Joyent & Node

2,335 views
Skip to first unread message

Ryan Dahl

unread,
Nov 9, 2010, 6:53:50 PM11/9/10
to nodejs
It has become clear over the last few months that Node's popularity
can support a larger, more structured development team. Indeed, for
Node to achieve massive success -- for it to be the next PHP -- it
needs investment that can support a team of full-time developers and a
corporate entity behind it that can, for example, enter into
agreements with large companies who are keen to bring Node to their
projects and make contributions back to the project. Someone spending
time and resources on evangelism and stuff wouldn't hurt either.

Joyent's core business (cloud computing) aligns well with a free and
open Node.js. For these reasons I've made an agreement with Joyent
that officially puts Node under their roof. The only effective change
for developers is that the contributor agreement will be directed to
Joyent rather than myself. Node will continue regular releases under
the MIT license. As part of the agreement -- in addition to supporting
my own development on Node -- we've recently hired Isaac Schlueter and
plan to add at least one more full-time developer on the core project
immediately. As a business we will begin selling instances of our
minimalistic no.de Node hosting service in several weeks and soon
offer support services developing custom bindings to Node.

Contact me at ry...@joyent.com if you have any questions.

Stephen Belanger

unread,
Nov 9, 2010, 7:03:47 PM11/9/10
to nod...@googlegroups.com
It's great to hear that node.js is getting some serious corporate consideration. It's a great project and it would be a shame if, for whatever reason, you eventually moved on and the project was put on hold. Putting it into the hands of Joyent like this should ensure it's survival long after you've moved on to other things. Not that I expect that'll happen any time soon. ;)


--
You received this message because you are subscribed to the Google Groups "nodejs" group.
To post to this group, send email to nod...@googlegroups.com.
To unsubscribe from this group, send email to nodejs+un...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/nodejs?hl=en.


Evan Meagher

unread,
Nov 9, 2010, 7:22:15 PM11/9/10
to nod...@googlegroups.com
This is good to hear. Emerging platforms generally need at least one solid corporate backer to succeed in the grand scheme of things. Joyent's investment in Node will undoubtedly prove fruitful for everyone.

Thanks for the update!

On Tue, Nov 9, 2010 at 3:53 PM, Ryan Dahl <r...@tinyclouds.org> wrote:
--
You received this message because you are subscribed to the Google Groups "nodejs" group.
To post to this group, send email to nod...@googlegroups.com.
To unsubscribe from this group, send email to nodejs+un...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/nodejs?hl=en.




--
Evan Meagher

Joshua Kehn

unread,
Nov 9, 2010, 7:23:31 PM11/9/10
to nod...@googlegroups.com
That sounds great! I look forward to seeing Joynet and Node work together.

Regards,

-Josh
____________________________________
Joshua Kehn | Josh...@gmail.com

Mikeal Rogers

unread,
Nov 9, 2010, 7:57:57 PM11/9/10
to nod...@googlegroups.com
Just to be clear, the copyright assignment is transferring from you personally to Joyent, and the CLA will now require contributors to assign the copyright to Joyent, correct?

Might also want to make it clear to people that the MIT license doesn't really enable the kinds of backdoor proprietary licenses companies do with the the GPL when they have these kinds of assignments in their CLA.

-Mikeal

On Tue, Nov 9, 2010 at 3:53 PM, Ryan Dahl <r...@tinyclouds.org> wrote:

Ryan Dahl

unread,
Nov 9, 2010, 8:26:53 PM11/9/10
to nod...@googlegroups.com
On Tue, Nov 9, 2010 at 4:57 PM, Mikeal Rogers <mikeal...@gmail.com> wrote:
> Just to be clear, the copyright assignment is transferring from you
> personally to Joyent, and the CLA will now require contributors to assign
> the copyright to Joyent, correct?

Correct

> Might also want to make it clear to people that the MIT license doesn't
> really enable the kinds of backdoor proprietary licenses companies do with
> the the GPL when they have these kinds of assignments in their CLA.

That's right. Linking to Node does not require your software to also
be open source.

Diego Correa T.

unread,
Nov 9, 2010, 9:31:35 PM11/9/10
to nod...@googlegroups.com
This is an inmediatisc solution to a self-created problem.

The case of Pure Data (software) continues to be the most promisory
business approach for the NeXT decade.

(the NeXT is of my iPod)

Diego

Enviado desde mi iPod

Andrew Lunny

unread,
Nov 9, 2010, 10:49:08 PM11/9/10
to nod...@googlegroups.com
+1 Diego

--
Andrew Lunny
Software Developer, Nitobi
604 685 9287
blogs.nitobi.com/andrew

Preston Guillory

unread,
Nov 10, 2010, 12:54:00 AM11/10/10
to nod...@googlegroups.com
On Tue, Nov 9, 2010 at 5:53 PM, Ryan Dahl <r...@tinyclouds.org> wrote:
Joyent's core business (cloud computing) aligns well with a free and
open Node.js. For these reasons I've made an agreement with Joyent
that officially puts Node under their roof. The only effective change
for developers is that the contributor agreement will be directed to
Joyent rather than myself. Node will continue regular releases under
the MIT license. As part of the agreement -- in addition to supporting
my own development on Node -- we've recently hired Isaac Schlueter and
plan to add at least one more full-time developer on the core project
immediately. As a business we will begin selling instances of our
minimalistic no.de Node hosting service in several weeks and soon
offer support services developing custom bindings to Node.

This is mixed news.

On the one hand, as a member of the Node community, I want to offer congratulations both to Isaac and yourself.  I count myself as one of no doubt many who have come to love Node and respect the two of you (and others) for your contributions to its development.  I hope however the business and legal environment surrounding Node evolves, people such as yourself always are rewarded appropriately, financially and otherwise.

On the other hand, as someone making investments in the future of Node, I find its corporate ownership a little unsettling.  Java got pushed into my CS program in the 90s, and now it's in the headlines because of Oracle's lawsuits.  Zend has a habit of trying to sell features that ought to be part of libre PHP.  And Microsoft... how much does Visual Studio cost again?

I don't know anyone at Joyent, personally.  Are they good people?  What did they get on their end of this deal?  Who has final say in the language's development?  There are other Node hosting providers, like Heroku.  Should they not be worried about selling something in competition with its creator?

joelparkerhenderson

unread,
Nov 10, 2010, 1:53:40 AM11/10/10
to nodejs
> This is mixed news.

This is difficult news because corporate interests
seem to harm open software, especially recently.

There are major problems right now with MySQL because of this.
Sun did buggy releases, and now Oracle is shutting off options.

I can donate $1000 in matching funds toward funding a Joyent Node
competitor.

Tim Caswell

unread,
Nov 10, 2010, 2:12:29 AM11/10/10
to nod...@googlegroups.com
There is a big difference.  Node is MIT.  And other companies with power and interest in node could simply fork if Joyent were to act foolishly.





Sent from my Palm Pixi on AT&T


Arnout Kazemier

unread,
Nov 10, 2010, 2:45:42 AM11/10/10
to nod...@googlegroups.com
So just to be clear, the CLA will say that you have transferred over your copyright to Joyent.
So basically Joyent as company can do anything they want with our Node.js code because
the CLA does not specifically state that the code is only intended to be used for the Node.js
project?

joelparkerhenderson

unread,
Nov 10, 2010, 2:55:41 AM11/10/10
to nodejs


On Nov 9, 11:45 pm, Arnout Kazemier <i...@3rd-eden.com> wrote:
> So just to be clear, the CLA will say that you have transferred over your copyright to Joyent.
> So basically Joyent as company can do anything they want with our Node.js code because
> the CLA does not specifically state that the code is only intended to be used for the Node.js
> project?

Correct. The first line is "You grant us the ability to use the
Contributions in any way."
The CLA is clear that this includes Joyent selling your code, changing
its license to be
proprietary, or one-sidedly deciding to omit your code from Node then
sell your code or
incorporate your code into other proprietary Joyent projects.

Erik Corry

unread,
Nov 10, 2010, 3:06:00 AM11/10/10
to nod...@googlegroups.com
2010/11/10 joelparkerhenderson <joelparke...@gmail.com>:

I am not, of course, a lawyer, nor am I speaking for anyone but
myself, but I personally don't think this adds much to what Joyent (or
anyone else) could already do with node source code. The BSD license
doesn't preclude 'additional restrictions' like the GPL does. As far
as I know MicroSoft had no claims on the BSD networking code, but that
didn't stop them putting it in Windows. That code is still covered by
the BSD license as far as I know, but MS have added extra
restrictions: More licenses that also cover the same code and the code
it is aggregated with.

--
Erik Corry

Micheil Smith

unread,
Nov 10, 2010, 3:11:48 AM11/10/10
to nod...@googlegroups.com
It is with general respect that Joyent would not be likely to sell your
code, given it is a contribution to an open source project. It is incredibly
unlikely that Joyent will ever change the licensing of Node.js, which
would end up being an extremely costly and complicated process.

The CLA only covers code contributed to the Node.js project, in other words, what
you get when you checkout ry/node from github or when you download a tarball.

The CLA does NOT cover any modules that you contribute to the community via
things like NPM.

The CLA has been in place for the last few months, all that has changed is now
instead of transferring your IP and such to Ryan Dahl, you are now transferring
them to Joyent, the text of the CLA is the same otherwise.

A really good example I have of CLA's is the Dojo Toolkit project. You sign a
CLA to say that the code you contribute to the Dojo Toolkit is owned / intellectually
owned by the Dojo Foundation. The reason for this is to protect the project from
people contributing things to http://github.com/ry/node and then later contact ryan
or someone and saying that they need the code they contributed removed from the
project because it's actually owned by their own company. This is something that
actually happened in the early days of the jQuery UI project, and part of the reason
as to why jQuery UI didn't launch with a Menu/Toolbar widget back when it was
first released.

It would be bad and unfriendly developer practice for Joyent to use the code you
contribute to Node.js under different license for other purposes. The CLA is purely
something to safe guard the project from bad Intellectual Property rights and licensing,
which have plagued various other projects.

If you have contributed any code to Node.js over the past few months, you should have
signed the CLA any way. The only difference is that now, rather than handing your IP
to Ryan, you're handing it to Joyent.

Hopefully this clears some stuff up a little bit. There's no need to over react to the
change in the name on the CLA.

Yours,
Micheil Smith
--
BrandedCode.com

Mark Mayo

unread,
Nov 10, 2010, 3:28:08 AM11/10/10
to nodejs
On Nov 10, 12:06 am, Erik Corry <e...@arbat.com> wrote:
> 2010/11/10 joelparkerhenderson <joelparkerhender...@gmail.com>:
Correct. The point here is that Joyent, like anyone else who uses the
Node
source code in any way, is bound by the terms of the source code
license,
namely, MIT. If we were using the code in a proprietary product, we
still
have to respect that license. If you're contributing code, it's under
MIT as well,
and published in public git repos where it could be rescued/forked/
whatever
if Joyent, Palm/HP, Yahoo!, Ryan, or anyone else that runs a node fork
"went rogue". The community around node is strong, and comparing this
to the MySQL/Oracle situation is far fetched, IMO. No GPL and
associated
"dual licensing" legacy is involved for one, and of course Joyent is
hardly
the $31B behemoth that Oracle is..

The CLA for node is identical to the CLA used by V8 itself, btw. Ditto
with
the choice of the MIT open source license.

Regarding competition, of course there will be zillions of options for
hosting
node apps. Node isn't GAE or some magic platform that only Joyent can
run. Got access to a computer? Run your node app! We expect to compete
with Heroku, VMware, Google, Amazon and the myriad of hosting
providers
out there by offering a better *service*. It's really that simple.

For a little more background on why we decided to do this, I wrote a
little blog post:

http://joyeur.com/2010/11/09/a-new-abode-for-node/

At the end of the day, I believe we did the honourable, right thing
for
both Ryan and node.

Cheers,
-Mark (Joyent)

Micheil Smith

unread,
Nov 10, 2010, 3:39:44 AM11/10/10
to nod...@googlegroups.com, Mark Mayo
Mark:

As a contributor to Node.js and an active user, I too believe you did the
right and honourable thing. I guess people are just scared after the chaos
that the Java/Android folks had to experience due to Oracle.

Yours,
Micheil Smith
--
BrandedCode.com

Paul Querna

unread,
Nov 10, 2010, 3:48:32 AM11/10/10
to nod...@googlegroups.com

Yes, strictly about the CLA change, the real world impact for
contributing code is minimal.

I personally would of preferred the code be licensed to a non-profit /
foundation steward. As you mentioned, Dojo is a good example of a
project that does exactly this. Non-profits and foundations
motivations don't change as business needs change. They are there to
last decades.

My concerns for the community about the change are much more on the
branding side of Node.js, as a product of Joyent.

I certainly want a company to promote the growth of Node -- someone to
sit down with enterprises and get them interested in Node. Joyent
seems to be in a good position to promote Node.

However, I believe advocacy in the enterprise is separate from
'owning' the project. I am concerned about things like branding it
"Joyent Node.js" and the like. Sponsored by logos in the footer of
the website aren't a big deal, but it will be interesting to see how
Joyent chooses to promote Node and vice versa.

In the end I'm sure Joyent wants what is best for the community. A
successful Node.js community is good for them. But I am still worried
about the depth of the embrace -- as long as Node is developed openly
and refereed to as an open source project, and not a Joyent product, I
think it will be fine.

These are just my concerns.

Thanks,

Paul

Paul Querna

unread,
Nov 10, 2010, 3:58:42 AM11/10/10
to nod...@googlegroups.com

No.

The Google v8 CLA is here:
<http://code.google.com/legal/individual-cla-v1.0.html>

It is different than the Node.js CLA here:
<http://nodejs.org/cla.html>

Node is under the MIT license:
<https://github.com/ry/node/blob/master/LICENSE>

Google's v8 however is under the New BSD License:
<http://code.google.com/p/v8/source/browse/trunk/LICENSE>

Unfortunately, Google did not release v8 under the Apache 2.0 license,
which includes a patent grant -- if you look at most projects Google
releases they pick the Apache 2.0 license by default, unless there is
a different expectation in the community (ie, a Linux kernel driver is
under GPL). Their choice of BSD does make me wonder if they were
worried about the pantent minefield that is developing a VM. Anyways,
enough musing about patent grants in licensing....

Sorry for being pedantic about the licensing/clas.

Thanks,

Paul

Jorge

unread,
Nov 10, 2010, 4:12:39 AM11/10/10
to nod...@googlegroups.com
On 10/11/2010, at 09:28, Mark Mayo wrote:
> hardly
> the $31B behemoth that Oracle is..

144.48B as of today. http://www.google.com/finance?q=NASDAQ%3AORCL
--
Jorge.

Isaac Schlueter

unread,
Nov 10, 2010, 4:16:27 AM11/10/10
to nod...@googlegroups.com
Let's say, hypothetically, Joyent turns evil, and decides to
dual-license node with a GPL version and a proprietary for-pay
license. It's unlikely, but stranger things have happened.

The day that happens, anyone can just take Node, fork it, call it
"OpenNode", and keep right on trucking. Certainly, damage could be
done to the community, and it could erode the confidence in Node as a
platform.

That is just so wildly unlikely, though. Not because Joyent is so
lovey dovey about the node community (though we totally are), but
because it'd just be such a profoundly *stupid* move. Joyent is in a
position to make sure that node continues to grow in awesomeness, that
the community stays free and vibrant, and most importantly, that
*other* companies invest in it and jump on this bandwagon. It's in
the greedy self-interest of Joyent to do these things.

If you look at the other stuff the company's been involved with,
there's a clear track record of using open source software, being a
good open source citizen, and reaping rewards from it. It's not a
dumb bunch of people.

Joyent didn't steal Node away from us. We all recruited Joyent to
help take node to the next level.

--i

Erik Corry

unread,
Nov 10, 2010, 4:35:35 AM11/10/10
to nod...@googlegroups.com
2010/11/10 Mark Mayo <mark...@gmail.com>:

> Regarding competition, of course there will be zillions of options for
> hosting
> node apps. Node isn't GAE or some magic platform that only Joyent can
> run. Got access to a computer? Run your node app!

I don't speak for Google and I'm not going to get into any form of
discussion on this, but I will just link to this blog post which
relates to the issue Mark brought up:
http://googleappengine.blogspot.com/2010/10/research-project-appscale-at-university.html

--
Erik Corry

Paul Querna

unread,
Nov 10, 2010, 4:40:16 AM11/10/10
to nod...@googlegroups.com
On Wed, Nov 10, 2010 at 1:16 AM, Isaac Schlueter <i...@izs.me> wrote:
> Let's say, hypothetically, Joyent turns evil, and decides to
> dual-license node with a GPL version and a proprietary for-pay
> license.  It's unlikely, but stranger things have happened.

I'll try to keep my replies realistic. I don't think many statements
made in your original email were realistic.

Regardless of the actual risk of it happening, this is one reason open
source non-profit foundations of all shapes and sizes exist -- so it
can never happen.

> The day that happens, anyone can just take Node, fork it, call it
> "OpenNode", and keep right on trucking.  Certainly, damage could be
> done to the community, and it could erode the confidence in Node as a
> platform.

You could not call it OpenNode, as a fork, the holder of the Copyright
and License would most likely enforce this as a confusing product
name, and it would need to be renamed something that didn't keep the
"node" name. See also, OpenSolaris -> Illumos.

> That is just so wildly unlikely, though.  Not because Joyent is so
> lovey dovey about the node community (though we totally are), but
> because it'd just be such a profoundly *stupid* move.  Joyent is in a
> position to make sure that node continues to grow in awesomeness, that
> the community stays free and vibrant, and most importantly, that
> *other* companies invest in it and jump on this bandwagon.  It's in
> the greedy self-interest of Joyent to do these things.

I agree. it is in Joyent's interest for Node to be successful.

I do believe companies priorities change over time, and in general
they will diverge from the communities in the long run. (See Zend ::
PHP, Sun/Oracle :: Java).

> If you look at the other stuff the company's been involved with,
> there's a clear track record of using open source software, being a
> good open source citizen, and reaping rewards from it.  It's not a
> dumb bunch of people.

What other open source projects has Joyent really produced with the
cooperation of an open community?

The smart platform? <https://github.com/joyent/smart-platform>

It doesn't seem at all like a community driven project, its a product
that happens to also be on Github.

You and Ryan both have open source track records, and that is great,
but Joyent -- as a company -- does not have the open source track
record with a community based project. This doesn't mean it can't or
won't happen, only I disagree with your statement about Joyent being a
"good open source citizen, and reaping rewards from it. ".

Well shit, now I'm being an ass. Best I go to bed.

Thanks,

Paul

joelparkerhenderson

unread,
Nov 10, 2010, 7:32:58 AM11/10/10
to nodejs
Mark (Joyent) wrote:
> The point here is that Joyent, like anyone else who uses the Node
> source code in any way, is bound by the terms of the source code
> license, namely, MIT.

Mark, that's a good intent, but it's not the way your CLA reads.
The CLA specifcally says Joyent can change a contributor's license.

> If we were using the code in a proprietary product, we still
> have to respect that license.

Actually, your CLA says that Joyent may, at its sole discretion,
change the license for whatever Joyent has received from contributors.


> If you're contributing code, it's under MIT as well, and published
> in public git repos where it could be rescued/forked/ whatever

If these repos use the word "Node" then Joyent could shut them down
for infringement. Sun did this for projects using "Java" in the name.


> The community around node is strong, and comparing this
> to the MySQL/Oracle situation is far fetched, IMO.

Then Joyent should have no problem striking the "proprietary" part of
the CLA.


> No GPL and associated "dual licensing" legacy is involved

What you write is contrary to the CLA, which specifically
says Joyent may add additional licenses.


> The CLA for node is identical to the CLA used by V8 itself, btw.

No, it's not. In fact, it's quite different, e.g. Joyent can create
proprietary licenses.
Whomever told you that the two are identical is mistaken; you can
compare them to see.


Mark, your comments each seem to contradict to what the CLA states;
Assuming you are being honest and truthful, then what would it take to
get
a Joyent lawyer to fix the CLA to bring it into line with what you're
writing?

Thanks,
Joel

rtweed

unread,
Nov 10, 2010, 7:37:16 AM11/10/10
to nodejs
I guess my only concern would be downstream chains of acquisitions -
eg company x buys out company y, then a few years later company z
acquires company x, and then realises along the way they've also
acquired this bit of Open Source software that was originally owned by
company y. So it isn't the original company y you need to worry
about. It's when big bad z or someone else down the acquisition chain
gets their hands on it....

Rob

rtweed

unread,
Nov 10, 2010, 7:52:13 AM11/10/10
to nodejs
Oh, however I forgot to say - congratulations Ryan! You've done a
great job!

I'm loving every minute of working with Node.js and hope this news is
nothing but positive for all of us in the community.

Rob

Renato Elias

unread,
Nov 10, 2010, 8:55:46 AM11/10/10
to nod...@googlegroups.com
Much more that a great work it opened the vision of many people !

Ryan Dahl

unread,
Nov 10, 2010, 9:59:29 AM11/10/10
to nod...@googlegroups.com
On Wed, Nov 10, 2010 at 4:32 AM, joelparkerhenderson
<joelparke...@gmail.com> wrote:
> Mark (Joyent) wrote:
>> The point here is that Joyent, like anyone else who uses the Node
>> source code in any way, is bound by the terms of the source code
>> license, namely, MIT.
>
> Mark, that's a good intent, but it's not the way your CLA reads.
> The CLA specifcally says Joyent can change a contributor's license.

I (and soon Joyent) have issued you the right to use Node under these terms:

https://github.com/ry/node/blob/v0.3.0/LICENSE

As a user this is the only license that you are bound to. Namely - if
this is what you're inferring - Joyent cannot take a version of the
software distributed to you in the past under the MIT, call you up,
dictate that your license is now commercial, and demand money. You
should read the terms of your license and consider if they suit you.
Specifically you are allowed "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".

Contributors to my repository must be bound by stricter terms.

Dean Landolt

unread,
Nov 10, 2010, 10:53:30 AM11/10/10
to nod...@googlegroups.com
On Wed, Nov 10, 2010 at 4:16 AM, Isaac Schlueter <i...@izs.me> wrote:
Let's say, hypothetically, Joyent turns evil, and decides to
dual-license node with a GPL version and a proprietary for-pay
license.  It's unlikely, but stranger things have happened.

The day that happens, anyone can just take Node, fork it, call it
"OpenNode", and keep right on trucking.  Certainly, damage could be
done to the community, and it could erode the confidence in Node as a
platform.

No, if going down this route Joyent would have long since trademarked the hell out of the node brand.
 

That is just so wildly unlikely, though.  Not because Joyent is so
lovey dovey about the node community (though we totally are), but
because it'd just be such a profoundly *stupid* move.  Joyent is in a
position to make sure that node continues to grow in awesomeness, that
the community stays free and vibrant, and most importantly, that
*other* companies invest in it and jump on this bandwagon.  It's in
the greedy self-interest of Joyent to do these things.

This is the most important point. I'd love to see a compelling scenario painted in which Joyent could actually benefit from acting in bad faith here. I can't think of one but I'm not always that creative.

Dean Landolt

unread,
Nov 10, 2010, 10:55:12 AM11/10/10
to nod...@googlegroups.com
On Wed, Nov 10, 2010 at 3:11 AM, Micheil Smith <mic...@brandedcode.com> wrote:
It is with general respect that Joyent would not be likely to sell your
code, given it is a contribution to an open source project. It is incredibly
unlikely that Joyent will ever change the licensing of Node.js, which
would end up being an extremely costly and complicated process.

Made much cheaper by the copyright assignment. But still...
 

The CLA only covers code contributed to the Node.js project, in other words, what
you get when you checkout ry/node from github or when you download a tarball.

The CLA does NOT cover any modules that you contribute to the community via
things like NPM.

The CLA has been in place for the last few months, all that has changed is now
instead of transferring your IP and such to Ryan Dahl, you are now transferring
them to Joyent, the text of the CLA is the same otherwise.

A really good example I have of CLA's is the Dojo Toolkit project. You sign a
CLA to say that the code you contribute to the Dojo Toolkit is owned / intellectually
owned by the Dojo Foundation.

The Dojo Foundation is very different than a corporate entity. Very different (as pointed out by Paul later in the thread). Still, as Mikeal pointed out, because Node is MIT there's very little Joyent could do to screw the community. They could take their ball and go home (e.g., change the license to a closed-source or add otherwise-unsavory terms), but everything that was released up to that point under the MIT can be forked by whomever. This would result in tumult, sure, but it would most certainly keep the project and community open and punish Joyent for being bad actors, likely pushing them out of the mix entirely.

This is not in Joyent's interests, thus, there's very little to fear IMHO.
 
The reason for this is to protect the project from
people contributing things to http://github.com/ry/node and then later contact ryan
or someone and saying that they need the code they contributed removed from the
project because it's actually owned by their own company. This is something that
actually happened in the early days of the jQuery UI project, and part of the reason
as to why jQuery UI didn't launch with a Menu/Toolbar widget back when it was
first released.

It would be bad and unfriendly developer practice for Joyent to use the code you
contribute to Node.js under different license for other purposes. The CLA is purely
something to safe guard the project from bad Intellectual Property rights and licensing,
which have plagued various other projects.

It's a promise from contributors to Joyent (and the community) that they're acting in good faith. The MIT license is Joyent's guarantee of good faith -- as has been pointed out, this guarantee should be all we need. Of course, this says nothing about patents (also pointed out by Paul later in the thread), which could royally screw us all -- including Joyent.
 

If you have contributed any code to Node.js over the past few months, you should have
signed the CLA any way. The only difference is that now, rather than handing your IP
to Ryan, you're handing it to Joyent.

Hopefully this clears some stuff up a little bit. There's no need to over react to the
change in the name on the CLA.


Indeed. 

Nathan Rajlich

unread,
Nov 10, 2010, 11:52:07 AM11/10/10
to nod...@googlegroups.com
So does this mean the GitHub repo is going to move (to maybe http://github.com/joyent/node)? And does this mean that Isaacs and possibly other people on the team at Joyent will be pushing commits/patches to the repo?

Isaac Schlueter

unread,
Nov 10, 2010, 12:16:27 PM11/10/10
to nod...@googlegroups.com
On Wed, Nov 10, 2010 at 01:40, Paul Querna <pqu...@apache.org> wrote:
> You could not call it OpenNode, as a fork, the holder of the Copyright
> and License would most likely enforce this as a confusing product
> name, and it would need to be renamed something that didn't keep the
> "node" name.  See also, OpenSolaris -> Illumos.

Actually, according to the terms of the license, you could call it
"NodeJS". It's not the PHP license.

> I do believe companies priorities change over time, and in general
> they will diverge from the communities in the long run. (See Zend ::
> PHP, Sun/Oracle :: Java).

Right. And if Joyent is a douche, they're going to find themselves
empty-handed, as the node community will find a new champion.

> What other open source projects has Joyent really produced with the
> cooperation of an open community?

Rails and Solaris come to mind. Joyent's had a lot of interaction
with both products. (The Reasonably Smart platform was an
acquisition, just like Node is.)

> The smart platform? <https://github.com/joyent/smart-platform>
>
> It doesn't seem at all like a community driven project, its a product
> that happens to also be on Github.

They tried to build a community around it. But it didn't catch like node has.

To what extent is node really "community driven" or "community based",
though? This has always been a monarchy. Node's success has always
relied in large part on the personality and esthetic sense of its
author.

If anything, my one fear about the Joyent-Node stuff is that it
becomes *less* unilateral. I certainly don't want the Joyent
engineering management team, brilliant guys though they may be, to
come in and say "make node do this, the community wants it". But I'm
really not worried about that happening. They're not morons, to say
the least.

> Well shit, now I'm being an ass. Best I go to bed.

I disagree. Your points are interesting and well stated, and I'm glad
you bring them up.


On Wed, Nov 10, 2010 at 08:52, Nathan Rajlich <nat...@tootallnate.net> wrote:
> So does this mean the GitHub repo is going to move (to maybe
> http://github.com/joyent/node)? And does this mean that Isaacs and possibly
> other people on the team at Joyent will be pushing commits/patches to the
> repo?

I sure hope not.

--i

Mikeal Rogers

unread,
Nov 10, 2010, 12:51:40 PM11/10/10
to nod...@googlegroups.com
On Wed, Nov 10, 2010 at 12:28 AM, Mark Mayo <mark...@gmail.com> wrote:
Umn, no you don't :)

That's not how copyright works. If you *own* the copyright then you can literally do whatever you want with it. The public is bound to the terms of license you grant the code to them under but you aren't as the owner.

Regardless of how great Joyent is, and I've spent a lot of time with Joyent lately and they are all wonderful people, I don't think you can count on that kindness because, as we've seen, any company could some day be bought by Oracle ;)

But we don't need Joyent or anyone else to protect us from the problems people have associated with recent Oracle crap because the MIT license saves us for an entirely different reason. Nobody would ever have any reason to buy a proprietary license/product from Joyent that uses the public node code because the MIT license allows them to use the code in basically anything without forcing them to open source it themselves.

Nor does owning the copyright afford Joyent much advantage to using the code in a proprietary product compared to anyone else since the code is available under liberal license.

This does however mean that Joyent has become the target of any litigation over patent or copyright issues that might be leveled against node in the future. Having someone who isn't Ryan be that target is probably a good thing for the community :)
 
MySQL was under the GPL and that was used to provide proprietary licenses to vendors at their discretion. Sun released a bunch of stuff under it's own license that granted Sun some special permissions that are now being severely abused. Sun also patented a lot of the technology they released as open source and that portfolio is now being used to sue competitors.

MIT doesn't save us from the abuse of possible patentable material so something in writing in this area might be a good idea.

-Mikeal

Mikeal Rogers

unread,
Nov 10, 2010, 12:53:32 PM11/10/10
to nod...@googlegroups.com
I would be +1 on moving to the Apache2 license or any license with a public patent grant.

This isn't a theoretical problem, we're seeing lots of patent litigation out of Oracle using Sun's portfolio.

Mikeal Rogers

unread,
Nov 10, 2010, 12:57:44 PM11/10/10
to nod...@googlegroups.com
I think the concern has less to do with people thinking Joyent might do something stupid/evil and more to do with the fact that at some point in the future Joyent could be bought by Oracle and do Oracle the stupid/evil things we expect Oracle to do.

But as you say, the licensing affords us a lot of flexibility if that happens and you've seen similar forks recently of projects Oracle is now ruining.

Mikeal Rogers

unread,
Nov 10, 2010, 1:03:34 PM11/10/10
to nod...@googlegroups.com
On Wed, Nov 10, 2010 at 9:16 AM, Isaac Schlueter <i...@izs.me> wrote:
On Wed, Nov 10, 2010 at 01:40, Paul Querna <pqu...@apache.org> wrote:
> You could not call it OpenNode, as a fork, the holder of the Copyright
> and License would most likely enforce this as a confusing product
> name, and it would need to be renamed something that didn't keep the
> "node" name.  See also, OpenSolaris -> Illumos.

Actually, according to the terms of the license, you could call it
"NodeJS".  It's not the PHP license.

This falls under trademark (even if the name isn't a registered trademark you can claim that you have been using it exclusively and therefor own it).

Trademark is not covered by the copyright license and in fact cannot be licensed in any kind of open sourcey way because trademark is fucked up and crazy law.
 

Eugene Lazutkin

unread,
Nov 10, 2010, 1:57:40 PM11/10/10
to nod...@googlegroups.com
Actual situation with Dojo and node.js is a little bit different.

Dojo's CLA assigns rights to Dojo Foundation
(http://dojofoundation.org/), which is a non-profit entity chartered to
promote its projects as open source. Besides Dojo it shelters Kris Zyp's
Persevere, Joe Walker's DWR, John Resig's Sizzle, James Burke's
RequireJS, and many other interesting open source projects.

Joyent is a private for-profit organization, which does the bidding of
its unnamed owners. There are no restrictions on what they may do with
node.js. Obviously, their ideas and the vision for node.js can be
completely noble and totally unselfish. Or they may have totally
different plans.

Money-wise I have no doubt that Joyent has much much deeper pockets,
than Dojo Foundation, or any other non-profit open source foundation for
that matter. If this is all that is required to develop node.js we have
no reason to complain.

Favoritism is the other matter. I just don't see Dojo Foundation keeping
jQuery's Sizzle from using in jQuery in favor of Dojo, or any other
project. Or preventing, say, Oracle from using Dojo in their products.
It never happened and I cannot even fathom how it can happen. But I saw
for-profit companies delaying open sourcing of certain code, until it is
fully integrated in their product line ("to battle-test it!") before
making it available to potential competitors --- to be completely honest
all such cases backfired and hurt respective open source projects and I
hope that Joyent is smarter than that.

Cheers,

Eugene Lazutkin
http://lazutkin.com/

Eugene Lazutkin

unread,
Nov 10, 2010, 2:21:15 PM11/10/10
to nod...@googlegroups.com
Just curious why Joyent even need to reassign CLA to them? Nobody
explained it. How is it different than supporting an open source
project, which is supervised by more appropriate entity?

Dojo Foundation supports dual-licensed projects --- AFL and BSD. Why not
ask Dojo Foundation to provide a legal shelter for node.js and help with
technical details? It knows JavaScript, and AFAIK its board members,
project leads council, and all voting members are well-known in the
JavaScript community, and are avid supporters and users of node.js.

And Dojo Foundation, while JavaScript-oriented, is not the only game in
town. Apache Software Foundation is well-respected and has the proven
track record. Why not ask them?

I am sure that Joyent would not have problems with such arrangement,
because I believe that their intentions are pure. This move will remove
any cause for criticism and uncertainty.

Cheers,

Eugene Lazutkin
http://lazutkin.com/

Micheil Smith

unread,
Nov 10, 2010, 2:48:12 PM11/10/10
to nod...@googlegroups.com
One thing to note on the Dojo Foundation though is their
100 point open-source policy. Which isn't terrible hard to
reach (I'm pretty sure node is already at that 100 point
level, but it's something to consider)

Yours,
Micheil Smith
--
BrandedCode.com

Dean Landolt

unread,
Nov 10, 2010, 3:15:38 PM11/10/10
to nod...@googlegroups.com
On Wed, Nov 10, 2010 at 1:03 PM, Mikeal Rogers <mikeal...@gmail.com> wrote:


On Wed, Nov 10, 2010 at 9:16 AM, Isaac Schlueter <i...@izs.me> wrote:
On Wed, Nov 10, 2010 at 01:40, Paul Querna <pqu...@apache.org> wrote:
> You could not call it OpenNode, as a fork, the holder of the Copyright
> and License would most likely enforce this as a confusing product
> name, and it would need to be renamed something that didn't keep the
> "node" name.  See also, OpenSolaris -> Illumos.

Actually, according to the terms of the license, you could call it
"NodeJS".  It's not the PHP license.

This falls under trademark (even if the name isn't a registered trademark you can claim that you have been using it exclusively and therefor own it).

Spot on. There actually exists the concept of common law trademark. But more to the point, Joyent could at any point trademark the ever-loving shit out of the node brand. This would not necessarily be a bad thing -- protecting the brand could be a very thing for the community. My point was that if Joyent did turn evil and close up the license they most certainly would get these trademarks and enforce them.
 

Trademark is not covered by the copyright license and in fact cannot be licensed in any kind of open sourcey way because trademark is fucked up and crazy law.


Dude, what? Trademark is easily the most sane of our intellectual property laws -- so much so that it's not really an intellectual property law at all but really a consumer protection law to prevent fraudulent labeling (marking) and consumer confusion. It is true that a trademark has to be enforced to be maintained, but this can be a good thing (no orphaned works). It definitely can be used in an open sourcey way -- it's only slightly complicated by the requirement for it to be used in commerce (but this is still applicable to many open source projects). Document and abide by a liberal mark usage policy and you're good to go.

Isaac Schlueter

unread,
Nov 10, 2010, 3:33:16 PM11/10/10
to nod...@googlegroups.com
>> Trademark is not covered by the copyright license and in fact cannot be
>> licensed in any kind of open sourcey way because trademark is fucked up and
>> crazy law.

Right, but there are already a few other projects that have used the
"node" name, and that's why you can't do "apt-get install node" to get
it. (See other threads on this topic.)

Anyway, you could call it awesomejs or whatever other name. It really
doesn't matter much.

None of us are lawyers here, and this is something that the courts
bicker over. So really, making any kind of strong claim one way or
the other is a bit silly. I retract all my previous claims about the
nodejs brand.

You can definitely fork the code, put a new logo on it, call it
something else, sell it, sublicense it, release it under the DWTFPL,
or anything else. So all of this is kind of a non-issue, imo.


> Joyent is a private for-profit organization, which does the bidding of
> its unnamed owners.

No, it's owners all have names.
http://www.joyent.com/about/management/
http://www.joyent.com/about/board/


> There are no restrictions on what they may do with
> node.js. Obviously, their ideas and the vision for node.js can be
> completely noble and totally unselfish. Or they may have totally
> different plans.

Seriously, think about this. What if Ryan had written the list to
say, "I've started up a company called NodeJS, taken some investment,
and hired Isaac. My new company owns the nodejs brand, and the CLA
points to it now."

How would that be any different?

Can we go back to talking about semicolons and promises?

--i

Mikeal Rogers

unread,
Nov 10, 2010, 3:40:00 PM11/10/10
to nod...@googlegroups.com
The enforcement requirement is incredibly rigid because you can't "enforce" your own terms of the usage, you actually have to enforce that every usage is explicitly permitted which is impossible in an open source context.

Successful usage of trademark in open source is very rare and some of the better usage has fallen under "crowd sourcing" the enforcement but it's still a giant pain in the ass.

Bert Belder

unread,
Nov 10, 2010, 4:04:59 PM11/10/10
to nodejs
It doesn't really matter to whom the CLA is assigned as long as node
itself is published under the MIT license. Right now anyone (me, you,
and Microsoft) can start publishing Node under a proprietary license,
but we generally aren't afraid of that happening because everyone will
happily ignore the proprietary 'fork' and keep on contributing to the
libre version.

The real risk IMHO is that Joyent hires the most skilled and active
Node contributors and exercises control over them. That this is a risk
indeed was shown by the ExtJS license change trick; it succeeded not
because they could change the license (everyone could have) but
because they could develop their semi-closed version of Ext much
faster than the community could maintain the open version. Another
example: the development of APC (an open php accelerator) has been
slow because Zend's PHP engineers wouldn't work on a feature competing
with one of Zend's own proprietary products.

But this is really a luxury problem. Right now a company is paying
people to work on a project that you love and get to use for free. If
you're afraid of Joyent, hire some node developers yourself. Or pray
that HP will :-)

- Bert

rtweed

unread,
Nov 10, 2010, 4:06:22 PM11/10/10
to nodejs

> Can we go back to talking about semicolons and promises?
>

how about this then?

http://www.usabilitycounts.com/2008/05/10/silly-saturdays-dilbert-and-bosses/

:-)

Mikeal Rogers

unread,
Nov 10, 2010, 4:31:12 PM11/10/10
to nod...@googlegroups.com
On Wed, Nov 10, 2010 at 1:04 PM, Bert Belder <bertb...@gmail.com> wrote:
It doesn't really matter to whom the CLA is assigned as long as node
itself is published under the MIT license. Right now anyone (me, you,
and Microsoft) can start publishing Node under a proprietary license,

No, you can't.

You can create a proprietary product that includes node code but you cannot release the code itself under a different license or copyright owner.
 
but we generally aren't afraid of that happening because everyone will
happily ignore the proprietary 'fork' and keep on contributing to the
libre version.

The real risk IMHO is that Joyent hires the most skilled and active
Node contributors and exercises control over them. That this is a risk
indeed was shown by the ExtJS license change trick; it succeeded not
because they could change the license (everyone could have) but
because they could develop their semi-closed version of Ext much
faster than the community could maintain the open version. Another
example: the development of APC (an open php accelerator) has been
slow because Zend's PHP engineers wouldn't work on a feature competing
with one of Zend's own proprietary products.

But this is really a luxury problem. Right now a company is paying
people to work on a project that you love and get to use for free. If
you're afraid of Joyent, hire some node developers yourself. Or pray
that HP will :-)

- Bert

Eugene Lazutkin

unread,
Nov 10, 2010, 4:34:20 PM11/10/10
to nod...@googlegroups.com
Inline.

On 11/10/2010 02:33 PM, Isaac Schlueter wrote:
>> Joyent is a private for-profit organization, which does the bidding of
>> its unnamed owners.
>
> No, it's owners all have names.
> http://www.joyent.com/about/management/
> http://www.joyent.com/about/board/

Sorry, but I have to ask: do you understand the difference between
managers, directors, board members, and owners? Being an officer may or
may not indicate that a person in question is an owner of the company,
and so on. Seriously, think about it.

The only ownership information I know of comes from
http://www.joyent.com/about/ --- "Joyent is privately held and backed by
Greycroft Partners and Intel Capital". If you still think managers and
board members on those pages are owners in charge (or have a reliable
information on that), please talk to your superiors to update the About
page with better information.

>> There are no restrictions on what they may do with
>> node.js. Obviously, their ideas and the vision for node.js can be
>> completely noble and totally unselfish. Or they may have totally
>> different plans.
>
> Seriously, think about this. What if Ryan had written the list to
> say, "I've started up a company called NodeJS, taken some investment,
> and hired Isaac. My new company owns the nodejs brand, and the CLA
> points to it now."
>
> How would that be any different?

The difference is obvious: your scenario is a hypothetical "what if..."
situation, while absorbing by Joyent is already happened.

Personally I don't worry too much about it, but in order to quell the
fears, it is better to prevent both of situations (ate by Joyent, or
Ryan going crazy) and shelter the project under non-profit entity
charged with nurturing it rather than going over "what if..." stuff.

> Can we go back to talking about semicolons and promises?

Ohhhh, I love promises of semicolons! ;-)

Cheers,

Eugene

Dean Landolt

unread,
Nov 10, 2010, 4:39:40 PM11/10/10
to nod...@googlegroups.com
On Wed, Nov 10, 2010 at 3:40 PM, Mikeal Rogers <mikeal...@gmail.com> wrote:


On Wed, Nov 10, 2010 at 12:15 PM, Dean Landolt <de...@deanlandolt.com> wrote:


On Wed, Nov 10, 2010 at 1:03 PM, Mikeal Rogers <mikeal...@gmail.com> wrote:


On Wed, Nov 10, 2010 at 9:16 AM, Isaac Schlueter <i...@izs.me> wrote:
On Wed, Nov 10, 2010 at 01:40, Paul Querna <pqu...@apache.org> wrote:
> You could not call it OpenNode, as a fork, the holder of the Copyright
> and License would most likely enforce this as a confusing product
> name, and it would need to be renamed something that didn't keep the
> "node" name.  See also, OpenSolaris -> Illumos.

Actually, according to the terms of the license, you could call it
"NodeJS".  It's not the PHP license.

This falls under trademark (even if the name isn't a registered trademark you can claim that you have been using it exclusively and therefor own it).

Spot on. There actually exists the concept of common law trademark. But more to the point, Joyent could at any point trademark the ever-loving shit out of the node brand. This would not necessarily be a bad thing -- protecting the brand could be a very thing for the community. My point was that if Joyent did turn evil and close up the license they most certainly would get these trademarks and enforce them.
 

Trademark is not covered by the copyright license and in fact cannot be licensed in any kind of open sourcey way because trademark is fucked up and crazy law.


Dude, what? Trademark is easily the most sane of our intellectual property laws -- so much so that it's not really an intellectual property law at all but really a consumer protection law to prevent fraudulent labeling (marking) and consumer confusion. It is true that a trademark has to be enforced to be maintained, but this can be a good thing (no orphaned works). It definitely can be used in an open sourcey way -- it's only slightly complicated by the requirement for it to be used in commerce (but this is still applicable to many open source projects). Document and abide by a liberal mark usage policy and you're good to go.

The enforcement requirement is incredibly rigid because you can't "enforce" your own terms of the usage, you actually have to enforce that every usage is explicitly permitted which is impossible in an open source context.


Even if this were completely true (it's not -- you can rubber stamp your permissions or issue blanket usage licenses) think about the only possible outcome for a second: dilution. It would be as if no trademark existed. From an "open sourcey" standpoint I fail to see he harm. Technically it means others could fork your project and use the same name, but we have social norms to prevent that :)

Bert Belder

unread,
Nov 10, 2010, 4:46:23 PM11/10/10
to nodejs


On Nov 10, 10:31 pm, Mikeal Rogers <mikeal.rog...@gmail.com> wrote:
> On Wed, Nov 10, 2010 at 1:04 PM, Bert Belder <bertbel...@gmail.com> wrote:
> > It doesn't really matter to whom the CLA is assigned as long as node
> > itself is published under the MIT license. Right now anyone (me, you,
> > and Microsoft) can start publishing Node under a proprietary license,
>
> No, you can't.
>
> You can create a proprietary product that includes node code but you cannot
> release the code itself under a different license or copyright owner.

Okay, sure, you would need to add a little bit of your own, but that's
just a technicality.

BTW, suggestion to Joyent: keep developing the 'community edition',
charge money for the closed-source 'corporate edition' with proper SSL
support.

(PS: this is just a joke, I'm not implying that Joyent may actually
consider this)

Mikeal Rogers

unread,
Nov 10, 2010, 4:51:04 PM11/10/10
to nod...@googlegroups.com
After thinking about this for a while I only have 2 real concerns and the designation of ownership itself isn't one of them.

1.) Patents
2.) Ability for contributors to sign this CLA and contribute to node.

We currently have no patent protection in place. Frankly, I'm wondering why Joyent isn't concerned about this. It's entirely possible for a contributor to grant the copyright of some source code to Joyent that they have a valid patent on and then sue Joyent for infringing on that patent some time later. That isn't theoretical, that's actually happened before, it's even happened in open standards.

The patent protection thing is going to be a big deal for adoption by large enterprises as well. One of the reasons that Apache projects make big companies feel warm and fuzzy is because all the patent issues are taken care of and if you sue a project for infringing on those patents they lose the right to run any software copyrighted by Apache.

This feeds in to the next concern I have. The CLA contains language that I'm pretty sure is unnecessary but will nonetheless be red flagged by any big company lawyers and probably be a blocker for a potential contributor employed by that company.

While I don't expect Joyent to divulge all of their strategic plans or their potential future plans for node.js it is worth having a real conversation about adoption and contributions they intend to gain by supporting node and how some of this may actually hinder those goals. The CLA reads like boiler for a proprietary CLA that has some open source language added. It also has language that is totally unnecessary, like the reservation of particular rights to code they are already getting the assignment of ownership for, which can only make it harder for people to sign.

To Issac's point that this wouldn't have been such a big issue if Ryan and him did a startup and took funding, I don't agree. I think, like most things in open source projects, any change like this brings to the surface a bunch of issues people may have with the project and is a potential opening for fixing and future proofing agains those problems.

-Mikeal


Mikeal Rogers

unread,
Nov 10, 2010, 4:55:15 PM11/10/10
to nod...@googlegroups.com
On Wed, Nov 10, 2010 at 1:46 PM, Bert Belder <bertb...@gmail.com> wrote:


On Nov 10, 10:31 pm, Mikeal Rogers <mikeal.rog...@gmail.com> wrote:
> On Wed, Nov 10, 2010 at 1:04 PM, Bert Belder <bertbel...@gmail.com> wrote:
> > It doesn't really matter to whom the CLA is assigned as long as node
> > itself is published under the MIT license. Right now anyone (me, you,
> > and Microsoft) can start publishing Node under a proprietary license,
>
> No, you can't.
>
> You can create a proprietary product that includes node code but you cannot
> release the code itself under a different license or copyright owner.

Okay, sure, you would need to add a little bit of your own, but that's
just a technicality.

No, that's not how it works. The original code is still MIT licensed and copyright to the original owner, the changes and code you wrote is under whatever license you like. You do not have the right under any circumstance to change the ownership of property, the code is property and sections of it can be separate properties from it as a whole. The product as a whole is property you can license however you like because the MIT license allows you to do so but the code you use that you didn't create is still owned by the original owner.
 

BTW, suggestion to Joyent: keep developing the 'community edition',
charge money for the closed-source 'corporate edition' with proper SSL
support.

(PS: this is just a joke, I'm not implying that Joyent may actually
consider this)

--

Ryan Gahl

unread,
Nov 10, 2010, 4:53:56 PM11/10/10
to nod...@googlegroups.com
Congratulations, Ryan!

Joyent will be good for node in the long run... or it won't. (but it will)

--rg
(I like semicolons)



--

Benoit Chesneau

unread,
Nov 10, 2010, 4:58:00 PM11/10/10
to nod...@googlegroups.com
On Wed, Nov 10, 2010 at 10:31 PM, Mikeal Rogers <mikeal...@gmail.com> wrote:
>
>
> On Wed, Nov 10, 2010 at 1:04 PM, Bert Belder <bertb...@gmail.com> wrote:
>>
>> It doesn't really matter to whom the CLA is assigned as long as node
>> itself is published under the MIT license. Right now anyone (me, you,
>> and Microsoft) can start publishing Node under a proprietary license,
>
> No, you can't.
> You can create a proprietary product that includes node code but you cannot
> release the code itself under a different license or copyright owner.
>
>>
Sorry but under MIT you can change the license. The only thing you
can't is removed the copyright.

Tim Caswell

unread,
Nov 10, 2010, 4:58:50 PM11/10/10
to nod...@googlegroups.com
They have :D

On Wed, Nov 10, 2010 at 1:04 PM, Bert Belder <bertb...@gmail.com> wrote:
>

Isaac Schlueter

unread,
Nov 10, 2010, 5:06:25 PM11/10/10
to nod...@googlegroups.com
On Wed, Nov 10, 2010 at 13:58, Benoit Chesneau <bche...@gmail.com> wrote:
> Sorry but under MIT you can change the license. The only thing you
> can't is removed the copyright.

You can change the license of the portions that you create or change.

If your new project contains a substantial portion of node's code,
then it needs to have the license included with any source code
distributions.

But you can compile node, put it on a CD, and say that no one has the
right to even look at it. That's explicitly allowed.

You can also put a little license agreement that says you don't have
the right to do anything with any of this code, and that it uses
substantial portions of nodejs under the following license: $(cat
node/LICENSE)

You can't change the license *of node itself* (or "substantial"
portions thereof), but that's hardly necessary if you can change the
license of whatever it is that you're releasing.

The MIT license isn't a no-op. But it's not much more.

There is basically nothing that you can't do with Joyent at the helm
that you could do before. This whole discussion is academic, except
also boring and no one's learning anything.

--i

Dean Landolt

unread,
Nov 10, 2010, 5:08:27 PM11/10/10
to nod...@googlegroups.com
What Mikeal's saying is that you cannot change the license of the work because you do not own the copyright. You can release a new work that includes the work in question that. You would own the copyright on this new work and can license it any way you want. But it's important that you realize you cannot change the license of the original work. You can use a more restrictive license for your new work, or even a less restrictive license (Public Domain, DWTFPL) -- but this would, of course, only apply to the new work -- not the original work. For instance you cannot unilaterally put node in the Public Domain. Only Joyent can.

Micheil Smith

unread,
Nov 10, 2010, 5:09:06 PM11/10/10
to nod...@googlegroups.com
Inline,


On 11/11/2010, at 8:58 AM, Tim Caswell wrote:
> They have :D

That, Bert, Is Tim speaking from experience as an employee :D

Dean Landolt

unread,
Nov 10, 2010, 5:10:06 PM11/10/10
to nod...@googlegroups.com
I disagree. Mikeal's points about patents are incredibly useful.

Mikeal Rogers

unread,
Nov 10, 2010, 5:17:38 PM11/10/10
to nod...@googlegroups.com
My issues don't have a lot to do with Joyent, maybe we should start a new thread.

The only thing this change does is bring these issues to the forefront. The patent issue being a big one as suing Ryan for patent infringement for monetary gain is laughable but suing a company with money in the bank is much more plausible.

We could assign the copyright to a grizzly bear and I wouldn't care all that much so long as some of the CLA language changed so that we wouldn't miss out on some contributors and there was some patent protection.

For my usage of the project and it's code the owner is irrelevant and if the owner changed to Oracle I would lead an effort to fork it :)

-Mikeal

brez

unread,
Nov 10, 2010, 5:52:24 PM11/10/10
to nodejs
A comparable example is Engine Yard's development of Rails 3 - they
funded it with two full time developers but never demanded they own
the Rails copyright / trademark, i.e. if a company if generally
sincere about an open source contribution why do they feel its
necessary to shore up the intellectual property associated with it?
The likely answer is to have a significant advantage over any emerging
competition for hosting - which is fine, Joyent is a business after
all, the primary focus is, and should be, revenue. Now whether or not
you want to be making it for them is up to you..

On Nov 9, 6:53 pm, Ryan Dahl <r...@tinyclouds.org> wrote:
> It has become clear over the last few months that Node's popularity
> can support a larger, more structured development team. Indeed, for
> Node to achieve massive success -- for it to be the next PHP -- it
> needs investment that can support a team of full-time developers and a
> corporate entity behind it that can, for example, enter into
> agreements with large companies who are keen to bring Node to their
> projects and make contributions back to the project. Someone spending
> time and resources on evangelism and stuff wouldn't hurt either.
>
> Joyent's core business (cloud computing) aligns well with a free and
> open Node.js. For these reasons I've made an agreement with Joyent
> that officially puts Node under their roof. The only effective change
> for developers is that the contributor agreement will be directed to
> Joyent rather than myself. Node will continue regular releases under
> the MIT license. As part of the agreement -- in addition to supporting
> my own development on Node -- we've recently hired Isaac Schlueter and
> plan to add at least one more full-time developer on the core project
> immediately. As a business we will begin selling instances of our
> minimalistic no.de Node hosting service in several weeks and soon
> offer support services developing custom bindings to Node.
>
> Contact me at r...@joyent.com if you have any questions.

djangst

unread,
Nov 10, 2010, 6:30:13 PM11/10/10
to nodejs
It's clear what didn't happen here, and that is that the community
didn't get excited about this partnership. It was like meeting the
"right now" person instead of the right person. No offense is intended
and this is purely observational, but the lack of a groundswell of
support and silence from many says that the announcement has had a
slightly chilling effect on the community.

If the author and other contributors finally got some help and
financial support for their hard work then that seems like a positive
thing (for Joyent too), but given what could have been this seems like
a missed opportunity, and instead we have good people who now have to
toe the company line with regard to honest and open communications.

If the new partner had been a different company, like a GitHub for
example, would the reaction have been different?

[This would probably make a good case study for commercial/OSS
partnerships.]

Dean Landolt

unread,
Nov 10, 2010, 6:42:45 PM11/10/10
to nod...@googlegroups.com
On Wed, Nov 10, 2010 at 6:30 PM, djangst <dja...@gmail.com> wrote:
It's clear what didn't happen here, and that is that the community
didn't get excited about this partnership. It was like meeting the
"right now" person instead of the right person. No offense is intended
and this is purely observational, but the lack of a groundswell of
support and silence from many says that the announcement has had a
slightly chilling effect on the community.

Really? I took the relative silence as tacit approval and a sign that most people don't get as amped up over IP issues as you or I :)
 

If the author and other contributors finally got some help and
financial support for their hard work then that seems like a positive
thing (for Joyent too), but given what could have been this seems like
a missed opportunity, and instead we have good people who now have to
toe the company line with regard to honest and open communications.

Excuse me? Ryan's been a Joyent employee for quite a while now -- whether he or his employer owns the copyright is irrelevant -- he's just as beholden to his employer as the day he started with them when it comes to what he says. If he doesn't like it he can quit, and continue developing node anyway. The only thing that would change is that he wouldn't be able to scrub Joyent from the copyright statements without getting the copyright back or doing a clean-room rewrite. This is getting silly, people.

[snip]

Micheil Smith

unread,
Nov 10, 2010, 6:44:07 PM11/10/10
to nod...@googlegroups.com
Djangst,

Node has been paying ryan (the lead developer) for the past few months
(I'm not sure exactly how many, but I know it's close to a year) to work on
Node.js. Recently they hired Isaac, and pay him too, to work on Node.

So, Joyent is actually giving financial support to node.

As for other contributors, I believe there's a few things in progress, however,
a lot of us are actually hired by other companies, such as Mikeal, Tim, TJ,
and myself to work on node and projects relating to it's continued development.

Yours,
Micheil Smith
--
BrandedCode.com

Ted Young

unread,
Nov 10, 2010, 7:54:55 PM11/10/10
to nod...@googlegroups.com
Congrats Ryan! Glad to hear you're getting a team, and that things
are working out with Joyent.

Ted Young
radicalDesigns
(415) 738 0456
t...@radicaldesigns.org

John

unread,
Nov 10, 2010, 7:34:16 PM11/10/10
to nodejs
On 11/10/2010 02:40 AM, Paul Querna wrote:
> On Wed, Nov 10, 2010 at 1:16 AM, Isaac Schlueter<i...@izs.me> wrote:
>> Let's say, hypothetically, Joyent turns evil, and decides to
>> dual-license node with a GPL version and a proprietary for-pay
>> license. It's unlikely, but stranger things have happened.
> I'll try to keep my replies realistic. I don't think many statements
> made in your original email were realistic.
>
> Regardless of the actual risk of it happening, this is one reason open
> source non-profit foundations of all shapes and sizes exist -- so it
> can never happen.
>
>> The day that happens, anyone can just take Node, fork it, call it
>> "OpenNode", and keep right on trucking. Certainly, damage could be
>> done to the community, and it could erode the confidence in Node as a
>> platform.
> You could not call it OpenNode, as a fork, the holder of the Copyright
> and License would most likely enforce this as a confusing product
> name, and it would need to be renamed something that didn't keep the
> "node" name. See also, OpenSolaris -> Illumos.
>
>> That is just so wildly unlikely, though. Not because Joyent is so
>> lovey dovey about the node community (though we totally are), but
>> because it'd just be such a profoundly *stupid* move. Joyent is in a
>> position to make sure that node continues to grow in awesomeness, that
>> the community stays free and vibrant, and most importantly, that
>> *other* companies invest in it and jump on this bandwagon. It's in
>> the greedy self-interest of Joyent to do these things.
> I agree. it is in Joyent's interest for Node to be successful.
>
> I do believe companies priorities change over time, and in general
> they will diverge from the communities in the long run. (See Zend ::
> PHP, Sun/Oracle :: Java).
>
>> If you look at the other stuff the company's been involved with,
>> there's a clear track record of using open source software, being a
>> good open source citizen, and reaping rewards from it. It's not a
>> dumb bunch of people.
> What other open source projects has Joyent really produced with the
> cooperation of an open community?
>
> The smart platform?<https://github.com/joyent/smart-platform>
>
> It doesn't seem at all like a community driven project, its a product
> that happens to also be on Github.
>
> You and Ryan both have open source track records, and that is great,
> but Joyent -- as a company -- does not have the open source track
> record with a community based project. This doesn't mean it can't or
> won't happen, only I disagree with your statement about Joyent being a
> "good open source citizen, and reaping rewards from it. ".
>
> Well shit, now I'm being an ass. Best I go to bed.

No, actually you've made some important points, especially in regards
to not using a foundation. There's already one company in the js
world that changed a license so that's always a possibility, even if
not immediately intended.

Wow, this announcement comes on the verge of doing a presentation of
Nodejs at Desert Code Camp and not sure if this is good news or bad
news.

How does this compare to Ruby ? tried to google for some info on this
but couldn't find out.



joelparkerhenderson

unread,
Nov 10, 2010, 11:57:11 PM11/10/10
to nodejs
+1 for using foundation, such as creating nonprofit organization or
adding to an existing one.

+1 for patent protection, for example as Apache does.

Are there any roadblocking reasons *not* to do a foundation and patent
protection?

Mikeal Rogers

unread,
Nov 11, 2010, 2:43:25 AM11/11/10
to nod...@googlegroups.com
Umn.... have you ever started a non-profit?

You can't just "start a foundation". It's a ton of work and require lots of things that you have to fill out on paper by hand, it's bullshit.

-Mikeal

Stephen Belanger

unread,
Nov 11, 2010, 4:09:46 AM11/11/10
to nod...@googlegroups.com
Yeah, I'd definitely go for handing it off to an existing non-profit. Creating a new foundation is extremely costly and is a very time consuming process. It can sometimes take years and millions of dollars to even start a non-profit foundation, then you need to worry about how it's going to get it's funding.

It's generally best to avoid making different foundations with overlapping goals, when they could be run as a single entity and share their resources. I think the Dojo Foundation would be the most ideal home for something like node.js, though that is entirely up to Joyent at this point.

Felix Geisendörfer

unread,
Nov 11, 2010, 4:24:12 AM11/11/10
to nodejs
I feel a little weird to see considerable amounts of my contributions
transfered to a corporate entity like this.

Personally I would have very much preferred Ryan to keep the sole
copyright on node, as I have unlimited trust in his intentions with
the project.

That being said, Ryan has made his decision and I am convinced he
would not have done so lightly. If he thinks Joyent is the best
possible home for node, then I will not argue with that.

So congratulations, I'm excited to see what this new future will
bring.

--fg

On Nov 11, 10:09 am, Stephen Belanger <cyruzdr...@gmail.com> wrote:
> Yeah, I'd definitely go for handing it off to an existing non-profit.
> Creating a new foundation is extremely costly and is a very time consuming
> process. It can sometimes take years and millions of dollars to even *start* a
> non-profit foundation, then you need to worry about how it's going to get
> it's funding.
>
> It's generally best to avoid making different foundations with overlapping
> goals, when they could be run as a single entity and share their resources.
> I think the Dojo Foundation would be the most ideal home for something like
> node.js, though that is entirely up to Joyent at this point.
>
> On Wed, Nov 10, 2010 at 11:43 PM, Mikeal Rogers <mikeal.rog...@gmail.com>wrote:
>
>
>
>
>
>
>
> > Umn.... have you ever started a non-profit?
>
> > You can't just "start a foundation". It's a ton of work and require lots of
> > things that you have to fill out on paper by hand, it's bullshit.
>
> > -Mikeal
>
> > On Wed, Nov 10, 2010 at 8:57 PM, joelparkerhenderson <
> > joelparkerhender...@gmail.com> wrote:
>
> >> +1 for using foundation, such as creating nonprofit organization or
> >> adding to an existing one.
>
> >> +1 for patent protection, for example as Apache does.
>
> >> Are there any roadblocking reasons *not* to do a foundation and patent
> >> protection?
>
> >> --
> >> You received this message because you are subscribed to the Google Groups
> >> "nodejs" group.
> >> To post to this group, send email to nod...@googlegroups.com.
> >> To unsubscribe from this group, send email to
> >> nodejs+un...@googlegroups.com<nodejs%2Bunsu...@googlegroups.com>
> >> .
> >> For more options, visit this group at
> >>http://groups.google.com/group/nodejs?hl=en.
>
> >  --
> > You received this message because you are subscribed to the Google Groups
> > "nodejs" group.
> > To post to this group, send email to nod...@googlegroups.com.
> > To unsubscribe from this group, send email to
> > nodejs+un...@googlegroups.com<nodejs%2Bunsu...@googlegroups.com>
> > .

Benoit Chesneau

unread,
Nov 11, 2010, 4:33:57 AM11/11/10
to nod...@googlegroups.com
public domain doesn't exist in all countries anyway. The current work
may stay under MIT and won't be polluted, people applying a license on
derivative work which may be rewriting some parts will use this new
license instead of MIT, so you won't be able to use the rights given
by MIT on this derivative code, even modifications of nodes if thhey
are done with this new license. But one thing that will stay is the
copyright. At least in most of the countries.

- benoît

Isaac Schlueter

unread,
Nov 11, 2010, 4:26:24 PM11/11/10
to nod...@googlegroups.com
2010/11/11 Felix Geisendörfer <fe...@debuggable.com>:

> I feel a little weird to see considerable amounts of my contributions
> transfered to a corporate entity like this.
>
> Personally I would have very much preferred Ryan to keep the sole
> copyright on node, as I have unlimited trust in his intentions with
> the project.
>
> That being said, Ryan has made his decision and I am convinced he
> would not have done so lightly. If he thinks Joyent is the best
> possible home for node, then I will not argue with that.
>
> So congratulations, I'm excited to see what this new future will
> bring.
>
> --fg
>

Once again, felixge nicely sums up the issue and its rational conclusion.

--i

joelparkerhenderson

unread,
Nov 11, 2010, 10:45:28 PM11/11/10
to nodejs
Mikeal wrote:
> Umn.... have you ever started a non-profit?

Yes, I have. It takes about an hour at City Hall and about $100 in
filing fees.
I would be happy to donate the money for the fees if that will help
this.

I'm also happy to donate time to help a Node.js nonprofit get set up,
and donate the Nolo Press book that has a step by step how-to.

It's easies to do this right now, and gets harder to do, the more time
passes,
because of how Joyent is forcing developers to turn over rights via
the CLA.


Eugene Lazutkin

unread,
Nov 11, 2010, 11:31:18 PM11/11/10
to nod...@googlegroups.com
Granted I am not a lawyer, but unless you want to restrict your
operations to your city (or county) sooner or later you have to deal
with state and federal authorities including IRS.

For example, the Dojo Foundation is a 501(c)(6) non-profit foundation
(so-called "business league"). IIRC "501(c)(6)" --- is a reference to an
IRS code section 501. The Apache Software Foundation is a non-profit
501(c)(3) corporation (so-called "charitable organization"),
incorporated in Delaware. Again "501(c)(3)" is another reference to the
same document.

Again, I didn't register any of those myself, and cannot tell how much
it costs, but I know that good lawyers who understand non-profits are
hard to find, and the process (the approval from IRS) takes a lot of
paperwork and a lot of time.

But all of this is a moot point now. My understanding is that Ryan's
decision is a done deal, which was officially announced on this list,
and now it is up to Joyent to decide what to do with their new property.
Somehow I don't see them rushing to existing non-profit foundations, or
founding new ones.

Cheers,

Eugene Lazutkin
http://lazutkin.com/

Mikeal Rogers

unread,
Nov 12, 2010, 1:49:45 AM11/12/10
to nod...@googlegroups.com
In order to start a new 501(c) you usually need to be sponsored for some time by an existing 501(c).

--
You received this message because you are subscribed to the Google Groups "nodejs" group.
To post to this group, send email to nod...@googlegroups.com.
To unsubscribe from this group, send email to nodejs+un...@googlegroups.com.
Reply all
Reply to author
Forward
0 new messages