IronPython-0.6 is now available!

"Martin v. Löwis" martin at v.loewis.de
Thu Jul 29 22:53:56 CEST 2004


Derek Thomson wrote:

> I haven't yet found the time to look at any of this stuff in detail,
> but I will, and it wouldn't surprise me at all to learn that it was
> really quite good in many respects - MS didn't just forge it all from
> the ground up in total isolation, ignoring all that had gone before,
> hoping to get it working right in the 3rd release.

I have worked a lot with C# and .NET lately, and found two things:
- the VM is very well designed. It does not really work for languages
   other than C#, though (and, yes: I believe that it does not work
   well for Python either, despite IronPython's existance (*))
- C# is ok. There are a few unfortunate (IMO) hacks, like the notion
   of delegates (which are the only callable objects), and the surprising
   behaviour wrt. to interfaces and overriding (where functions in a
   subclass won't implement an interface operation defined in the base
   class unless that implementation is virtual), but all in all, I
   can live with it.
- The IDE is great; MSVC's type-ahead interface (or whatever it is
   called really saves a lot of typing an documentation lookup)
- The standard library is both huge and awfully, terrible, and a huge
   hack.
   There is API for nearly any aspect of the system, but many aspects
   are ill-designed, apparently in an ad-hoc manner - the entire library
   appears to be written under great time pressure. The API is often
   counter-intuitive and does not provide features that people would
   expect, as very similar features are provided. For example, it has
   an interface API for relational databases, but it is not possible
   to write an application that is agnostic of the specific interface
   implementation, as the application always needs to create instances
   of specific implementation classes - they forgot to provide central
   factory functions.

Regards,
Martin



More information about the Python-list mailing list