[IronPython] If not IronPython, what?
tristanz at gmail.com
Mon Aug 23 00:22:46 CEST 2010
The popularity of the JVM and the CLR suggest not everybody agrees with your
assessment. IronPython provides a compelling interoperability story for
some applications. I use IronPython and F# on Mac and Linux, so I'm not
certain where the Windows requirement comes from. I happily use CPython as
I do encourage Microsoft to clarify its position on DLR and Iron* support
sooner than later. Given the dominance of dynamic languages such as Python,
Perl, R, and Matlab as the glue language in science, I'm skeptical that
Microsoft's new technical computing initiative will succeed without some
dynamic language support. I'm hopeful that the NumPy announcement means
IronPython will get support as an infrastructure component in the technical
On Sun, Aug 22, 2010 at 4:57 PM, Vernon Cole <vernondcole at gmail.com> wrote:
> Sorry for chiming in late. I've been on vacation, and since my "smart"
> phone runs Windows Mobile, it does not work very well, so I have not kept up
> on my email.
> When I first started in this industry, there where two groups of computer
> users: IBM and everybody else. I was in the "everybody else" group, and
> developed a deep disrespect for 'Big Blue". There were lots of people who
> were always willing to believe that anything IBM did was the greatest thing
> in the world and everything else was trash. "We are an IBM shop" was the
> excuse for a lot of what looked -- to me -- like mismanagement and waste..
> Today IBM has faded into a niche, and Microsoft is the 900 pound gorilla in
> the computing cage. There are those today who chant "WE ARE A MICROSOFT
> SHOP" as if it were a mantra embodying all good, and all else is evil. If
> you work in such a place I am sorry for you. I have worked in such places
> and I am glad I am out. By all means USE IronPython, it's the best tool in
> your shop.
> But let's set the record straight -- at least my humble opinion of the
> record -- dotNET was a stupid idea to start with. Why invent a new
> pseudo-machine to compile everything in, when 99% of the world is using the
> same hardware machine anyway? Pseudo machines were great when 32 K bytes
> was a large computer, by why now? The "just in time" compiler should be
> called a "waste my time" compiler. If you think I'm wrong, try running
> "hello world" in CPython and IronPython and see which one is faster.
> _Anything_ running on CLR starts up like as sleepy teenager. [And why oh
> why did they let the sales department stick it with a name which conflicts
> with an Internet domain? But I digress....]
> Do I support IronPython? Yes, I do. Because some poor saps are stuck
> working in "WE ARE A ... SHOP." places. Do I personally use IronPython for
> production software? Nope! I use CPython -- on Linux where possible and on
> Windows where necessary. The concept of "scientific data acquisition" using
> Windows is an oxymoron. *
> I strongly encourage you to continue to look at alternatives which were
> _not_ made in Redmond, Washington, USA. If nothing else, the competition
> pushes Microsoft to continue to improve the quality of their products.
> IronPython is the greatest thing which has happened in Redmond since NT was
> Vernon Cole
> * this e-mail is being written on an Ubuntu Linux box which is also running
> a closed-loop environmental control system (written in Python) while I type.
> On Wed, Aug 18, 2010 at 11:41 AM, Max Yaffe <maxyaffe at gmail.com> wrote:
>> I've been reading the buzz around Microsoft's reduced commitment to Iron*
>> languages and wondering if I should rethink my own commitment to
>> To fill you in, I'm a designer of instruments and software for scientific
>> data acquisition and analysis. My current software uses a dynamic
>> for scripting in a Win32 based framework program for acquisition and VBA
>> scripting in a VB program for analysis. We decided to rewrite the whole
>> thing in C#+.Net4+IronPython.
>> On stackoverflow, someone asked about using Iron* vs PowerShell for
>> scripting in a C# application in light of Microsoft's changing
>> This was my answer:
>> > I'm in a similar position.
>> > I decided to use IronPython scripting but ever since I saw Anders
>> > talk "The Future of C#", I've had a feeling IronPython was doomed.
>> > It was in Microsoft's interest to get the DLR developed but they
>> > want us to use tools and languages they control. After all, aren't you
>> > C# and not Java? So what will a Microsoft dynamic language look like?
>> > about dynamic, interpreted C# (Iron C#)? Hejlsberg's talk made it clear
>> > isn't that far away. He even had a console window with a REPL interface.
>> > That said, there's always a possibility for Iron VB. Talk about closing
>> > loop.
>> > On the plus side for us programmers, Iron C# also solves another problem
>> > I'm having trouble with -- the existence of two parallel object
>> > one of .Net objects, one of Python objects. It takes work to get from
>> > the other. I assume an Iron C# would utilize the .Net class structure.
>> > My advice: Stick with Iron Python and .Net classes. When Iron VB or Iron
>> > happens, it'll be a quick, maybe automatic, language translation.
>> > if enough of us use IronPython, Microsoft may change their mindset.
>> So my question to you is a) am I thinking correctly about the future of
>> IronPython, and 2) if not IronPython, what scripting language should I be
>> considering for a .Net C# application? I should let you know I'm also
>> considering switching to Qt/PyQt/Cpython.
>> Thanks for your input.
>> Users mailing list
>> Users at lists.ironpython.com
> Users mailing list
> Users at lists.ironpython.com
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Ironpython-users