Discussion:
Chromium 10 ported to Solaris 11
Ruben
2011-04-11 06:54:11 UTC
Permalink
Hello, I recently ported Chromium 10, the open source base of the
current Google Chrome 10 stable browser, to Solaris 5.11 snv_151a. A
test build for i386 is available here:

http://chromium.hybridsource.org

There are a couple issues I'm still ironing out, as written in the
notes. Thanks to James Choi for his early patches from last year that
got this Solaris port going.
Alasdair Lumsden
2011-04-11 08:49:12 UTC
Permalink
Hi Ruben,
Post by Ruben
Hello, I recently ported Chromium 10, the open source base of the
current Google Chrome 10 stable browser, to Solaris 5.11 snv_151a. A
http://chromium.hybridsource.org
There are a couple issues I'm still ironing out, as written in the
notes. Thanks to James Choi for his early patches from last year that
got this Solaris port going.
That's great news indeed! The OpenIndiana project would love to support
your work, and we'd love to make this available to "pkg install" once
the major issues are worked out.

Would you be interested in a permanent OI zone for development work?

Cheers,

Alasdair
Alasdair Lumsden
2011-04-11 09:31:15 UTC
Permalink
Hi Ruben,
Post by Ruben
http://chromium.hybridsource.org
There are a couple issues I'm still ironing out, as written in the
notes. Thanks to James Choi for his early patches from last year that
got this Solaris port going.
It appears the build was performed with the OpenCSW gcc - does chrome
refuse to build with the stock gcc 3.4.3?

Regards,

Alasdair
Guido Berhoerster
2011-04-11 10:13:22 UTC
Permalink
Post by Alasdair Lumsden
Hi Ruben,
Post by Ruben
http://chromium.hybridsource.org
There are a couple issues I'm still ironing out, as written in the
notes. Thanks to James Choi for his early patches from last year that
got this Solaris port going.
It appears the build was performed with the OpenCSW gcc - does
chrome refuse to build with the stock gcc 3.4.3?
gcc 4.5.2 is available from SFE if needed. A bigger question is
whether the patchset will be made available free of charge.
--
Guido Berhoerster
Gordon Ross
2011-04-11 15:39:26 UTC
Permalink
Post by Ruben
Hello, I recently ported Chromium 10, the open source base of the
current Google Chrome 10 stable browser, to Solaris 5.11 snv_151a. A
http://chromium.hybridsource.org
There are a couple issues I'm still ironing out, as written in the
notes. Thanks to James Choi for his early patches from last year that
got this Solaris port going.
Great work - Thanks!

If you will publish your patches somewhere, others can help
finish up any remaining details, optional features, etc.
James Choi
2011-04-11 22:27:49 UTC
Permalink
Ruben!

First FreeBSD, now Solaris! And it looks like you've solved all the
difficult problems for the port. You are a Chromium ninja!
Post by Ruben
http://chromium.hybridsource.org
There are a couple issues I'm still ironing out, as written in the
notes.  Thanks to James Choi for his early patches from last year that
got this Solaris port going.
_______________________________________________
desktop-discuss mailing list
Nikola M.
2011-04-13 02:15:45 UTC
Permalink
Post by James Choi
First FreeBSD, now Solaris! And it looks like you've solved all the
difficult problems for the port. You are a Chromium ninja!
I guess "best" thing about this "Stealium" is turning product that is
for unrestricted use and under open source license into binary-only
restricted pay-for only build, not worth to be called by Chromium name.
And trying to push changes depending on he's closed source binary parts
upstream,
classify it as very bad behavior (submarine code) and might hinder true
open porting effort.

That build is for people foolish enough to accept restricted and
unchecked/unaudited
spyware that is maybe included in this build on their
Open/Solaris/Express boxes.
Hans J. Albertsson
2011-04-11 08:49:24 UTC
Permalink
Does this mean Google Chrome can be made available for Solaris later on?
Post by Ruben
Hello, I recently ported Chromium 10, the open source base of the
current Google Chrome 10 stable browser, to Solaris 5.11 snv_151a. A
http://chromium.hybridsource.org
There are a couple issues I'm still ironing out, as written in the
notes. Thanks to James Choi for his early patches from last year that
got this Solaris port going.
_______________________________________________
desktop-discuss mailing list
Hans J. Albertsson
2011-04-11 08:57:31 UTC
Permalink
Oh. built for b151
Post by Hans J. Albertsson
Does this mean Google Chrome can be made available for Solaris later on?
Post by Ruben
Hello, I recently ported Chromium 10, the open source base of the
current Google Chrome 10 stable browser, to Solaris 5.11 snv_151a. A
http://chromium.hybridsource.org
There are a couple issues I'm still ironing out, as written in the
notes. Thanks to James Choi for his early patches from last year that
got this Solaris port going.
_______________________________________________
desktop-discuss mailing list
_______________________________________________
desktop-discuss mailing list
Ruben
2011-04-12 03:00:16 UTC
Permalink
Alasdair, I'll test my Solaris builds out on OpenIndiana at some point,
let me know if you have any problem running this test build on OI. As
for gcc 3.4, I didn't bother trying it since Chromium has listed gcc 4
as a dependency since the beginning:

http://code.google.com/p/chromium/wiki/LinuxBuildInstructionsPrerequisites

I will not be open sourcing all my patches right now, but I do
periodically spin off portions of my patches and submit them upstream.
The plan is to offer paid subscriptions for Solaris to fund this porting
effort, just as I do for other Unix systems now.

Hans, Google Chrome is 99% the same as Chromium, just a few extra closed
bits added:

http://code.google.com/p/chromium/wiki/ChromiumBrowserVsGoogleChrome

So if Google were to port those closed bits also, there's no reason
there couldn't be a Google Chrome for Solaris. However, it appears that
they have no plans to support anything other than linux right now.

James, thanks for your enthusiasm :) and early patches for this port.
Nikola M.
2011-04-12 09:25:17 UTC
Permalink
Post by Ruben
I will not be open sourcing all my patches right now
So you actually made closed source fork of Chromium.
And you put your binary build as a commercial to the list.

I doubt you have rights to call it Chromium then,
since it is clear that Chromium brand name is not at your possession
and second, Chromium by default is open source project
and base for Google Chrome and Chromium browser in all those Linux
distributions they have it ported in their repositories right now.

So you actually want to make money out of distributing binary port.
That is quite fine.
Just release source and I think people will be more then happy to
contribute to your porting project. And you can continue to sell it if
no one have objections over Chromium brand name.

You do not need to make binary available for free, sell it freely
(IF it is allowed to you by license)
Just put your source changes there somewhere with build instructions, so
OI folks and anyone else can build it fully in open. (share knowledge)

If otherwise, your porting effort goes against free and open source
software implemented in Illumos and OpenIndiana.
Your Chromium with closed parts is a threat to rights of free
distribution of OI and at the end, it is unlikely that closed source
product like that might end up in OI repositories/publishers unless full
open sourced.

Illumos project is changing and replacing closed parts in OS/net and
adding closed things on top of other distributions is like shooting
yourself in a foot, actually.
It is not only open vs closed, it also the case of security - Not having
your arbitrary closed parts in binary releases.

And making it under closed and usage restricted Oracle Solaris Express
primary,
but collaborating to the open source OpenSolaris based
Illumos/OpenIndiana folks also does not help advocating your effort. But
may be helping your selling, sure.

Maybe then you could ask Oracle for porting sponsorship and solve your
financial support problems, right now. I guess they might be willing to
help you advocating their platforms together.
For the lack of Chromium brand name usage rights, you can contact Google
directly.
Post by Ruben
but I do periodically spin off portions of my patches and submit them
upstream.
I suppose that might be from 1-5 years time, and in between, no one can
actually make Chromium to compile it and use it. And there are no
insurances you will ever distribute your source changes to the public in
full, beside this e-mail/message.

I only guess, you maybe do not even have rights to distribute changes in
the first place?
Maybe even you do not have rights to those binary parts added to your build?

I'll be happy to support OPEN SOURCE porting project with donations,
fund raising or something, (or Google and Oracle with selling those fine
shiny servers to the enterprises might do that) just I do not like
anyone to force me to support closed source things I am forced to support.

I wonder, how would anyone react if one would make fork of X server ,
GNOME, Libre/OpenOffice, Firefox, Pidgin, Thunderbird, adding to it
closed and binary parts and try to sell it with the _same name_ like
original free software and open source products?
I guess reaction would be not quite positive and legal.

We had such attempt locally, someone tried to sell OpenOffice via SMS to
windows users without releasing same-licensed source changes. That is
,like a NO-NO thing.

I guess someone might say that you are basically doing what Google is doing.
Well, Google is not selling Chrome. And Google have rights to core Chromium.
And even selling is not the point it might be OK under license,
but problem are binary additions to the open source product, who's
sources are unavailable to the public inspection, expansion,
contribution and future development.

Therefore, anyone thinking that your Chromium port is a solution might
be stuck with your code, depending on closed parts you might never
release. (That is OpenCore on top of Open..) And being insecure by
default is Microsoft's job, not Solaris folks.
Orvar Korvar
2011-04-16 14:29:25 UTC
Permalink
Im not going to use closed binaries made by a random guy. could be virus. ok if google/oracle/... makes closed binaries, but not a private person. too risky
--
This message posted from opensolaris.org
Jerry Kemp
2011-04-16 20:51:39 UTC
Permalink
Solaris virii??? That'll be the day.

If you are going to beat the guy up over his efforts, that's one thing,
but at least come up with a plausible reason.

Open source. Most admins don't have the skill or persistence to go
through the source code of a large project and find malicious code in
the first place.

Chrome has been out for some time, and this has been the first
real/successful effort that I have seen to get Chrome on our "chosen
platform" to date.

I am a Firefox user, but I am envious of the skills required to get the
port as far along as it is.

.......

To James Choi and Ruben. Thank you for your porting efforts. It is
appreciated.

Jerry
Post by Orvar Korvar
Im not going to use closed binaries made by a random guy. could be
virus. ok if google/oracle/... makes closed binaries, but not a private
person. too risky
Nikola M.
2011-04-17 00:37:11 UTC
Permalink
Post by Jerry Kemp
Solaris virii??? That'll be the day.
Exactly is what is proposed with this closed source binary release of a
one man.
Post by Jerry Kemp
If you are going to beat the guy up over his efforts, that's one thing,
but at least come up with a plausible reason.
Developing open software and porting is not a battle. It is more like
learning and exchanging ideas and making life of of other porters better
by sharing findings.

Best reason is: Guys intentions are not clear neither look promising in
the future.
Post by Jerry Kemp
Open source. Most admins don't have the skill or persistence to go
through the source code of a large project and find malicious code in
the first place.
Exactly, that is why Admins should go and use only audited software that
someone else (maintainer) patched, changed and who's changes are audit
able from a trusted party and visible by others.
Aether some guy that is actually looking at the source code before
putting it out
or a company that have employed people for doing that (Oracle) who you
maybe could potentially trust, while source changes are also released.
Post by Jerry Kemp
Chrome has been out for some time, and this has been the first
real/successful effort that I have seen to get Chrome on our "chosen
platform" to date.
Unfortunately, this binary build/blob is clearly made from guy who's
perusing small temporary personal gain and turning inside out the level
of trustfulness that open source tools and applications give to it's users.

There is nothing wrong making money.
But there IS something terribly wrong smelling around here, with closing
source.

If he made he's build for sale, while also contributing sources (in the
same way he actually got real source code from REAL Chromium), I would
personally advocate to pay the guy for it's porting and packaging
effort, providing a binary release.
That way he wold be CREDIBLE source, someone you can trust.
This way, You can not possibly trust this guy's intentions, because it's
binary is Closed, unaudited and therefore, too risky

It is making Closed source out of Open in the clear light.
That binary is NOT a Chromium. Chromium is open source.
It is sort of binary that looks and smell like it, but you actually
never know what is under hood, until you have ABILITY to test it by
taking full audited open source and with possibility to build it
yourself, without closed undefined UFO parts that came from no one knows
where.
Post by Jerry Kemp
I am a Firefox user, but I am envious of the skills required to get the
port as far along as it is.
It would be great to learn, but obviously that kind of people can only
teach bad and unsustainable market behavior.
Post by Jerry Kemp
To James Choi and Ruben. Thank you for your porting efforts. It is
appreciated.
I guess level of uncertainty that this particular closed source build of open software and possible problems with it induces in short and long term, disqualify it from ever using it.

We will yet to see source code changes in the clear light and binary build that can be built by anyone else.
Using binary source from This guy is at least very bad idea.
If I have to use binary build, I would definitely not use this guy's binary. Never.
I would rather wait for source changes to "land" (if ever) and to see independent and open build Everyone can use. And providing that sort of "open" changes this Bag of tricks could provide , does not always intentionally DEPEND on it's new closed addons!..
And since he will continue to add closed things to it's build without releasing them, with this kind of behavior he will always REMAIN UNTRUSTED source.

If you value your life/business/job/Administration practice and have stepped away from
"Everything closed, lets use untrusted sources programs from internet, and 'Antivirus'": Microsoft way of doing things, then you will say:
Not in here, not now. Not on my Solaris. (Or OpenIndiana/Illumos, whatever you get to have installed)
Alan Coopersmith
2011-04-17 15:47:27 UTC
Permalink
Post by Nikola M.
Exactly, that is why Admins should go and use only audited software that
someone else (maintainer) patched, changed and who's changes are audit
able from a trusted party and visible by others.
Aether some guy that is actually looking at the source code before
putting it out
or a company that have employed people for doing that (Oracle) who you
maybe could potentially trust, while source changes are also released.
Most packagers, whether individuals working on projects like SFE or employees
of companies like Sun & Oracle, don't have the time (or in many cases the
knowledge) to audit all the source code they're building. I know I've not
looked at all of the millions of lines of code I brought in from open source
projects and integrated to Solaris, and I know I'm not alone.

The whole open source world is held together by distributed webs of trust.
You trust Oracle or OpenIndiana to build and deliver X packages from the
sources I provide. I trust the developers at X.Org to not intentionally
put malicious code in the system (though I know they're human and bugs will
be there). The X.Org developers trust each other to vet new code and
new people being considered for commit access, and trust the developers of
tools like git & gcc to not insert trojan horses into our code via our
use of them. The list could go on for a long time - see the classic paper
"Reflections on Trusting Trust" by Unix co-creator Ken Thompson.
( http://cm.bell-labs.com/who/ken/trust.html )
--
-Alan Coopersmith- alan.coopersmith-QHcLZuEGTsvQT0dZR+***@public.gmane.org
Oracle Solaris Platform Engineering: X Window System
Nikola M.
2011-04-18 08:02:31 UTC
Permalink
Post by Alan Coopersmith
Post by Nikola M.
Exactly, that is why Admins should go and use only audited software that
someone else (maintainer) patched, changed and who's changes are audit
able from a trusted party and visible by others.
Aether some guy that is actually looking at the source code before
putting it out
or a company that have employed people for doing that (Oracle) who you
maybe could potentially trust, while source changes are also released.
Most packagers, whether individuals working on projects like SFE or employees
of companies like Sun & Oracle, don't have the time (or in many cases the
knowledge) to audit all the source code they're building. I know I've not
looked at all of the millions of lines of code I brought in from open source
projects and integrated to Solaris, and I know I'm not alone.
The whole open source world is held together by distributed webs of trust.
You trust Oracle or OpenIndiana to build and deliver X packages from the
sources I provide. I trust the developers at X.Org to not intentionally
put malicious code in the system (though I know they're human and bugs will
be there). The X.Org developers trust each other to vet new code and
new people being considered for commit access, and trust the developers of
tools like git & gcc to not insert trojan horses into our code via our
use of them. The list could go on for a long time - see the classic paper
"Reflections on Trusting Trust" by Unix co-creator Ken Thompson.
( http://cm.bell-labs.com/who/ken/trust.html )
Thank you Alan.
That is exactly what is important. No ode can audit everything alone. So
we rely on per-patch auditing of changes to in the end have software
that is maintainable and audited as a result.

Undisturbed chain of trust and being dependent on someone maintaining
software upstream, before user, and on the other sides, is what
everything relies on.
Auditing those changes as they come, regarding he's or she's project and
checking them out when they go upstream or are changed from upstream
when porting.

Time and money needed to re-link the chain of trust like this again is
too great,
from any personal gain or reason of breaking it intentionally.

Loading...