Issue 117: Linux support

[Freddi]

As suggested by jlivni on Google Code, I bring that issue to this
group.
http://code.google.com/p/earth-api-samples/issues/detail?id=117
I really don't want to sound nerving.

The word "future" has it that it's never now. We ("543 people starred
this issue") have been waiting over two years, when the Google Earth
Plugin was first mentioned to come "in future" to Linux.
Although the activity of Google Guides on that issue was rather low, I
appreciate very much that you still see it planned - only that there
is not yet a timeline. But I'm affraid that this doesn't anymore sound
to most people as serious as you meant it to be understood.

We don't have to be geeks or programmers to decide using Linux or
Ubuntu as our operating system. I just like the way how easy and
secure some things are. On the other side, _I'm_used_to_it_ that there
is not much support from commercial software manufacturers.

Now the point is that Joe Public coming from Windows or OS X
_is_not_used_ to such a bad support, he can't understand why he can't
just use any software. I have several friends who got a chrome netbook
for beta testing and they actually expected to see Google Maps in 3d
or to adopt the cr48 as a development tool - for developing GE-plugin
applications. The more Chrome OS will be widespread, the more will
this become an issue of ordinary people that noone can just put down
as "geeks".

I have patience and I'm still looking into the undefined "future" ;-)

[Earth PM Peter]

Hi Freddi,

I know that it is frustrating to not have support for the Google Earth plug-in for Linux, and to be waiting for an indefinite and undefined "future".  The reality is that getting the Earth plug-in to work on Linux desktops is not a trivial effort, and we have had to make trade-offs given that there are only so many things we can do at once.  Over the past couple of years, we have had to add support for multiple new browser/OS combinations, including IE8, IE9, Firefox 4, Chrome, Windows 7, and Snow Leopard.  We did some major refactoring to take advantage of new APIs in Snow Leopard to significantly improve Mac plug-in performance.  On the Linux side, we have done a major repackaging of the client binary using LSB, in order to address many of the problems our desktop Linux users have had with incompatible distributions.  

Another thing to note is that while ChromeOS is based on Linux, it is very different from a standard desktop Linux OS such as Red Hat or Ubuntu.  It is a completely different operating system which is build around the Chrome browser, and standard plugins are not supported.  We are definitely interested in supporting ChromeOS, but this is a significant undertaking, and the native code support for ChromeOS (Native Client) is still under development.

We are working on this, and I hope that we will have better news for you in the coming quarters.  I apologize that I don't have a better answer for you now.

peter

[Chad Burt]

Seems like the end goal would be to port the Earth API to a standards-based approach like WebGL, then in theory it's just up to the browser vendors to fully implement the spec. Not trivial obviously since these apis aren't in common usage now. A big project like Google Earth could surely spur adoption though.

[oneman...@googlemail.com]

I think the point is that bringing this to linux desktops would open up a world of development and support that is sorely missing at the moment.
The fact that you make a point of trying to distance yourself from the geek image is proof that people who choose linux, like it or not, are a bunch of geeks. Sad fools like you or I chose to use /bin/gnu/linux/debian/ubuntu/etc/distributions for a reason. I have never in all my years (nearly 30 now) of messing with electronic processing devices come across such a wide ranging, exciting, cutting edge technology, with such an active helpful friendly willing bunch of REAL people supporting it at all levels, from the kernel artists to the graphic hackers.

Yes, we know how to run win VM's, but If we were to hold some form of election, the amount of us who want to run windows would be close to zero. and it's not just us:

I am writing this from a youth hostel in Granada (south Spain if you were curious). There are 2 desktop machines here for the guests to use (most of them spend alot of time on facebook). They run ubuntu. My girlfriend has an 11 year old daughter. Like prety much every other schoolkid in andalucia she has been given her very own netbook computer. It runs a gnome desktop on what they call Gaudalinex. If you press ctrl-alt-F1 you won't be all that surprised to see a log-in terminal which says: Welcome to Ubuntu! Every government official from the tax office (Hacienda) on the next street, to the town hall across gran via has a desktop computer running the same thing. And it's not just spain (pop.about 80million) all over europe, governments, businesses, and private individuals are using linux desktops. The last time I saw a macintosh it was a us-american tourist.

I read something recently that some CEO of Canonical estimated there were 12million instances of ubuntu in use worldwide. This is a vast under-estimate. There may be 12 million people like you and I who have chosen to install and use the best GNU/LINUX distribution for their own reasons. Every university in the world with a computer science department has people learning and understanding GNU/LINUX. The argument about different distributions doesn't hold any water, just get it working for debian. The rest will follow.

The fact that only 547 english speakers have made the effort to express their dissatisfaction about the lack of linux support from google-earth, particularly the browser plug-in means that there are talented, creative, knowledgeable people who want to bring an immense amount of positive work to the project. And at least 547 of us will not do it while google expects us to use windows for testing, and ignore all of the end users who have GNU/LINUX desktops.

Please start giving us some proper answers. If it's a commercial based decision, it's a bad one, but please tell us anyway. If it's a technical issue, let us know. 547 knowledgeable hobbyists/developers/hackers want to help. The fact that we can do nothing or know nothing is just a bit more than 'frustrating' as EarthPM Peter says.

Moving the issue away from a forum, cancelling comments, etc is just plain evil.

[oneman...@googlemail.com]

Oh, and peter, Ive just been looking at LSB. good common sense again
from the linux foundation to ensure portability, but what exactly are
you talking about with client binaries and packaging? We need the
browser plugin.
ChromeOS is uninteresting because it doesn't already have a user base
of hundreds of millions, and it's not serious as a developer platform.

[oneman...@googlemail.com]

Yeah, and just noticed that you say chromeOS is based on linux, but completley different. How so? Different process scheduler? different virtual memmory management? different file system management? different IO layers? tcp/ip stack? I don't think so. Why would there be any need for that. please just make it work for chrome or mozilla gecko debian.
As you said: There are only so many things you can do at once, so do the most important ones first, and if necessary, ask for help.

[Freddi]

ChromeOS does not give the possibility to install any binary software,
but as it comes with some binary plugins (flash, pdf?, native client?)
preinstalled under the hood, I had thought it's not too much beside
the point that it would provide a more complete experience of Google
services and Google's earth vision.
Trying to make Google Earth work on all distributions is in my eyes a
very huge and ambitious project (they could also just make it work
with standards and debian, forcing the excrescence of smaller
distributions to go compliant with bigger distributions). The move of
the Google Earth client to LSB might be a step in that direction. Now
as there exists a native client for some years, the problem of
creating a native plugin isn't the same as creating a completely new
linux port of an application.
Unfortunately Google Earth is not open source (in contrast to chromium
etc.) so there is much will but not much chance to help Google. One
opportunity for Google to get some help could be the GOOGLE SUMMER OF
CODE.

I think the most current numbers on canonical.com is something around
15 millions (growing 2-3 mio. per year, but that doesn't say much
about Ubuntu-only users or now-and-then users).

I once made a Google Earth Plugin web app, (programming blindly in the
dark, and sometimes testing on windows). But as I almost never used
it, I had not much motivation to do more (although I like Google
Earth).

[Greg-WBSR]

'There are only so many things you can do at once, so do the
most important ones first' - best quote

In my opinion - Google should spend more time developing Google Earth
for mobile platforms rather than linux - that is where the world is
heading and that is where they have a lot more people to provide for.
The last poster mentions a number of linux users at 15 million (is
that all? - must be more than that?) - regardless, it is a nice number
in this case, because 1 telecom company alone sold 15 million iPhones
last year in the US (numbers unconfirmed but similar numbers get
repeated on various websites) - now add in other countries iPhones
sales and then all the Android / BB sales and you have a lot more
users.

It is a shame that an iPhone can link Google Earth with a kml in
Google My Maps - but an Android phone cannot.

That's my $0.02

Google Earth Rocks!
Greg

p.s. I have an Android phone, and a 'use' in development, but have to
use a friends iPhone to see how it is working.
p.p.s I also just saw a 'statement' claiming the number of iPhones
sold since it's release is over 50 million

[Freddi]

Now as Google's Native Client API is almost complete, there would be
another way to realize more platform-independence.

I see the difficulties to port an application to any new platform,
also that it depends on the size of the userbase whether it's worth
the effort. As far as I understood, Native Client aims to run binaries/
C-code etc. in any browser on any platform. So the thought suggests
itself that ChromeOS will not be limited to Javascript-webapps only.
Maybe there is no reason why it shouldn't be capable of things that
browser plugins do up to now, without conflicting with its security
principles.
Native Client with binary code would not only be a solution for
performance concerns (like pure webGL/javascript), but it would also
put the GEplugin on a sound footing in the long term: A platform-
independent model allowsGoogle to offer one solution even for all
currently supported platforms and browsers.

Regarding mobile platforms: It's always the case that one supported
mobile device makes the userbase of another platform demand it too. No
Google Earth on iOS without Google Earth on Android, and vice versa.
Important to note is that those are not the only mobile platforms,
Symbian and Blackberry exist too, and HP's WebOS is very very
promising. Even if one argues that mobile platforms had a bigger
userbase than desktop linux, they cause also an increasing variety of
operating systems. It's now not anymore the era of a single OS'
monopoly which makes it easy for developers. Instead of adding support
for each single platform (be it a mobile OS or desktop Linux), Google
would do better a pioneering task and lead the industries into an era
of common standards.

Like the body browser webGL demo and many others, Google always
creates some demos to push its new features/ new technology forward.

-----
May I suggest that Google sees another Google Earth plugin port not
only as something "planned for the future" but rather as a SHOWPIECE
of Google's Native Client technology? I think for both purposes, as
presentation of NativeClient and for increasing Google Earth's
platform independence it would be really worth to look into this,
discuss it, the advantages and obstacles etc.
-----

@Greg: "15 million (is that all?)"
Sorry, I meant 15 million users solely of the Ubuntu distribution.

[Jeffrey Seger]

Alright, I am new to this forum, but I came here looking for some answers as to why Linux was not supported under the GE plugin, thinking that it might be a fairly recent plugin and that perhaps support might be right around the corner.  Sadly what I have found is that users have been asking for support for linux for over 2 years. Amazing.  And as for why, so that you could better support Windows and MacOS.  They are getting new, upgraded versions, while we get excuses.  I am sorry but this feels very much like a raised middle finger as you turn away from us toward the masses.  Yes, I understand that getting support for those OS' first may have been more commercially important to you.  No, I do not understand why new versions and features for them is more important than basic support for us.

Linux did not make Google what it is, but it sure helped.  Show some loyalty and quit ignoring us.

That being said, I have no expectation that this comment will help.  You are a huge corporation and will make decisions based on what you think is in your interest, but I think you are making a huge mistake.

[Gerd]

We have to appreciate that jlivni at least does not keep his fingers away from our wish: He recently merged issue 443 into this - whatever that means. But still we are only acknowledged but have no internal number and there are endless other issues that have internals but have not visibly advanced. Afterall we have already a native Google Earth client, so it's not like if there was no code yet to create the browser plugin.

Initially the Linux version was planned for August 2008 (almost 3 years). We can try to be patient, but we hope this is still seriously on the agenda (and not for fun).

[Freddi]

I agree!

[Gerd]

Any news?

[Freddi]

Have a look at Alexandro's post:
https://groups.google.com/forum/#!topic/google-earth-browser-plugin/tMFmK7CGCqE

[William Beaufoy]

Still waiting for this

[tony]

Any time frame on this yet?

[Bruce Wolfe]

What I don't understand about this statement is that all of Google is built on mostly native open source platform. How is it that at bare minimum it can't be rendered by the same/similar platforms? Plus, Macintosh Snow Leopard was a major shift to a *nix platform for some years now. We are now two upgrades since. 

So, your response is not quite accurate.

Please offer an updated response and timeline on implementation.

Thank you in advance.