Chrome Port for the Mac is Out

I found that real Google Chrome in Windows XP under Parallels is a lot better on my MacBook Pro than Crossover Chromium.

On 9/17/08 9:10 AM, "Fearghas McKay" <fm-listsst-kilda.org> wrote:

>> I don't know if this could be called exactly "native". You don't need
>> Wine itself, but the code is Windows code that uses their "Wine
>> Technology" to quickly port the code over to Mac and Linux. I take it
>> that it uses translator libraries to map Windows calls to Mac calls.
>
> So it is not a native Apple OSX framework on top of BSD & Quartz -
> just a "native" third party OSX framework on top of BSD & Quartz.
>
> Kinda feels like splitting hairs from here... by the way that is is a
> general comment rather than directed at David.

But that creates a lot of issues. For example, can chrome make use of OS
Services? Can you send and receive data with it from other applications?

It seems like splitting hairs, "Well, it runs with a double-click, what more
do you want", but there is really, rather a lot to being a "native"
application.

--
John C. Welch

On Wed, 17 Sep 2008, LewisGmail wrote:

> Too be fair it is a bit of a quandry now. What IS native, really?

Fair enough -- there's clearly a spectrum here, as you yourself have
pointed out in the past (talking about Firefox at the time):

http://emperor.tidbits.com/webx?229.3cc3c9eb/4

Anyway:

> kreme 19276 0.2 0.6 2975108 36896 ?? S 10:51AM
> 0:17.83 /Applications/CrossOver Chromium.app/Contents/SharedSupport/
> cxchromium/lib/../bin/wineloader c:/chromium/chrome.exe --no-sandbox --
> in-process-plugins
> kreme 19172 0.0 0.1 2928592 4740 ?? Ss 10:46AM
> 0:00.60 /Applications/CrossOver Chromium.app/Contents/SharedSupport/
> cxchromium/bin/wineloader winewrapper.exe --wait-children --run -- c:/
> chromium/chrome.exe --no-sandbox --in-process-plugins
>
> There are the two main Chrome processes.

How many tabs were open when you did this check?

Does Chromium-via-WINE do the "one process per tab" thing that the
Windows version does, and if so, is it a Mac process, or just some kind
of memory space within the wineloader process? Hmm, interesting...


--
Chris Devers
DO NOT LEAVE IT IS NOT REAL

On 18/09/2008 1:10 AM, "LewisGmail" <gkremegmail.com> spake thus:

> What IS native, really?

Well, there's native to the _hardware_ (i.e., Intel executable code), which
this port clearly is. This category includes anything that can run on the
base hardware, including natively compiled X11 and Mac OS X (Carbon, Cocoa,
et al.) apps. Java apps may or may not be included in this depending on the
compiler used.

Then there's native to the _OS_, which this port clearly isn't. It doesn't
directly use any of the Mac APIs (i.e., they're not compiled in), instead
calling them indirectly through a runtime API translation layer. I'd also
argue that most Java apps fall into this category.

So, while Chrome runs natively on the hardware, it's NOT a native Mac OS X
app.

--
Nigel Stanger, Dunedin, NEW ZEALAND.
http://xri.net/=nigel.stanger

On 18/09/2008 1:10 AM, "Robert Buice" <rgbuicegmail.com> spake thus:

> but the code is running natively on the processor and not through an emulation
> layer.

So by that argument Windows XP/Vista running under VMWare would count as
"native" (and by extension, so would all Windows software). There's no
emulation, and the code is running natively on the processor.

--
Nigel Stanger, Dunedin, NEW ZEALAND.
http://xri.net/=nigel.stanger

On 18-Sep-2008, at 01:36, Mike Cohen wrote:
> I found that real Google Chrome in Windows XP under Parallels is a
> lot better on my MacBook Pro than Crossover Chromium.

I found Crossover's Chromium locked up and stopped responding after a
while of running. It also seems kinda flakey even when it is running.

On 18-Sep-2008, at 01:36, Chris Devers wrote:
>> kreme 19276 0.2 0.6 2975108 36896 ?? S 10:51AM
>> 0:17.83 /Applications/CrossOver Chromium.app/Contents/SharedSupport/
>> cxchromium/lib/../bin/wineloader c:/chromium/chrome.exe --no-
>> sandbox --
>> in-process-plugins
>> kreme 19172 0.0 0.1 2928592 4740 ?? Ss 10:46AM
>> 0:00.60 /Applications/CrossOver Chromium.app/Contents/SharedSupport/
>> cxchromium/bin/wineloader winewrapper.exe --wait-children --run --
>> c:/
>> chromium/chrome.exe --no-sandbox --in-process-plugins
>>
>> There are the two main Chrome processes.
>
> How many tabs were open when you did this check?

Hmm.. good question. I can redownload it I guess.

> Does Chromium-via-WINE do the "one process per tab" thing that the
> Windows version does, and if so, is it a Mac process, or just some
> kind
> of memory space within the wineloader process? Hmm, interesting...

The entire process is what locks up, not individual tabs.

On 9/18/08 at 2:36 AM, jwelchbynkii.com (John C. Welch) wrote:

>On 9/17/08 9:10 AM, "Fearghas McKay" <fm-listsst-kilda.org> wrote:
>
>>> I don't know if this could be called exactly "native". You don't need
>>> Wine itself, but the code is Windows code that uses their "Wine
>>> Technology" to quickly port the code over to Mac and Linux. I take it
>>> that it uses translator libraries to map Windows calls to Mac calls.
>>
>>So it is not a native Apple OSX framework on top of BSD & Quartz -
>>just a "native" third party OSX framework on top of BSD & Quartz.
>>
>>Kinda feels like splitting hairs from here... by the way that is is a
>>general comment rather than directed at David.
>
>But that creates a lot of issues. For example, can chrome make use of OS
>Services? Can you send and receive data with it from other applications?
>
>It seems like splitting hairs, "Well, it runs with a double-click, what more
>do you want", but there is really, rather a lot to being a "native"
>application.

I would argue that to be considered 'native', it needs to a) use
code compiled natively for the processor; b) use native platform
API's; c) use a native platform interface, including
platform-native widgets and not emulated ones.

So I don't think there's any question about Chromium or any
other WINE applications - they still use Windows APIs translated
for Mac, so they're not native. X11 apps might be arguable, if
you consider X to be a native API in OS X; I think that's pretty
questionable, but even if you do accept the APIs as native, the
interface certainly isn't. So no, X11 isn't native.

Firefox is, to me anyway, a key illustration of what it means to
be 'native' - or in this case, not really native. While it tries
to look and act like an OS X application, it doesn't use the
native platform widgets - scroll bars, buttons, dialogs, and on
and on. As a result, its look and behavior differs from
applications that *do* use native controls, in lots of little
subtle ways that drive me crazy. So I don't consider Firefox a
true native application, either.

Games translated with CIDER are an interesting case. (I
understand the Mac version of Spore is one such.) I don't know
enough of the technical details to know whether CIDER
re-implements the Windows APIs as native OS X APIs, or is more
of a translator like WINE. However, even most native games
create their own interface, and have little or no interaction
with the rest of the OS; in that case, can you truly call CIDER
ports non-native, if they behave just like native games?


Travis Butler
tbutlermac.com

Hi,

I have a question about Chrome and the Mac. As it is based on Webkit
what can it offer over Safari? I guess I can play with Webkit in
different ways than Safari, but if Webkit has a security hole, do both
browsers? Are Safari and Chrome basically wrappers over Webkit? I am
assuming Safari and Chrome install Webkit when installed on a pc? Does
this mean they can use different version of Webkit. Would this be
possible on the Mac?

I am interested on how this will work moving forward. Are we going to
have two developers competing over who can implement the latest Webkit
version fastest?

Best wishes

Michael

*****************************************************

This E-Mail has been scanned by Sophos Anti-Virus

*****************************************************

On Thu/9/18/08 04:21 37, "LewisGmail" <gkremegmail.com> saith :

> I found Crossover's Chromium locked up and stopped responding after a
> while of running. It also seems kinda flakey even when it is running.

I had a similar experience.
It was fast but quirky and then stopped after a while. I have turned it off

--
Max Manshel

On Sep 19, 2008, at 12:23 PM, Travis Butler <tbutlermac.com> wrote:
> Games translated with CIDER are an interesting case. (I
> understand the Mac version of Spore is one such.) I don't know
> enough of the technical details to know whether CIDER
> re-implements the Windows APIs as native OS X APIs, or is more
> of a translator like WINE. However, even most native games
> create their own interface, and have little or no interaction
> with the rest of the OS; in that case, can you truly call CIDER
> ports non-native, if they behave just like native games?

It's hard to say about Spore. I bought it for my 7 year old daughter
(really, it wasn't for me, honest). When you click on it's icon in
the Dock, a second Spore icon shows up in the Dock, just like X11 in
the early days of Leopard (since fixed). It takes a long time to
start. So it looks like it's starting some kind of emulation layer,
though I haven't checked the process table to see what's actually
going on. As you say, since it's a game it's hardly native anyway.

--
Paul Schinder
schinderpobox.com

On 19-Sep-2008, at 13:56, Paul Schinder wrote:
> On Sep 19, 2008, at 12:23 PM, Travis Butler <tbutlermac.com> wrote:
>> Games translated with CIDER are an interesting case. (I
>> understand the Mac version of Spore is one such.) I don't know
>> enough of the technical details to know whether CIDER
>> re-implements the Windows APIs as native OS X APIs, or is more
>> of a translator like WINE. However, even most native games
>> create their own interface, and have little or no interaction
>> with the rest of the OS; in that case, can you truly call CIDER
>> ports non-native, if they behave just like native games?
>
> It's hard to say about Spore. I bought it for my 7 year old daughter
> (really, it wasn't for me, honest). When you click on it's icon in
> the Dock, a second Spore icon shows up in the Dock, just like X11 in
> the early days of Leopard (since fixed). It takes a long time to
> start. So it looks like it's starting some kind of emulation layer,
> though I haven't checked the process table to see what's actually
> going on. As you say, since it's a game it's hardly native anyway.

CIDER reimplements the graphics calls into native APIs and does
everything else very much like Wine, as I understand it.