Groklaw says "Watch out, Ruby!"

ruby-talk
21

Ari Brown wrote:
> Does any know any good instructions for installing IronRuby on Mac OS X?
There are only just working instructions for building it on Windows :slight_smile:
  It may in time be supported by Mono, but the DLR is still a moving
target so I wouldn't expect that just yet.

> Or of any good (easy to use) compiler for Ruby?
IronRuby will be the first, assuming all is as it appears. At least, to
the best of my knowledge. I may be spouting gibberish - I've been
writing c# for 16 hours,

Good then it should still be possible to stop, now <ducking>

···

On 7/31/07, Alex Young <alex@blackkettle.org> wrote:

so all bets are off :slight_smile:

--
Alex

--
[...] as simple as possible, but no simpler.
-- Attributed to Albert Einstein

22

SonOfLilit wrote:

With all due to respect to John Lam and the other IronRuby
developers, is anyone taking IronRuby all that seriously?

It'll probably be the first "ruby" compiler to reach 1.0 and it allows
one to replace ActionScript with Ruby. Or so I hope.

You must mean "after jruby":

No, I don't. jruby is an interpreter.

JRuby is a mixed-mode interpreter and compiler (to JVM bytecode) similar to how HotSpot works. The compiler doesn't compile all Ruby syntax, but it compiles more every day. What it can compile, it does compile in a JIT fashion as the code runs.

It's likely we'll have a 100% complete compiler to bytecode along with the current "very high" level of compatibility before IronRuby achieves either, but I'm the only one working on the compiler at the moment.

- Charlie

···

On 7/30/07, Ben Bleything <ben@bleything.net> wrote:

On Tue, Jul 31, 2007, SonOfLilit wrote:

23

Oh, he can't use the implementations because I think the license is
incompatible :-/

···

On 7/30/07, Alex Young <alex@blackkettle.org> wrote:

Gregory Brown wrote:
> On 7/30/07, Alex Young <alex@blackkettle.org> wrote:
>
>> Another question: is there any reason that, say, Rubinius' standard
>> libraries couldn't be used with IronRuby? Has anyone looked? What's
>> their implementation status at the moment?
>
> The MRI stdlibs are (mostly) Ruby and should be freely reusable. I'd
> be sort of surprised if alternative implementers bother implementing
> more than they have to with these, unless they need to tweak the
> existing libs for their implementations.
>
I'm sure you're right, I can't see many technical reasons not to use the
standard implementation. However, John Lam is specifically asking for
contributions to the IronRuby stdlib. I don't know what assignments
you'd have to make to get them accepted, and I haven't seen it discussed
anywhere. John, are you reading? Can you give us a steer on this?

24

I'm sure you're right, I can't see many technical reasons not to use the
standard implementation. However, John Lam is specifically asking for
contributions to the IronRuby stdlib. I don't know what assignments
you'd have to make to get them accepted, and I haven't seen it discussed
anywhere. John, are you reading? Can you give us a steer on this?

There are quite a few stdlib libraries that are implemented in C - those are the ones that we need help with. We're making good progress on the built-in types (another dev on our team just checked in a nearly complete implementation for Hash yesterday).

As for assignments, we will have a standard contributor agreement that will assert that you have the legal right to contribute the code that you are submitting, as well as a copyright assignment. This is a very standard process in open source projects (see Apache's contributor agreement here: http://apache.org/licenses/icla.txt\).

Thanks,
-John

25

Alex Young wrote:

Robert Dober wrote:

Ari Brown wrote:

Does any know any good instructions for installing IronRuby on Mac
OS X?

There are only just working instructions for building it on Windows :slight_smile:
  It may in time be supported by Mono, but the DLR is still a moving
target so I wouldn't expect that just yet.

Or of any good (easy to use) compiler for Ruby?

IronRuby will be the first, assuming all is as it appears. At least, to
the best of my knowledge. I may be spouting gibberish - I've been
writing c# for 16 hours,

Good then it should still be possible to stop, now <ducking>

It's only by having ruby-talk on one monitor and Visual Studio on the
other that I'm staying sane :slight_smile:

Monitors? You work with lizards? How cool is that??

:slight_smile:

···

On 7/31/07, Alex Young <alex@blackkettle.org> wrote:

26

Gregory Brown wrote:

···

On 7/30/07, Alex Young <alex@blackkettle.org> wrote:

Gregory Brown wrote:

On 7/30/07, Alex Young <alex@blackkettle.org> wrote:

Another question: is there any reason that, say, Rubinius' standard
libraries couldn't be used with IronRuby? Has anyone looked? What's
their implementation status at the moment?

The MRI stdlibs are (mostly) Ruby and should be freely reusable. I'd
be sort of surprised if alternative implementers bother implementing
more than they have to with these, unless they need to tweak the
existing libs for their implementations.

I'm sure you're right, I can't see many technical reasons not to use the
standard implementation. However, John Lam is specifically asking for
contributions to the IronRuby stdlib. I don't know what assignments
you'd have to make to get them accepted, and I haven't seen it discussed
anywhere. John, are you reading? Can you give us a steer on this?

Oh, he can't use the implementations because I think the license is
incompatible :-/

That's why I mentioned rubinius - it's BSD-licensed, isn't it?

--
Alex

27

John Lam (CLR) wrote:

I'm sure you're right, I can't see many technical reasons not to
use the standard implementation. However, John Lam is specifically
asking for contributions to the IronRuby stdlib. I don't know what
assignments you'd have to make to get them accepted, and I haven't
seen it discussed anywhere. John, are you reading? Can you give
us a steer on this?

There are quite a few stdlib libraries that are implemented in C -
those are the ones that we need help with. We're making good progress
on the built-in types (another dev on our team just checked in a
nearly complete implementation for Hash yesterday).

There's a current discussion on ruby-core about packaging stdlib as gems - have you looked into that at all, or is that just not on the radar?

···

--
Alex

28

Actually, there are many differences between the MS-PL
and the *real* open source licenses approved by OSI,
as pj@groklaw has pointed out in further detail

http://www.groklaw.net/article.php?story=20070730120109643

Would you want to ellaborate more on this ?
Hopefully the dialog will result in Microsoft moving closer to
what open source really stands for.

Regardless of the outcome, thanks John for your outstanding
work reaching out to the community.

···

On 30/07/07, John Lam (CLR) <jflam@microsoft.com> wrote:

> I'm sure you're right, I can't see many technical reasons not to use the
> standard implementation. However, John Lam is specifically asking for
> contributions to the IronRuby stdlib. I don't know what assignments
> you'd have to make to get them accepted, and I haven't seen it discussed
> anywhere. John, are you reading? Can you give us a steer on this?

There are quite a few stdlib libraries that are implemented in C - those are the ones that we need help with. We're making good progress on the built-in types (another dev on our team just checked in a nearly complete implementation for Hash yesterday).

As for assignments, we will have a standard contributor agreement that will assert that you have the legal right to contribute the code that you are submitting, as well as a copyright assignment. This is a very standard process in open source projects (see Apache's contributor agreement here: http://apache.org/licenses/icla.txt\).

Thanks,
-John

--
best,
                        UG
---
Uma Geller

29

What I'm saying is that my guess is that Rubinius would use the MRI
implementations rather than wasting time building them from scratch.
MIT License is compatible with the License of Ruby + GPL, i think
IronRuby's license isn't.

When you use parts of a project that are under a different license,
you usually can't change the license terms without permission. Which
means the Rubinius standard library would be under License of Ruby,
not MIT, if they use the MRI implementations.

···

On 7/30/07, Alex Young <alex@blackkettle.org> wrote:

Gregory Brown wrote:
> On 7/30/07, Alex Young <alex@blackkettle.org> wrote:
>> Gregory Brown wrote:
>>> On 7/30/07, Alex Young <alex@blackkettle.org> wrote:
>>>
>>>> Another question: is there any reason that, say, Rubinius' standard
>>>> libraries couldn't be used with IronRuby? Has anyone looked? What's
>>>> their implementation status at the moment?
>>> The MRI stdlibs are (mostly) Ruby and should be freely reusable. I'd
>>> be sort of surprised if alternative implementers bother implementing
>>> more than they have to with these, unless they need to tweak the
>>> existing libs for their implementations.
>>>
>> I'm sure you're right, I can't see many technical reasons not to use the
>> standard implementation. However, John Lam is specifically asking for
>> contributions to the IronRuby stdlib. I don't know what assignments
>> you'd have to make to get them accepted, and I haven't seen it discussed
>> anywhere. John, are you reading? Can you give us a steer on this?
>
> Oh, he can't use the implementations because I think the license is
> incompatible :-/
>
That's why I mentioned rubinius - it's BSD-licensed, isn't it?

30

Uma Geller wrote:

Actually, there are many differences between the MS-PL
and the *real* open source licenses approved by OSI,
as pj@groklaw has pointed out in further detail

Groklaw - Uh Oh. Another Smooth Move from Microsoft: Watch out, Ruby. Watch out OSI. -- Updated

Would you want to ellaborate more on this ?
Hopefully the dialog will result in Microsoft moving closer to
what open source really stands for.

Regardless of the outcome, thanks John for your outstanding
work reaching out to the community.

Groklaw is saying that real open source licenses obligate licensees to
distribute source code of derived works IF they distribute binaries. OK.
Ms-PL doesn't do that.

But how does the MIT license require this? There is no wording in there
anywhere that looks anything like that, but MIT is in the OSI license
database.

Groklaw also says that Ms-PL allows licensees to charge for modified
works. Well, the MIT license specifically allows fees as well.

It seems that Ms-PL is as much an open source license as the MIT
license. Those MIT bastards!

···

------------------------

The MIT License

Copyright (c) <year> <copyright holders>

Permission is hereby granted, free of charge, to any person obtaining a
copy
of this software and associated documentation files (the "Software"), to
deal
in the Software without restriction, including without limitation the
rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or
sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included
in
all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
IN
THE SOFTWARE.

-------------------------

best,
Dan
--
Posted via http://www.ruby-forum.com/\.

31

Actually, there are many differences between the MS-PL
and the *real* open source licenses approved by OSI,
as pj@groklaw has pointed out in further detail

Groklaw - Uh Oh. Another Smooth Move from Microsoft: Watch out, Ruby. Watch out OSI. -- Updated

Would you want to ellaborate more on this ?

OSI approves more than just GPL. Unfortunately groklaw believes that open source == GPL which is clearly not the case.

As a more concrete example, the Ms-PL allows Novell to redistribute the DLR however they wish, which is something that Miguel has already publicly suggested that they will do.

It also allows any of you to redistribute DLR or IronRuby or IronPython however you wish. You could do so freely, you could charge for the distribution etc. You just have to abide by the conditions imposed by the license.

As for *interpreting* those conditions, please make sure that you consult with your friendly neighborhood attorney before making any business decisions around this. Unfortunately, email threads on RubyTalk and blog posts / comments on the internet don't count as legal advice :slight_smile:

Thanks,
-John

32

Rubinius are building ruby implementations of the part of the standard
library that is implemented in C, AFAIK.

Do they also reimplement the ruby part?

Aur

33

Gregory Brown wrote:

Gregory Brown wrote:

Gregory Brown wrote:

Another question: is there any reason that, say, Rubinius' standard
libraries couldn't be used with IronRuby? Has anyone looked? What's
their implementation status at the moment?

The MRI stdlibs are (mostly) Ruby and should be freely reusable. I'd
be sort of surprised if alternative implementers bother implementing
more than they have to with these, unless they need to tweak the
existing libs for their implementations.

I'm sure you're right, I can't see many technical reasons not to use the
standard implementation. However, John Lam is specifically asking for
contributions to the IronRuby stdlib. I don't know what assignments
you'd have to make to get them accepted, and I haven't seen it discussed
anywhere. John, are you reading? Can you give us a steer on this?

Oh, he can't use the implementations because I think the license is
incompatible :-/

That's why I mentioned rubinius - it's BSD-licensed, isn't it?

What I'm saying is that my guess is that Rubinius would use the MRI
implementations rather than wasting time building them from scratch.
MIT License is compatible with the License of Ruby + GPL, i think
IronRuby's license isn't.

Taking another look at the Rubinius project page, it seems that they have, indeed, imported the 1.8 MRI stdlib. I need more coffee.

When you use parts of a project that are under a different license,
you usually can't change the license terms without permission. Which
means the Rubinius standard library would be under License of Ruby,
not MIT, if they use the MRI implementations.

Yup, agreed - this tangent was entirely inspired by the false assumption (on my part) that for some reason there was a large part of stdlib being recoded as part of the rubinius project.

···

On 7/30/07, Alex Young <alex@blackkettle.org> wrote:

On 7/30/07, Alex Young <alex@blackkettle.org> wrote:

On 7/30/07, Alex Young <alex@blackkettle.org> wrote:

--
Alex

34

Daniel Lucraft wrote:

It seems that Ms-PL is as much an open source license as the MIT
license. Those MIT bastards!

The Apache License 2.0 too. "You may reproduce and distribute copies of
the Work or Derivative Works ... in Source or Object form" and no
obligation to redistribute the source code of a derived work is
mentioned.

Those Apache bastards! Truly the Open Source Community is under a
greater threat than we realized.

best,
Dan

···

--
Posted via http://www.ruby-forum.com/\.

35

Argh! Why people think that forced source distribution is some how "the
only way to be free" is beyond me.

···

On Fri, Aug 03, 2007 at 05:14:32PM +0900, Daniel Lucraft wrote:

Uma Geller wrote:
> Actually, there are many differences between the MS-PL
> and the *real* open source licenses approved by OSI,
> as pj@groklaw has pointed out in further detail
>
> Groklaw - Uh Oh. Another Smooth Move from Microsoft: Watch out, Ruby. Watch out OSI. -- Updated
>
> Would you want to ellaborate more on this ?
> Hopefully the dialog will result in Microsoft moving closer to
> what open source really stands for.
>
> Regardless of the outcome, thanks John for your outstanding
> work reaching out to the community.

Groklaw is saying that real open source licenses obligate licensees to
distribute source code of derived works IF they distribute binaries. OK.
Ms-PL doesn't do that.

But how does the MIT license require this? There is no wording in there
anywhere that looks anything like that, but MIT is in the OSI license
database.

Groklaw also says that Ms-PL allows licensees to charge for modified
works. Well, the MIT license specifically allows fees as well.

It seems that Ms-PL is as much an open source license as the MIT
license. Those MIT bastards!

--
CCD CopyWrite Chad Perrin [ http://ccd.apotheon.org ]
John Kenneth Galbraith: "If all else fails, immortality can always be
assured through spectacular error."

36

John Lam (CLR) wrote:

As for *interpreting* those conditions, please make sure that you
consult with your friendly neighborhood attorney before making any
business decisions around this. Unfortunately, email threads on RubyTalk
and blog posts / comments on the internet don't count as legal advice :slight_smile:

Doesn't your saying that we now need to get lawyers to tell us what we
can and cannot do sort of tell us that the source is not open anymore?
I thought that the idea was to relax and share openly, hence the term.
If we have to cower and run to lawyers, who cannot tell us the right or
wrong of it but only whether they can defend it when some other lawyer
is paid to fight for the other team, then why would we want that around
anyway? Should open source be open source?

This is taking a most disheartening turn.

···

--
Posted via http://www.ruby-forum.com/\.

37

>> That's why I mentioned rubinius - it's BSD-licensed, isn't it?

> What I'm saying is that my guess is that Rubinius would use the MRI
> implementations rather than wasting time building them from scratch.
> MIT License is compatible with the License of Ruby + GPL, i think
> IronRuby's license isn't.

Taking another look at the Rubinius project page, it seems that they
have, indeed, imported the 1.8 MRI stdlib. I need more coffee.

> When you use parts of a project that are under a different license,
> you usually can't change the license terms without permission. Which
> means the Rubinius standard library would be under License of Ruby,
> not MIT, if they use the MRI implementations.

Yup, agreed - this tangent was entirely inspired by the false assumption
(on my part) that for some reason there was a large part of stdlib being
recoded as part of the rubinius project.

To set the record straight:

1) Rubinius is BSD licensed.
2) There is a mixup in terminology here. We are not recoding the
standard library (stdlib). We are recoding the core library. The core
library is all methods on Array, Hash, String, etc. everything that is
available in ruby without a require. All of that code in MRI is in C,
and 95% of ours is in ruby.
3) I'd love to see IronRuby using our core library rather than
recoding it themselves. The JRuby guys and I have talked about that
very thing.

- Evan Phoenix

···

On Jul 30, 3:24 pm, Alex Young <a...@blackkettle.org> wrote:

--
Alex

38

Doesn't your saying that we now need to get lawyers to tell us what we
can and cannot do sort of tell us that the source is not open anymore?
I thought that the idea was to relax and share openly, hence the term.

You are certainly free to read the license yourself. It's very short and simple. I just wanted to make it clear that *my* interpretation doesn't matter, and neither does anyone else's if you happen to violate a term of the license based on a misinterpretation on your part. But if you feel comfortable in understanding the terms of your license without consulting your own lawyer then that's a risk that you are certainly welcome to take.

Consulting a lawyer is something you should do regardless of the license. The more complex ones eg GPL are things that you should look at very closely to make sure you understand the implications of the license on your business.

BTW, MsPL has and continues to be described as a BSD-style license by virtually anyone who has actually read the two licenses and compared them side-by-side.

-John

39

John Lam (CLR) wrote:
> As for *interpreting* those conditions, please make sure that you
> consult with your friendly neighborhood attorney before making any
> business decisions around this. Unfortunately, email threads on RubyTalk
> and blog posts / comments on the internet don't count as legal advice :slight_smile:

Doesn't your saying that we now need to get lawyers to tell us what we
can and cannot do sort of tell us that the source is not open anymore?
I thought that the idea was to relax and share openly, hence the term.
If we have to cower and run to lawyers, who cannot tell us the right or
wrong of it but only whether they can defend it when some other lawyer
is paid to fight for the other team, then why would we want that around
anyway? Should open source be open source?

This is taking a most disheartening turn.

You are so right, that is why having some framework like OSI approval
might help in relaxing a bit more...
But I guess John is giving sound advice for this low world.

Concerning the MS license I can only say "Timeo danaos et donas
ferentes" and yes I now exactly what happened to the guy who said that
first, funny to see Bill Gates as Ulysses, but my horse sense is
telling me so...

Cheers
Robert

···

On 8/3/07, Lloyd Linklater <lloyd@2live4.com> wrote:

--
Posted via http://www.ruby-forum.com/\.

--
[...] as simple as possible, but no simpler.
-- Attributed to Albert Einstein

40

Whoops, I somehow threw MIT in there in place of BSD, but what I said
holds the same for BSD. :slight_smile:

···

On 7/31/07, evanwebb@gmail.com <evanwebb@gmail.com> wrote:

On Jul 30, 3:24 pm, Alex Young <a...@blackkettle.org> wrote:
> >> That's why I mentioned rubinius - it's BSD-licensed, isn't it?
>
> > What I'm saying is that my guess is that Rubinius would use the MRI
> > implementations rather than wasting time building them from scratch.
> > MIT License is compatible with the License of Ruby + GPL, i think
> > IronRuby's license isn't.
>
> Taking another look at the Rubinius project page, it seems that they
> have, indeed, imported the 1.8 MRI stdlib. I need more coffee.
>
> > When you use parts of a project that are under a different license,
> > you usually can't change the license terms without permission. Which
> > means the Rubinius standard library would be under License of Ruby,
> > not MIT, if they use the MRI implementations.
>
> Yup, agreed - this tangent was entirely inspired by the false assumption
> (on my part) that for some reason there was a large part of stdlib being
> recoded as part of the rubinius project.
>

To set the record straight:

1) Rubinius is BSD licensed.