Ian,

see below....

On Thu, 11 Sep 2008 15:09:57 -0700, you wrote:

>We're not trying to support a browser for every platform. 

You are under no legal obligation to do so.  The issue is whether you
are violating the GPL (GNU Pulbic License), which your website says
you are using in Chrome: 

http://code.google.com/chromium/terms.html

I am sorry I included my frustration over your team's reponses
regarding W2K because my main development platform is now 2003 Server
and XP.  I do have 2000 Server platforms, but I like the v9 C++
compiler best, and it does not run on 2000 Server.  As you will see
below, platform is a red herring.

>We're not planning
>on supporting Amiga, OS/2 Warp, Windows 95/98/ME, or Sun Solaris. We're
>planning on supporting Windows XP/2003/Vista/2008, Mac OS X, and Linux.
>Marc-Antoine has already given plenty of reasons why we don't want to
>support Windows 2000. 

>We can't test on it, which is a huge issue. (You can't
>honestly expect us to support something we can't test.)

Obviously true.  Not the point at all.  The point is that by not
disclosing all of the source code, you are violating the GPL and
nobody else can support what they want to.  That is really the point
of free software- you supply the source code and then others make it
do things that you did not have the time, resources, desire, whatever
to do in the first place.

>Besides that, we
>have hacks for W2K that we would love to get rid of. SafeBrowsing can't use
>MACing on W2K because of a lack of support for certain required functions
>(WinCrypt API was not updated on windows 2000 to support everything we used
>on XP/Vista). 

But this WinCrypt API is available in Linux?  if it is, why can't you
compile the code to run under W2K?  If it isn't. why can't you use
whatever you are using in Linux for W2K?  Look, I know there may be
good technical answers to those questions.  The problem with your
responses on W2K is that the responses just lead to more questions.

What you need to do is go beyond the questions and say "Here is the
source code, suit yourself."  You do say that below, but you don't
deliver the source.

>There are certain APIs that we want to use that we can't.
>Supporting W2K would cause us extra work, not just to get it working but to
>keep the tree W2K friendly (if we could even test that.)

But you cannot use proprietary APIs in GPL-protected software, unless
they are normally distributed with the platform.  It does not matter
what you want or why you want it.  You have to make the source
available and you cannot do that with proprietary code.

>I don't understand where you are going with free as in beer.

That is alright. Just click on the next link and I will guide you
through it:

http://www.gnu.org/

Do you know what "GNU" stands for?  It means "GNU is Not Unix".  The
recurrsion is neat, huh?  That was Richard Stalman's invention when he
started the free software movement.  He started it to free software
from Digital Equiptment Corporation and Unix.  That is why we have
Linux now.  

The link is to the GNU operating system.  "Linux" is a particular
kernel written for GNU.  The popularity of that kernel has caused the
entire operating system and accessory programs to be called Linux. But
you really need to understand that I was not making any reference to
beer- I quoted what free software means according to the front page of
the real, honest to God, official "Linux" operating system.  Yes, the
same one Google claims Chrome will support.

> It's released
>under the BSD license. 

No, that is only partially true.  Every line of code written by Google
IS covered by the BSD license.  But Chrome is a work derived from many
components, including several that are covered by the GPL.  One of the
most important aspects of the GPL is that it covers derived works. The
GPL is the most restrictive license, so it covers everything in the
final product.

(http://code.google.com/chromium/terms.html again)

That means that Chrome, as a whole, is subject to the GPL.  The GPL
does not allow the distribution of derived works that include
proprietary code.  As you explain above, you are using proprietary
Microsoft libraries in Chrome, and that is a violation of the GPL and
the copyright on every line of code subject to the GPL, which is every
line in every component subject to the GPL, and every line that
Chromium wrote because Google chose to incorporate GPL-restricted code
into Chrome.  

Your violation does not go the other way.  For example, Ross Johnson
has not violated the GPL for Pthreads just because you included it in
a derived work that violates the GPL.

Do you follow that?  The BSD license is less restrictive than the GPL.
But when any code is mixed with GPL code, the result is then subject
to the GPL.  To remove the influence of the GPL requires that the GPL
restricted code be removed from the derived work.

So there is your choice.  Dump the proprietary Microsoft code or dump
the GPL-restricted code.  You cannot have both.

>That's about as free as you can get. You are free to
>take the source code and make it work on Windows 2000 - nothing is
>preventing you there. 

Yes, the BSD license is as free as it reasonably gets.  Most of the
others are variations on that theme.  It is only the GPL and LGPL that
really result in free software in the sense of GNU.  The Mozilla code
is also covered by the GPL.  The analysis I made for Mr. Walker (see
below) is an approximately 500K PDF file, but it contains all of the
licenses, if anyone is interested.

Chrome does not meet the requirements of the GPL because you don't
supply the source for what is in the proprietary libraries.  Your
license from Microsoft does not allow you to do that, assuming that
Microsoft supplies the source in a SDK and not just the binaries- I do
not know the facts about that, but it does not matter because
Microsoft does not allow the free distribution of either the source or
the binaries.  The GPL absolutely requires that the source be
available; that is the whole point of open source.

For your "take the software", Ian, I already tried.  I don't have the
time or inclination to work around the source code that you did not
disclose.  You give me the complete source that you are using in
Chrome, and I will agree that Google has met its duty.  But it is
insulting for you to say, in effect, "go recreate the proprietary
Microsoft libraries we used".  I have a better kernel than Microsoft
does, and my libraries are better desgned, more efficient, and already
are independent of all code not proprietary to me.  I am not
interested in any part of Microsoft or its code. 

 It is also deceitful to use those libraries and say your product,
Chrome, is open source.

>We are just saying that we don't want to take on the
>burden of supporting that in our tree. (You can take your beer and do what
>you want, but don't expect the brewer to take your modified recipe and start
>distributing it when the brewer thinks it's going to hurt him to do so.
></bad beer analogy>)

OK, I understand what you mean, but now you are insulting the entire
Linux world, although you do not realize it.  You "take your browser"
and click here:

http://www.gnu.org/

Now scroll down past the picture of Stephen Fry and the "What is GNU"
item on the left.  One click on the scroll bar was enough for me.  You
should now see the "What is Free Software" item that I quoted on the
left.  You thought I made up that saying and you insulted me.  You
really insulted the entire Linux communiy, because that quaint little
saying has a point, and Linux is the result of that point.

>I'm sorry you're not happy, 

No you are not.  Your team wants to pretend that it is part of the
open source community, but you want to dictate solutions just as if
you were the thugs in Redmond.  Well, maybe that is a bit strong, but
you probably get the point.

If you were really sorry, you would not have made me unhappy by using
proprietart Microsoft code.  Instead, you made decisions that may well
have been pragmatic and well reasoned.  Now you have to make decisions
that will result in Chrome being GPL-compliant.

>but please just let this thread die. 

Last night, when thinking about how it could be that you could support
the Linux kernel and the XP kernel, but not the W2K kernel, I realized
that you had to be violating the GPL by mixing proprietary Microsoft
software with the open source GNU software.  I contacted Peter Kasting
privately, because he had pretended to have authoritative knowledge
about supporting W2K, and he was more rude than you and refused to
even refer me to anyone with real authority.  He insisted that this
discussion be public where it can be seen by the bloggers, and that is
why I replied to the two posts here.  Now you want to stiffle the
public discussion entirely, but that is not going to happen now.  I
tried that already, lost the battle, and here we are.

I had to spend quite a bit of time today finding the email address of
Kent Walker because of Mr. Kasting's obstruction.  I left a telephone
message and I wrote the kind of careful explanation that is necessary
when lawyers get involved and emailed it to Mr. Walker.

I heard nothing from Mr. Walker, or anyone else in Mountain View or
any other part of Google, apart from your post, so I am going to see
if Mr. Fry cares about the foundation of his organization, which is
the foundation of the Linux world.

>We've done
>our best to explain the reasons that we are not supporting Win2K, and I
>don't think that it's reasonable to expect us to do something that we don't
>feel is in the best interest of the project.

In truth, I do think you have done your best to explain.  It is a real
shame that nobody on your team, apparently not even the Linux
developers, understood that you cannot just take other people's work
and use it how you want, even when your intentions are good.  When you
use GPL-restricted code, your work becomes GPL-restricted, and you
have to use it according to the GPL.  The GPL protects the freedom of
the software, not your  perogatives, or even the perogatives of the
original author.  Software is not free when it is chained to
proprietary Microsoft code.  I had trouble understanding that free
software concept at first too, but it is as it says here:

http://www.gnu.org/

Do you want a very public showdown between Google and the open source
community?  Peter has the analysis I did last night trying to get some
reasonable dialog with your team.  His claim to being a professional
sofware developer is clouded by his inability to understand software
copyright concepts.  Some of us own our creative work and we do care
about copyright, licenses, and know how to read them.

Mr. Walker has a more detailed analysis, and very shortly Mr. Fry will
also have it and copies of these emails.  I think the world just
assumed that Google had written all the software in Chrome, or
obtained it from other open sources, because that is what you thought
and said.  Do you think they will be surprised to find out that Chrome
depends on proprietary Microsoft code, and that to work on Chrome you
have to buy an expensive Microsoft compiler, and then your work will
be tainted just as Chrome is?  I think that is news.

If you want this thread to die without ever explaining how you plan on
relieving Chrome of proprietary code, it is up to you, and the rest of
the list.  Maybe some blogger is writing his scoop as you read this. I
think Mr. Fry, Mr. Johnson, and Mozilla should know first.

In my opinion, it will be enough for Google to simply comit to doing
whatever it takes to get rid of the proprietary Microsoft code.  I do
not think it is necessary to stop work until that is done.  Hire an
additional team to remove the Microsoft code if you want- just don't
expect me to do it for free.

Everybody who looks at Chrome should respect the work that has gone
into it.  If Google makes the world believe that the production
version of Chrome will really be open source by complying with the
GPL, I don't see what the GPL copyright holders would gain by
obstructing development.  

But that is just my opinion.  I am not a copyright holder and
therefore do not have standing to file suit.  Google has deep pockets,
so you really should not be tempting people to find reasons to sue. 

Also, before you you blow me off like Mr. kasting did, do you remember
this blog:

http://tapthehive.com/discuss/This_Post_Not_Made_In_Chrome_Google_s_E..

Tap the Hive retracted their initial comments when the Google License
Agreement was corrected.  I think the story of Google violating the
GPL will have a lot more traction. 

Do you have anybody on your team that actually understands what GNU is
about?  If you do, that person might be a good one to continue this
discussion with the plan Google has for making Chrome open source.

If you do contemplate responding, first look at the end of Section 3
in the GPLv2.  That will give you a glimer of hope that using the
proprietary Microsoft libraries is permitted.  But consider the words
"normally distributed".  Now consider that I have W2K Server with MS
C++ v6, XP with C++ v9 Express, and W2K3 Server with C++ v8 Express.
In no case do I have the proprietary Microsoft libraries that you are
using in Chrome.  I can have them if I give the predator in Redmond
about $800, but I think that is not "normal distribution" of any of
the platforms I mentioned.

Before making up your mind what GPLv2 Section 3 really means, look at
the definition of "system library" in Section 1 of GPLv3.  GPLv3
resolves the ambiguity in GPLv2 in the way I explained.  While I
presume Mr. Walker is thinking how to twist the words of GPLv2 to
justify gutting the meaning of free software (that is what corporate
lawyers are for, sometimes), I'm betting that Mr. Fry and the people
he represents see it my way.

You can find the GPL versions on the GNU site- that is where they were
invented, along with the free beer / freedom of speach analogy.

http://www.gnu.org/

Good site, good ideas, have a look.

I think I have explained my viewpoint.  If you have any questions, I
will do my best to answer.  Otherwise I think my role has come to an
end and all that is left to do is see if anybody cares.

Ken

>
>Thanks.
>
>On Thu, Sep 11, 2008 at 2:12 PM, Ken Berry <[EMAIL PROTECTED]> wrote:
>
>>
>> Peter,
>>
>> Your analysis does not make sense, if Google is really going to
>> support many platforms, particularly Linux.  There is little
>> difference between the W2K and XP kernels.  There is a substantial
>> difference between the XP and Linux kernels.
>>
>> To make sense of your argument, it must be supposed that there will be
>> large differences between the versions for each platform, with the
>> Windows version continuing to depend on proprietary Microsoft
>> libraries.
>>
>> That leads to a question of whether Google is at all sincere in
>> developing a platform-independent browser.  If you were, you would
>> maximize the amount of common code and minimize the amount of code
>> that depended on a particular platform.  But basing Chrome on
>> proprietary Microsoft libraries is doing just the opposite.  You will
>> have to create the functionality of those libraries under Linux.  If
>> you did base the functionality on the kernel, there would not be a
>> significant difference between W2K and XP.
>>
>> I would like the ability to look at the dependancies to understand
>> what the common code needs from the host OS, but I have to interpret
>> your excuse as indicating that Chrome is heavily dependent on
>> proprietary features Microsoft has built into XP and Vista, and that
>> is precisely what I want to avoid.
>>
>> I have taken your advice, and in spite of your refusal to help, I did
>> find Kent Walker's email address.  I hope he takes copyright
>> infringement and misrepresentation more seriously than you.
>>
>> I hope there is truly an open source version available that works on a
>> platform other than Windows.  I'll take another look then.
>>
>> Ken
>>
>> On Wed, 10 Sep 2008 12:56:39 -0700, you wrote:
>>
>> >On Wed, Sep 10, 2008 at 12:44 PM, burgersoft777 <[EMAIL PROTECTED]
>> >wrote:
>> >
>> >> When you take account of the size of
>> >> the 2k user base I would have thought goggle would have bent over
>> >> backwards to help. Its not to late to change direction and do
>> >> something for the users rather than act as an agent of MS effectively
>> >> attempting to fool users into changing an OS that remains more than up
>> >> to the task of running your browser.
>> >
>> >
>> >Much as I love a good conspiracy theory, acting as an agent of
>> >Microsoft wasn't the deciding factor here.  Win2k lacks a large number of
>> >APIs Chromium uses, especially with regard to the sandbox (but also in a
>> >number of other core and UI functions).  We intended to support win2k for
>> a
>> >long time and only changed our minds after it became clear that the
>> >engineering cost would be very large and ongoing (despite your assertions
>> to
>> >the contrary).
>> >
>> >Also, the size of the win2k user base is not, in fact, particularly
>> >large, and many of these users are in locked-down corporate environments
>> >where Chromium will have little penetration.
>> >
>> >I don't have a ton of sympathy here.  Electing to run an
>> >eight-year-old operating system is a choice that carries many tradeoffs
>> with
>> >it, including no support from the manufacturer and an inability to run
>> >various newer programs, including Chromium.  I think there are more
>> >impactful changes we can spend our effort on as a development community.
>> > Obviously, if a fully-working port and an ongoing maintainer are both
>> >available, the cost of this choice is reduced, and the decision might
>> >change.
>> >
>> >PK
>> >
>> >
>>
>> >
>>
>
>

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups 
"Chromium-dev" group.
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to [EMAIL PROTECTED]
For more options, visit this group at 
http://groups.google.com/group/chromium-dev?hl=en
-~----------~----~----~----~------~----~------~--~---

Reply via email to