[IronPython] Jim Hugunin's web log.

[David Wilson]

For anyone who is excited about IronPython and it's consequences, you
might find Jim Hugunin's web log to be of particular interest. I didn't
see an announcement for this anywhere so here it is:

http://blogs.msdn.com/hugunin/
http://blogs.msdn.com/hugunin/Rss.aspx

Anti-trust conspiracy theories aside, if Microsoft adopts Python as a
part of it's development toolset, the repercussions for Windows RAD and
scripting are massive, not to mention the effects it would have on the
average Python developer's wages and availability of work.

It would also greatly ease the need to advocate Python in places of
employment, as a Microsoft adopted product, it couldn't be wrong(tm).
This is perhaps the first time where I have been glad to see Microsoft
hijack something. Even if IronPython becomes a commercial offering, I'm
still sold.

Other random thoughts: my experiences of the Python community versus,
eg., the perl community make me believe that Pythonistas are generally
more accepting of commercial solutions than their open source weenie
perl counterparts (*duck*). I'm still unsure as to whether or not this
should be considered a Microsoft marketing strategy for making in-roads
into the "open source scripting market".

My apologies for the poor terminology, this really isn't my department.
Just got a feeling. :)
David.

[Peter Hansen]

David Wilson wrote:

It would also greatly ease the need to advocate Python in places of
employment, as a Microsoft adopted product, it couldn't be wrong(tm).
This is perhaps the first time where I have been glad to see Microsoft
hijack something. Even if IronPython becomes a commercial offering, I'm
still sold.

Will you still be "sold" if Microsoft embraces and mutates (or
whatever is the pat phrase describing their amoebic methods)
Python by introducing a half dozen incompatible new forms of
syntax and such, confusing newcomers with market mumbo jumbo
and vapourware announcements, stealing away some of the best
minds from the open source community with obscene wages, and
all those other little "couldn't be wrong(tm)" <wink> things
that they have done in the past?

This might be somewhat like what one of our (Canadian) past
Prime Ministers referred to as "sleeping with an elephant" in
reference to living next to the US. "No matter how friendly and
even-tempered the beast, one is affected by every twitch and grunt."

(See http://en.wikiquote.org/wiki/Pierre_Trudeau for a possibly
accurate quote.)

-Peter

[Martin v. Löwis]

David Wilson wrote:

Other random thoughts: my experiences of the Python community versus,
eg., the perl community make me believe that Pythonistas are generally
more accepting of commercial solutions than their open source weenie
perl counterparts (*duck*).

I don't think this is an accurate description. Instead, Python is not
tied so much in Unix as Perl is (IMO). It is not Python's philosophy
to make all platforms look alike, but rather to expose all features of
a platform to the Python programmer - whether this is /dev/tty on Unix
or the registry on Windows. For features where it makes sense, a common
interface is established; other features are by nature restricted to
a single platform.

As a result of that philosophy, people are often tempted to port Python
to "strange" platforms (be that Mac OS 9, BeOS, VMS, or the JVM). They
then found that Python maintainers where open to changes resulting from
these ports as long no harm was done to Python "proper" (laissez-faire);
this continues to encourage people to experiment with the language, and
with various platforms.

Regards,
Martin

[Donn Cave]

In article <41**************@v.loewis.de>,
"Martin v. Lowis" <ma****@v.loewis.de> wrote:

David Wilson wrote:

Other random thoughts: my experiences of the Python community versus,
eg., the perl community make me believe that Pythonistas are generally
more accepting of commercial solutions than their open source weenie
perl counterparts (*duck*).

I don't think this is an accurate description. Instead, Python is not
tied so much in Unix as Perl is (IMO). It is not Python's philosophy
to make all platforms look alike, but rather to expose all features of
a platform to the Python programmer - whether this is /dev/tty on Unix
or the registry on Windows. For features where it makes sense, a common
interface is established; other features are by nature restricted to
a single platform.

As a result of that philosophy, people are often tempted to port Python
to "strange" platforms (be that Mac OS 9, BeOS, VMS, or the JVM). They
then found that Python maintainers where open to changes resulting from
these ports as long no harm was done to Python "proper" (laissez-faire);
this continues to encourage people to experiment with the language, and
with various platforms.

I'll go along with that, it's sure a positive factor for me -
sort of like the NetBSD of programming languages. And to my
way of thinking it's essentially an open source phenomenon.
Python does generally get there before Perl, but it also gets
there before ... Smalltalk? Mathematica? Kind of stumped for
closed source Python competitors. Perl has its own problems
(as does Tcl: "Tk".) I would propose GNU C as an example of
open source software that manages to take root in strange soil.
Written partly by Stallman himself if I remember right.

It might be true that the Python community is relatively free
of stridently ideological open source advocacy, but if Python
had not been open source, none of us would have ever heard of it.
Whether or not it makes sense for all software to be open source,
when it comes to "middle ware" like compilers and interpreters
it makes an awful lot of sense.

Donn Cave, do**@u.washington.edu

PS: Yes I know there has been an open source Smalltalk for years,
but for a long time it seemed like Smalltalk was dominated
by commercial, closed source implementations, and GNU Smalltalk
was a footnote. At that time it seemed that that language
had some promising traction in business applications, and I
guess it's a miracle there's no Microsoft Visual Smalltalk.
Meanwhile, Smalltalk hasn't exactly taken over the world, and
Python has emerged from its relative obscurity at the time.

[flacco]

David Wilson wrote:

Anti-trust conspiracy theories aside, if Microsoft adopts Python as a
part of it's development toolset, the repercussions for Windows RAD and
scripting are massive, not to mention the effects it would have on the
average Python developer's wages and availability of work.
why in the world would we want to put aside the concerns about microsoft
that we have, which you have "misnomered" under the umbrella term
"anti-trust conspiracy theories"?

This is perhaps the first time where I have been glad to see Microsoft
hijack something. Even if IronPython becomes a commercial offering, I'm
still sold. [...] My apologies for the poor terminology, this really isn't my department.
Just got a feeling. :)

are you in advertising by any chance?

[A.M. Kuchling]

On Mon, 30 Aug 2004 20:11:51 +0100,
David Wilson <[e-mail]> wrote:

Anti-trust conspiracy theories aside, if Microsoft adopts Python as a
part of its development toolset, the repercussions for Windows RAD and
scripting are massive...

Hugunin has said his job is to "make the CLR better for dynamic languages".
It's unclear to me that directive is synonymous with "develop IronPython
further"; he could invent his own new language and implement that, or work
with other CLR-based projects. I wouldn't conclude that Microsoft is
adopting Python as a supported tool, or even that Python is Hugunin's
primary focus, until he actually says that's his goal.

--amk

[yaipa h.]

flacco <fl*******@spambadTwilight-systems.com> wrote in message news:<10*************@corp.supernews.com>...

David Wilson wrote:

Anti-trust conspiracy theories aside, if Microsoft adopts Python as a
part of it's development toolset, the repercussions for Windows RAD and
scripting are massive, not to mention the effects it would have on the
average Python developer's wages and availability of work.

why in the world would we want to put aside the concerns about microsoft
that we have, which you have "misnomered" under the umbrella term
"anti-trust conspiracy theories"?

This is perhaps the first time where I have been glad to see Microsoft
hijack something. Even if IronPython becomes a commercial offering, I'm
still sold.

[...]

My apologies for the poor terminology, this really isn't my department.
Just got a feeling. :)

are you in advertising by any chance?

Hey, if Python can reuse the .net GUI builder... Cool.
Give something, get something back. So it goes...

Cheers,

--Al

[Neuruss]

> Hugunin has said his job is to "make the CLR better for dynamic languages".

It's unclear to me that directive is synonymous with "develop IronPython
further"; he could invent his own new language and implement that, or work
with other CLR-based projects. I wouldn't conclude that Microsoft is
adopting Python as a supported tool, or even that Python is Hugunin's
primary focus, until he actually says that's his goal.

I agree. I think the primary focus of Microsoft is to make the CLR the
best and most attractive development platform, and a good way to
achieve this is by making it more friendly to as many languages as
possible, including scripting ones.
That's the reason they hired Jim Hugunin. Not only to develop python
for the CLR, but to improve the CLR for all the scripting languages
and to investigate the best ways to implement these languages for the
CLR.

As for the concerns about Microsoft "hijacking" python, I think they
probable go too far. We all know what MS did in the past, but this
time, I think they shifted their policies drastically.

There's a whole move of opening their source and submitting standards
that confirm that their new strategy is stabilishing .NET as "the"
platform.
In this task, open source developers have an important place.

Regarding Jim Hugunin, he already implied in his comments that his
intentions are to impement python as close to the standard
implementation as posible. He even consults Gudo regularly on this
subject. And seeing what he did with Jython, there's no reason to
disbelieve.

[@(remove)yahoo.it]

David Wilson ha scritto:

For anyone who is excited about IronPython and it's consequences, you
might find Jim Hugunin's web log to be of particular interest. I didn't
see an announcement for this anywhere so here it is:

nice to hear, thanks
Other random thoughts: my experiences of the Python community versus,
eg., the perl community make me believe that Pythonistas are generally
more accepting of commercial solutions than their open source weenie
perl counterparts (*duck*). I'm still unsure as to whether or not this
should be considered a Microsoft marketing strategy for making in-roads
into the "open source scripting market".

Imo MS has always played with scripting languages.
The Shares Source CLI used perl.exe in its build process (maybe it still
does), and IIRC MS owns part of ActiveState. Plus, in a presentation of
MSH/Monad/"the new cmd.exe" they mentioned it being 'as powerful as perl
,python or ruby'.

[John Marshall]

On Tue, 2004-08-31 at 07:46, Neuruss wrote:

As for the concerns about Microsoft "hijacking" python, I think they
probable go too far. We all know what MS did in the past, but this
time, I think they shifted their policies drastically.

There's a whole move of opening their source and submitting standards
that confirm that their new strategy is stabilishing .NET as "the"
platform.

I'll believe they have changed after a decade of proof.

Skeptical.

John

[Baalbek]

Neuruss wrote:

As for the concerns about Microsoft "hijacking" python, I think they
probable go too far. We all know what MS did in the past, but this
time, I think they shifted their policies drastically.

Yeah, sure, after more than 25 years (since Microsoft's inception) of
foul play, they have "shifted their policies drastically".

The latest statements from their PR dudes must be the proof of this!

How naive are you?

QQQ

[Neuruss]

> How naive are you?

I don't think I'm naive. I just don't suffer paranoia.

Regarding Ironpython, I just love to see it alive.
I love python, and I want to see it running everywhere. What's the
problem with python in the CLR? None! One place more to see it
running!

And if you don't like Microsoft, or you just hate it, or if you're one
of those guys who think that Redmond is ploting to destroy the world,
well, you still can use Cpython or Jython. You choose, it's up to you.

I just think that, being Microsoft and .NET (and now Mono and soon
dotgnu) so ubiquitous, it wont harm Python if someone creates an
implementation to take full advantage of this framework.

Hugunin already did an excellent job with Jython, and it's hard to
believe that while working for the evil empire he will come up with
the seed of destruction for python...

[Baalbek]

Neuruss wrote:

How naive are you?

I don't think I'm naive. I just don't suffer paranoia.

Regarding Ironpython, I just love to see it alive.
I love python, and I want to see it running everywhere. What's the
problem with python in the CLR? None! One place more to see it
running!

Sorry, it came of more harshly than I intended.

I only reacted to "this time, I think they shifted their policies
drastically", which has been believed time and again, and one has lived
to regret trusting Microsoft, as far as I know.

As far as your other statements, no problems at all with me, I too would
like to see Python going "where no other programming language has dared
to go" :-)

[Neuruss]

> Sorry, it came of more harshly than I intended.

Don't worry. I'm the bully here ;-)