, before creating IronPython.  It will be interesting to see what he does at Google.  That he is going to Google does seem to say a lot about the differences, in many ways, between Google and Microsoft.  Or at least, for his sake, we should hope it does.

3 comments:

Vfp likes me said...

I beleive also that ipy can now breathe healthy air. It was previsible as 3 or 4 months ago the same happened with ironruby.
Camil

Hank Fay said...

Yes indeed. And in fact, Jeff Hardy confirms that as of 9/1 of this year, the fate of the IPy team had been settled. The air is much healthier out here. <s>

johnf said...

I still do not see the advantage of continuing to stay with MS.

In your case this is the second time you have have discarded. If you would really like to be independent then just use python and whatever GUI lib you like.

If not then stay with C#.

BTW I'm sorry you invested so much into IronPython only to see go away just like VFP.

Also a word of warning - many believe Novell is in major trouble and that means so is Mono.