OSNews

I don't know the details of how the UA strings are parsed by the tool they used, but as far as Chromium goes if it is a "vanilla" build it will be indistinguishable from Chrome - the UA strings are constructed identically.

However, Chromium is often packaged by 3rd parties and they do sometimes modified the UA string. Ubuntu for example adds a chromium specific identifier "Chromium/xx.xxx", which does not actually exist in vanilla builds. It still contains the "Chrome/xx.xxx" identifier though, which is what most UA detection routines would pick up on and identify it as Chrome.

I would say in all likelihood that yes, Chromium usage would be bundled with the Chrome numbers.

On my windows partition, I only enable Flash in IE (mainly for viewing YouTube), and most of the time I use FireFox. On Linux I did not even bother to install Flash, as I have Chrome installed and use VP8 for YouTube playback.

On my windows partition, I only enable Flash in IE (mainly for viewing YouTube), and most of the time I use FireFox. On Linux I did not even bother to install Flash, as I have Chrome installed and use VP8 for YouTube playback. "

Chrome ships with Adobe Flash pre-installed as part of the browser install. Thus, you do have Flash installed. Just check the plugins page in Chrome, and you'll see it listed there. Even on Linux.

Yep. And while I agree with Thom insofar as Flash not going away anytime soon... generalizing OSNews reader trends regarding Flash from his dataset is error prone.
Some users may actually be "uninstalling" flash... but since Chrome integrates its own binary blob of the flash plug-in, Chrome's jump from 8% to 30% may have blurred those "uninstall" numbers. Others may have flash installed (since it's still obnoxiously necessary for functionality on certain sites), but they may have a permission-based blocker ("click to show element") sorta thing running. Both of these serve as confound to the reliability and validity of flash "trends"...

No, I made sure Flash is disabled in Chrome. I could not view Flash sites, and in fact some YouTube videos that are not VP8 encoded are not viewable in my Chrome installation.

Flash is the most annonying piece of software, and I always make sure that Flash is disabled in my default broswers. I don't want to use FlashBlock, and I don't want Flash, period.

IOW, you have Flash installed. It may be disabled, but it's still installed.

The last few weeks YouTube has been pushing out changes and A/B testing* changed to perfect their support for WebM/HTML5 video.

For example, if you enable html5-support for Youtube at: http://youtube.com/html5

And you visit a different site with a YouTube video enabled, it is very much possible it will be displayed in HTML5/WebM.

I've also seen that YouTube has started to enable HTTPS support.

That means that https://youtube.com/ could be the default in the future. This is to protect the login-cookie users use (used by people who upload and comment).

Only the video is downloaded over HTTP from the caching servers (it is a different domain, thus the cookie won't be sent in that case).

Also they are enabling HTTPS so they can enable the use of SPDY.

SPDY is the faster 'HTTP-transport'-method of loading pages from Google as implemented in Chrome and maybe Firefox 10, but probably Firefox 11.

* A/B testing is that certain users temporarily see the new version and other users see the original. This is to test what happends when you enable new code without impacting all users if something does not work.

SPDY is not good in the sense that with Google Services Chrome looks like it is faster. Chrome kinda uses it secretly, which means things like gmail etc run faster than they do in other browsers.

Yeah it is an "open" standard but only Google are using it.

SPDY will be adopted by Firefox and is actually a good thing.

It helps solve the Bufferbloat problem too: http://www.youtube.com/watch?v=qbIozKVz73g

http://www.osnews.com/permalink?502566

Edited 2012-01-09 10:17 UTC

There is one difference with browser wars I, they are all open specifications. Most of the time there is real discussion, it is all out in de open.

You have to remember this is the W3 and IETF process:
- someone has an idea
- discuss/make a draft specification
- 1 or more vendors implement something
- in CSS/JavaScript they will get a "vendor-prefix" so it does not conflict with the real standard if it works differently.
- there will be a lot of in the field experience
- a more "final" specification is made
- all or atleast most of the vendors implement the "final" specification
- the specification is announced as final
- the remaining vendor will probably also implement it

So the time between: have an idea and get something in one or 2 vendors browser is actually not all that long. Maybe half a year, but it can take years before all vendors adopt it.

There is nothing proprietary about it.

Have to admit, the Google developers do create a lot of new ideas/code to try it out in real life.

Something like SPDY needs a lot of operational experience before any other vendors would even think of deploying it. Deploying it on the Google websites is thus a really good idea.

Other vendors are free to come up with their own ideas and implement them. If it turns out they have a better idea, other vendors will adapt that.

From here,

http://www.pcmag.com/article2/0,2817,2397158,00.asp

I have seen websites that only work properly with webkit these days, and suddenly everyone thinks that is alright ... but everyone seems to be up in arms if a site only work properly in IE.

As for the vendor prefix ... Nothing annoys me more.

It is abused quite a bit by developers, go to Smashing Magazine and find one of their "top 20 sites doing coolest <something or other>" and check how many of those sites work in IE8 or even IE9, some won't even work correctly in Firefox.

Here is a good example from Zurb

http://www.zurb.com/playground/osx-dock

It won't work even in the latest Firefox, to be fair to Zurb ... this is only a bit of an experiment.

Yes, "one browser vendor widely deploys it early in few of the most popular web services around" should be on that list itself / it really doesn't seem to you similar to the fuss about MS & bundling of IE with Windows? (plus http://www.osnews.com/permalink?503344 )

Edited 2012-01-13 23:58 UTC

How is that not good?



Again, how is that not good?



Blame those not using it, not those that do.

Edited 2012-01-08 06:31 UTC

Because it makes it appear that Google Services work faster in Chrome.

TBH SPDY on its own isn't a problem, instant pages only works in Chrome.

SPDY is only one example of where they are doing it.

I honestly thought we got away from this "It works better in <browser x>", that is a relic of the 1990s. Now I have actually seen some pages that say "Works best in Chrome" ... brilliant ... Browser Wars II

Edited 2012-01-09 10:19 UTC

Sure instant pages only works in Chrome currently, AFAIK.

Interresting enough I do not see it being used in Google Search with Chrome currently (!)

There is nothing stopping other sites and browser vendors also implementing it.

As you can see it shows up in draft form on the W3C site before they announced it:

"2 June 2011"

"This is a First Public Working Draft of "Page Visibility"."

http://www.w3.org/TR/2011/WD-page-visibility-20110602/

"June 14, 2011"

http://blog.chromium.org/2011/06/prerendering-in-chrome.html
http://googleblog.blogspot.com/2011/06/knocking-down-barriers-to-kn...

There is a white paper which talks about it from a date I don't know:

http://code.google.com/chrome/whitepapers/prerender.html

Microsoft also wanted atleast the Page Visibility API, not sure what their angle was:

http://ie.microsoft.com/testdrive/Performance/PageVisibility/Defaul...

Firefox 10 (currently beta) also supports it, here is a test page:

http://davidwalsh.name/dw-content/page-visibility.php



Stupid people that build things which should work cross-browser but don't or make stupid claims are not something Google did. And people who work at browser vendors and on specifications moan about it on Twitter too so I don't think any of them encourage it.

A bigger problem is: having Apple create some CSS3/HTML5 demos on their site ( http://www.apple.com/html5/ ) and specifically blocking other browsers than Safari even though they support the specifications.

Obviously demos are not real sites and people have to be really clear about that.

It seems they have improved the site, now at specific parts of certain demos it says:

"Download Safari to view 3D transforms.

The 3D transforms in this demo are not yet supported by Firefox. If you’d like to experience these transitions, simply download Safari. It’s free for Mac and PC, and it only takes a few minutes."

or

"Enable GPU Accelerated Compositing in Chrome to view 3D transforms.
The 3D transforms in this demo are only supported by Chrome if you enable GPU Accelerated Compositing.

If you’d like to experience these transitions:

Open a new tab in Chrome with the URL “about:flags”
Click on Enable GPU Accelerated Compositing
Restart Chrome
Watch it again"

But it should never have happend.

The thing is a couple of years ago, we were in the position where we could build nice XHTML apps. Now because of HTML 5 and because everyone is chasing the new "Shiny Shiny", we are in another state of Churn yet again.

The only really positive thing to come out of this are the Geolocation APIs, but they seem to work differently on pretty much every browser (BTW the way IE9's impletation works is how I think it should work, Mozilla does a best guess based on IP).

Well it is somewhat reminiscent of Embrace, Extend...

I mean, come on, Opera is one of the most standards complaint browsers (also according to Chrome guys! http://blog.chromium.org/2010/03/does-your-browser-behave.html ), but it often works horrible / borderline unusable in js-heavy Google services; and now they're starting to neglect other browsers too, I hear.

Right now, for example, Gmail is basically broken - well, at least its js version ...and, again, that's despite Chromium guys claiming not a long time ago (when Gmail also worked poorly in Opera) that Opera is most standards-compliant in js!

Seems to me it's at least somewhat akin to bundling of IE in the case of Microsoft - strength in one area propping up another, and not strictly by virtue.

Edited 2012-01-14 00:09 UTC

And who should should we blame for the lack of ipv6 device ?

Who are to blame ?

If you ask the content providers: access providers

If you ask the access providers: hardware vendors

If you ask the hardware vendors: there is no demand

If you ask the users: most probably don't know what it is

I'm in the content provider business, I would say the hardware vendors are really slow to adopt it and I don't believe they didn't get requests from their customers about it as they kept claiming in the past that they did not get them.

The real problem of slow adoption by vendors is probably because the hardware (ASICS) needed to be created first which was probably an investment they wanted to delay for as long as possible.

I view YouTube and other video sites in HTML5 but still have Flash installed just in case. I use ClickToFlash so most of the time I'm not using Flash even though it is installed.

Having flash installed does hurt. Battery power on a notebook is adversely affected, and pages are slower to load.

That's what FlashBlock is for. I have flash installed because there's still too much content online that requires it, but with FlashBlock it loads only when I want it to load.

That's fine and all, but then you aren't sending a message to website owners that you don't have flash. Just look at OSNews as the perfect example, Thom thinks we all use flash. That isn't an accurate number, because of all you "ClickToFlash" and "Flashblock" people.

Better bet is to leave it uninstalled completely, for Safari or Firefox, then jump to Chrome when absolutely necessary. This is how I run, and I rarely find myself jumping to Chrome. Everywhere I go works without flash (except swtor web site... sigh).

Your comment motivated me to uninstall Flash and see how usable (or crippling) Firefox is without it. Should be interesting. I just looked through the 52 tabs I had open after uninstalling Flash and restarting Firefox and all seem reasonable.

My little experiment is meeting with mixed-success. Most techy sites have been reasonably usable. But apparently Taco Bell felt that having Flash on their front page was critical and told me I needed Flash ("This site requires that you have Adobe Flash installed on your computer. Click here to download and install Flash."). Thankfully I didn't really need their site to get the info I wanted.

Right, you can pick up a laxative at your local Walgreens for much cheaper than taco bell sells them for.

--
Sorry, I had to. Taco Bell is the funkiest of the funky fast food joints.

Edited 2012-01-07 05:26 UTC

I'm guessing installing the add-on NoScript is probably an answer you weren't aware of.

You can choose whether things like flash objects run or not, on a permanent or temporary basis.

Everyone? My computer has no Flash installed since about two years. Maybe I'm lucky, but all websites I consider worth visiting work. Embedded youtube videos don't show, but they work when watched on youtube directly, so I can live with that. Who knows, perhaps I even stumble upon a solution to that one day.

I use IE8 a lot, and only load Flash on selected white-listed websites (OSnews obviously not being one of them).

That's actually one of the neatest features in IE (second only to InPrivate Filtering): open the Tools menu > Manage Add-ons, select Shockwave Flash Object, click on More information (lower pane), then click on Remove from all sites.
Now every website that wants to run Flash will show a non-intrusive gold bar at the top, and let you choose whether you agree or not.

And then you realize that virtually every web site wants you to run it, even when there's no visible Flash content, and that its primary use is NOT as a multimedia framework, but as yet another user tracking system.

(Disabling local storage in Flash also somewhat limits its usefulness for trackers; but not running it at all if not needed it definitely the best option.)

Edited 2012-01-10 02:53 UTC

It would be very interresting to see the other 1% specified, especially on a site like OSNews.

+million

I was wondering about the mobile stuff as well (incl. Android). TBH, I generally just use gReader on my Android phone. My guess is that the feeds are being pulled from Google to my phone with Google's servers doing the actual hits to the feed. The only time I hit the site itself from my phone is if I wanted to check out the comments.

OSNews just sort of sucks on Android's Browser. I do it to look at the topics. I definitely would use it more if it just let me view the regular version of the Website. But that's kind of a pain in the but.

I don't mind the mobile OSNews too much. Only complaint is it's difficult to follow threaded conversations due to (wait for it) there are no threads! Understandably so, ever hit a site on a mobile browser that uses Disqus for the comment system? If not check out engadgets, it's pretty crap-tastic in my opinion. But then again, I don't particularly care for Disqus (or endgadget) in my full fledged desktop browser either.

Agree 100%. But, just because no one has made a decent mobile threaded discussion interface before, doesn't mean it shouldn't be tried.

The site looks great on my Galaxy Tab 10.1. It is not the browser that is the problem. You should know better than making such a generalized comment like that. What device are you using?

Sorry, I posted the same comment multiple places, in the other spot I specified I was using Cyanogen Android 7.1, which is pretty close to stock. It still gives it the mobile site, which is terrible.

Opera is still quite amazing... and its position in those stats is also from how OSNews seems to mostly cater to, draws audience from so called "western" cultural sphere.

But if it were largely a CIS audience site... http://gs.statcounter.com/#browser-RU-monthly-200812-201201 (check out also at least Belarus and Ukraine; similar in most of the CIS; or in Georgia; few central European states also noticeably more than is typical)

Edited 2012-01-14 00:07 UTC

The post-PC era is 99% hype, 1% reality. Well, not quite. But anyone thinking Windows is dead is simply a delusional fool.

Microsoft lost the tablets and smartphones -at least for now-. It is still not too late and Windows 8 and Windows Phone 8 can be their comeback, but if they miss the boat the second time it is over.

Well, you might be an order of magnitude off already... http://gs.statcounter.com/#mobile_vs_desktop-ww-monthly-200812-2012...
(plus, those stats are most likely sort of skewed towards the desktop, in how browsing on desktops tends to be more "intensive" and/or how many mobile users are very cautious about data costs - and how many people use mobile, for many of them this being the only access, is also important )

I had no idea! Honoured . And thanks!

Android, iOS, WP7, etc. get the full OSNews. Only non-capable browsers get the true mobile version of OSNews, but even there there's an opt-out link at the bottom to load the full page instead.

You keep saying that, but my Android phone running Cyanogenmod which is pretty close to stock Android has never gotten the full site by default. Furthermore, its a pain in the but to hit the opt out link every time ( its supposed to stay full for a month, but never more than a browser session).

Android devices are notoriously hard to detect correctly. Some include the "mobile" identifier and some don't (although it is getting much better as most new ones identify as "Mobile Safari" compatible). I expect what is happening is your browser's UA string does include the mobile flag, but for some reason is not being parsed as an Android device (and thus gets the mobile site by default). I would check to see what your User Agent string actually is:

http://whatsmyuseragent.com/

It has "Android" as its user agent. It shouldn't be that diffucult to detect if you want to.

Of course, Thom, David and Co, don't get me wrong I get that you're all short on time and money and I appreciate the fact that you are able to provide the service as it is. But, it really isn't detecting Android browsers correctly. If you have the time, energy and resources to fix it great! If not, no worries. I'm more than happy to use it as is.

Just don't take the stats too seriously to make an inference as to the popularity of mobile devices.

Using WURFL or similar server side libraries would prevent this.

I went to try it and it loaded the mobile site. Then I realized that the bookmark I'd used was set to mobile.osnews.com back from when I'd started using this phone.

Once I loaded osnews.com by typing it in, it did load the full site. I wonder when that changed.

Thanks Thom.

I browse OSNews using Opera Mobile on Symbian^3 and it operates just fine. The site initially loaded in it's mobile form, however opting out was a pretty obvious and simple option.

No advertising budget? What about all those free ubuntu CDs they gave out for years?

I agree that apple has done more advertising than the Linux community, but there are several commercial players in the Linux space. Saying they have no budget is silly.

I'd also argue that apple has spent little to prop up Mac OS X or Mac computers in recent years. Most of their ads are focused on iPhone, iPod, iPad and iTunes.

I really wish someone would come out full force for linux on the desktop. I'm a BSD guy, but even I can see the benefits of getting people off windows.

Well, considering that you are forced by law to buy a Mac if you want to develop stuff for iOS...

Utter bullshit.

Web apps can be built with any platform.

iOS-native ones can be built on any machine running a legal copy of MacOS and Xcode (just don't resell it yet).

Web apps ? Let's be serious for a second there. There's only so much things which a bit of Javascript which is not backed by a robust internet connection and a powerful distant server can do, and it's nowhere near what native iOS software (or Java 2 ME software for that matter) is capable of.

As for Hackintoshes, which you seem to suggest as an alternative to Macs for iOS development, I believe they do not count as a legal copy of Mac OS as far as Mac OS X's EULA counts. Don't know in which countries exactly this agreement is legally binding though.

It is quite a lot, I have written HTML apps, using JS and HTML and nothing else that are quite useful.

If the app has to connect to the outside you need network anyway ... so I don't understand the problem?

Last I checked, you're forced to use Windows to develop for windows phone. This isn't an apple specific problem.

I'm sure google would do the same thing if they sound desktop computers.

Well, they don't even offer instructions for building Android on Windows anymore, so their rule appears to be "use a Unix-like OS with gcc if you want to build Android."

No, it's not Apple-specific. I just challenged the idea that OS X does not have that much of Apple's marketing behind it anymore.

If Apple do heavy OS X/Mac marketing, they may get a few more OS X users. If they do heavy iOS/iStuff marketing, they get A LOT more iOS users since that market is not mature yet. A significant share of these users will want to develop native applications on their device, and need a Mac to do so, so OS X's market share will also go up.

While I despise the way Apple manage their ecosystem, I have to admit that it makes commercial sense.

Edited 2012-01-07 17:00 UTC

It is the .NET framework, what do you expect?

It is not like you can't get hold of Windows or VS Express licenses? Both end up being "free" as much as it matters.

Edited 2012-01-07 19:07 UTC

It's much less of a problem, when sort of everybody has a Windows machine; hardly anybody has an OSX one. Vista, the "flop" release, has notably more users than OSX.

http://gs.statcounter.com/#os-ww-monthly-200807-201201
And those few percentages not only possibly overstate things (the way OSX users are probably more web-active on average, particularly on sites less likely to be overlooked by Statcounter), they are also a reflection of maybe dozen, tops, atypical countries.
(check out the graphs for S. America, Africa, Asia - you know, the continents where MOST PEOPLE, by now probably most computer users, live)

Trying to portray it as the same kind of issue is hilariously inaccurate.

And the closest Google has to a desktop OS, they mostly give away for free...

Two/four percent are almost nothing, and may also just be random noise. I'd say, when the changes in usage exceed 10% it gets interesting.

I don't mean to be rude but this post displays such strange views about statistics that I'm finding it hard to figure out if you're being sarcastic.

The basic fact is that without access to the raw traffic data (as well as broader industry data by which some context can be gained), nothing of any substance can be said about operating system trends.

Linux usage expanded by almost 50% in 2011 alone, using statistics just as meaningless as Thom's.

http://www.netmarketshare.com/operating-system-market-share.aspx?qp...

Ahem.

I'd also point out that the statistical change in operating systems at a single website could simply be indicative of a change in the site's viewership. Say, for example, if OSNews ran fewer Windows stories and more Linux stories, there would naturally be fewer Windows readers and more Linux readers.

So it doesn't necessarily mean that people are switching their operating systems, only that different people are reading the site from last year.

Which is the point I was trying to put across.

--

Discover Amiga, Discover AmigaOS
www.amigaos.net

I don't know... http://gs.statcounter.com/#os-BR-monthly-201110-201112-bar

Yes, such web stats tend to underestimate Linux installations for various reasons ("in general, not only for geeks" usage: because, say, they're more likely to be company or school PCs used less for browsing, and if - then probably more on ~local sites), and overestimate OSX ones (owned by the most affluent & web active users, more on ~international sites), but still ...the exact situation in Brazil, how you describe, seems unlikely?