[notfanboy]

The Z10 has a very good browser, but there's been a lot of hype and plain old misinformation about it in the past several months. You all remember those posts stating that it is "better than most desktop browsers" and other posts that raised the Z10 to this mythically perfect browser. Now that we're a couple of months in, it's about time we check out which claims are true and which aren't.

The Z10 browser was completely written in html5. FALSE. Looking back this sounds really hard to swallow, but some people really believed it back in the Dev Alpha days. The trouble was that RIM bragged about this a lot, but in fact the rendering part of the browser was in webkit, but the UI elements around it was in html5.

The Z10 browser runs each tab in a sandbox. If a webpage dies in a tab, the rest of the tabs are unaffected. FALSE. In the screenshot below, you can see how the "Bubbles" page brought down the entire browser.



The Z10 browser is sandboxed by BB10/QNX. If the browser crashes, the rest of the phone is unaffected. FALSE. You can try this out for yourself. Open a couple of active frames, the run the Aston Martin test from the link below. Shift gears upward until the browser crashes. At that point you will see that all the active frames have also been taken down. (By the way you can see how the Aston Martin test runs to completion on the S3 in this video:
)

Site Map

Z10 browser has the best Flash of all mobile browsers. TRUE. The Z10 was able to run all of the Flash pages in this video I made some months back:
and the Z10 felt more responsive somehow. Follow along on your Z10 and you might notice the same thing. Since Flash is officially not supported by Android anymore, the BB10 browser is indubitably the king of mobile flash.

Z10 browser has the highest HTML5 compliance of all released browsers. FALSE. This misconception was based on scores from html5test.com. Html5test doesn't actually test for html5 compliance, it only asks if the browser says it is compliant.

It is easy to check for yourself how the html5 score is misleading. The Z10 has a higher score than all desktop and mobile browsers right. Go to Microsoft's html5 testdrive site on your Z10 and on your desktop browser. Just in the first few tests, you will see the Z10 browser crash or misbehave. What if you compare it with other mobile browsers? I ran the same tests on the stock browser, Chrome, and Dolphin on the S3. The Z10 had the most crashes on the Microsoft testdrive site.

Z10 browser is the fastest mobile browser. SOMETIMES it is, and sometimes it isn't. It depends on the website. In general the Z10 is good with html5, but is poor at javascript. The Z10 also does a good job at quickly rendering a page, so pages usually show up a split second faster than on my S3, as in the video below.



Z10 browser does not lag and does not have checkerboarding. FALSE. For this one, please refer to the video above. It is a side by side test with the Z10 and the S3 on a mainstream website, namely CNN. Notice that the Z10 shows a still picture where there ought to be a video. With the video playing on the S3, the two browsers are made to scroll side by side. The lagginess of the Z10 browser is very evident.

[yvpan1]

so?? what's the verdict? is the Z10 browser GOOD? or is it RUBBISH?

honestly i have no idea what you wanna jot down here, sorry dude.. excellent explanation but no conclusion.

[notfanboy]

so?? what's the verdict? is the Z10 browser GOOD? or is it RUBBISH?

honestly i have no idea what you wanna jot down here, sorry dude.. excellent explanation but no conclusion.

The conclusion was right in the very first sentence. "The Z10 has a very good browser, but there's been a lot of hype and plain old misinformation about it in the past several months." I just wanted to test some of the wilder claims made in the past few months now that we actually can.

[donnation]

Thanks for the info, really well done and thought out post. Cheers!

[tek1255]

It seems very technical but I can tell you my browser is awesome. I've done simple test runs against other people holding there phones and I've beat them with loading Web pages. I have had my phone for almost 2 months and yet to have my browser crash as your test shows. Also, multiple active frames still the browser responds great. So, I think your reaching with all the technicalities.

Posted via CB10

[notfanboy]

tek1225: You are welcome to recreate all my tests. I've provided the links in the post and explained how to do the test.

Here are a few more observations.

Website compatibility is often a problem.There are times where a website appears to render properly but you're actually not getting all the content. Classic example is the CNN mobile website in the video below where what is supposed to be a video is just a static picture.



On TED.com, you also have a situation where the page appears to render properly:



This is how it is supposed to look:



Here is thenextweb.com



Google Docs is unusable on the Z10 (works in iOS and Android)



Now in fairness, this may not be BBRY's fault at all. The page simply hasn't been written to recognize the Z10 browser and coded accordingly. The best way to fix this is for BB10 users to send feedback to the website owners and report the problem. Because iOS and Android have a large marketshare, websites make sure they are optimized for these browsers. It is not BBRY's fault that the website authors have not gotten around to optimizing it for Blackberry. I find myself often going to the browser settings and toggling between desktop and mobile in order to get a satisfactory result.

[tek1255]

Also, I would like to do that test run you did with cnn. First off, I think it is stupid that you would load a video, not watch it then scroll through the site like a chicken with its head cut off. Sorry just isn't practical. I'm the browser does lag at times but who ever said he doesn't???

Posted via CB10

[STV0726]

So the question is I want the OP to order off which mobile browser he/she thinks is the best, fourth place obviously being worst.

Between BlackBerry 10, Android, iPhone, and Windows Phone.

Please and thank you. Oh and no "Depends". I'm looking for overall experience assessment.

~STV

[KermEd]

Some of this is inaccurate.

#justsayin

Posted via CB10

[STV0726]

Some of this is inaccurate.

#justsayin

Posted via CB10

I would be delighted if you could please provide counter evidence.

Would you kindly?

~STV

[notfanboy]

Also, I would like to do that test run you did with cnn. First off, I think it is stupid that you would load a video, not watch it then scroll through the site like a chicken with its head cut off. Sorry just isn't practical. I'm the browser does lag at times but who ever said he doesn't???

Posted via CB10

Here is the link to the CNN story. No noon meal for kids in debt at middle school - CNN.com

"First off, I think it is stupid that you would load a video, not watch it then scroll through the site like a chicken with its head cut off. Sorry just isn't practical."

It's a talking head video. I do this all the time - read the rest of the article or comments while listening to the audio. Besides, the Z10 didn't have the video playing, only the S3. I was actually trying to make the test harder for the S3

[diegonei]

There are a lot of points in here where we don't see eye to eye, and either we just don't have enough info or can spin it any way we want. I'll give some exemples.

Browser being full HTML5: The browser is a full HTML5 app. The UI is HTML5, the code behind how it reads the screen taps and all the other stuff is HTML5. The rendering engine is WebKit. The engine isn't the browser. It's just a rendering engine. Google is dropping WebKit but it isn't dropping Chrome. Chrome will just have a different engine. See what I mean? we can just spin this any way we want.

HTML Compliance: Well, the score on HTML5Test, let's see what it say:

The HTML5 test score is an indication of how well your browser supports the upcoming HTML5 standard and related specifications. Even though the specification isn't finalized yet, all major browser manufacturers are making sure their browser is ready for the future. Find out which parts of HTML5 are already supported by your browser today and compare the results with other browsers.

So, a high score means high compliance. It may not mean best implementation. It may not mean best compliance. But it means compliance. So yes, the BB10 browser has the highest compliance (and I wish it had the highest compliance to javascript as well, although it's leaks and bounds ahead of the PlayBook).

Sandboxed Browser: This one is tricky. Whay I think happened on your screenshot is: The WHOLE browser had an issue and crashed (not that mine ever did, but I admit that all browsers eventually crash). Now if the issue was just with that one page, I believe it would have worked as advertised. I may be wrong on this one though.

On the test that crashes the browser. Have you ever wondered that Active Frames may have some HTML5 content on it? If that's so, crashing the browser may as well force the whole HTML5 rendering engine to boot. Food for thought and something to get an expert to actually tell us about. I'm not that person.

FLASH. Ah-Ah.: And very soon it will be one of the few devices, if not the only one, to support Flash.

SPEED: Depends on the site, on the speed of the connection,signal strenght, your data plan, on the site's usage of javascript (and how properly coded that was)... Soo many variables.

LAG/RENDERING: I am yet to see what you showed on your video. Not saying it may not happen. Just saying not everyone is going through that. That said, also never had it checkerboard and scrolling is so much better than that.

So all that said, my point is, I agree, the browser is great but not perfect. No browser is really. No smartphone is either. Most of what we see here is debatable and prone to change depending on a lot of varibalbes.

[Canuck671]

I did recreate the tests. I monitored the memory during the tests. When the memory ran out the lag and the crashes occurred. That being said, some of the active tiles closed, some however, did not. I'm interested in testing my other phones in the same way. The result of any testing must meet the criteria of being repeatable, and have finite measurable data. Stay Tuned

Posted via CB10

[notfanboy]

There are a lot of points in here where we don't see eye to eye, and either we just don't have enough info or can spin it any way we want. I'll give some exemples.

HTML Compliance: Well, the score on HTML5Test, let's see what it say:

So, a high score means high compliance. It may not mean best implementation. It may not mean best compliance. But it means compliance. So yes, the BB10 browser has the highest compliance (and I wish it had the highest compliance to javascript as well, although it's leaks and bounds ahead of the PlayBook).

I'm not interested in spinning anything. I'm interested in repeatable experiments.

Here is what html5test.com says regarding HTML5 compliance ( The HTML5 test - How well does your browser support HTML5? )
"We cannot distinguish between a browser that supports a particular feature and a browser that lies about supporting that feature."
I encourage you to test this for yourself. Go to the Microsoft testdrive website that I linked to above. Run the first few tests on your Z10 and your desktop browser. The desktop browser, with a lower html5 compliance score, will run those test without crashing and with better results. This test, which you can do in five minutes, will show just how much we can trust the Z10 self-reported compliance.

LAG/RENDERING: I am yet to see what you showed on your video. Not saying it may not happen. Just saying not everyone is going through that. That said, also never had it checkerboard and scrolling is so much better than that.

Again, I encourage you to try out the same webpage whose link I provided above. Record a video like I did and post it here.

I'm not claiming that my tests are infallible, any corroborating or contradicting evidence is most welcome.

[KermEd]

I would be delighted if you could please provide counter evidence.

Would you kindly?

~STV

I'll try. It's going to be tough from my phone.

Posted via CB10

[crohns]

The web browser is great!!

[SDTRMG]

Lol it says if a Web page dies, not if the browser crashes. Are getting paid to write this none sense, because my iPhone 5 browser, MBP, browsers, ipad and nexus have all crashed to.


Qnx sandboxs the os, if the apps crash it doesn't effect the rest of the phone. And qnx shuts things Dow if there's a problem to keep the main backend of the os running. I'm not saying your wrong or anything but I've had to restart my nexus 7, MBP, and Windows 7 laptop due to browse crashes before, and when recreating your test I actually like that if it's a serious crash it doesn't reboot my phone.

it's not perfect just really good.


Posted via CB10

[KermEd]

I feel the need to preface this by saying generally I agree with you.

But there are some points where I think the accuracy could be improved.

For example, with a Microsoft test. For accuracy it should be tested on several html5 benchmark + compliancy testers by company's not affiliated with a mobile handset.

The Z10 browser was completely written in html5. FALSE.
...the rendering part of the browser was in webkit, but the UI elements around it was in html5...

I see a lot of back and forth on this one. We have to keep in mind that most people have no idea what HTML5 means. And that HTML5 in BBRY world is simply an extension of webkit.

Webkit is the driving engine for both - and both feed directly into it. So although technically you are right (it is not using one of the dozen or so HTML5 tags to draw all the content) it is a moot point as the underlying framework (webkit) is used to drive both. It becomes more of an argument around semantics as both terminate in the same framework.

The Z10 browser runs each tab in a sandbox. If a webpage dies in a tab, the rest of the tabs are unaffected. FALSE.

I'm surprised people said that. RIM has struggled with multi tabs on QNX on all versions. Part of this seems to be the device has a hard time with discarded tabs. When a tab is discarded, the OS interprets it as the Window needs to be discarded. Because the app is using tabs inside a Window it results in the entire browser being discarded. So if it misses catching and interrupting that event...

(figured that out writing Secure Browser)

The Z10 browser is sandboxed by BB10/QNX. If the browser crashes, the rest of the phone is unaffected. FALSE. ... run the Aston Martin test from the link below. Shift gears upward until the browser crashes....

Actually, here is where I need to disagree. Technically, this statement made by fans is true.

When your device has reached its maximum load on a resource, most noticble RAM, the OS reacts (not the browser). What you see is a symptom of the OS. The reason why BB10 rarely hangs (unless you have a bad device) is the OS will automatically close all non essential apps when low on resources to free them up.

So what you are seeing is a website asking the browser to overload it's resources too quickly for the browser to catch it. As a result, the OS sees the device is at risk of freezing, and closes all non-essential apps until it sees the system as stabilized.

So in effect, all this test proves is that the OS protects it's core systems during an overload.

The "browser crash" is a symptom of the OS closing everything down. Not the other way.

What you don't see in active frames, is run times are shutdown too including Android. That's how I built my android reset tool (Overload). So the next time you run an android app, it will reinit the runtime.

Interestingly this is also what made Shoppers Paradise customers so angry on PB. They were storing app data in Android temporary file space. So when you overload your RAM, android reinits, and you lost your game data.

Z10 browser has the highest HTML5 compliance of all released browsers. FALSE. ...This misconception was based on scores from html5test.com. Html5test doesn't actually test for html5 compliance...

See my note at the top. One test (Microsoft or html5test.com) does not to prove it either way (for or against your statement). Would love to see a more comprehensive test is all I'm saying here.

I agree with your statement about not relying on html5test. But I don't agree that it's been proven false yet without more testing.

Posted via CB10

[mikeo007]

Here's another test to see some rather poor implementations in the browser, in this case: canvas.

CanvasMark 2013 - HTML5 Canvas 2D Rendering and JavaScript Benchmark

Pay close attention to the visual artefacts, especially in tests 4 & 8.

[KermEd]

Also, I want to post and say thank you for the level of detail. It helps a lot in recreating and is such a nice change from:

Omg I hate my z10 so much you love your z10 compared to how much I hate mine the browser lags where's netfkix

Posts

Posted via CB10

[JMDBERRY]

All I know is that no other current smart or dumb phone's browser has beat my Z10's browser; hence,...is all I care about! Z10 wins!!

[zten]

Nice write up!!

Posted via CB10

[diegonei]

I'm not interested in spinning anything. I'm interested in repeatable experiments.

Here is what html5test.com says regarding HTML5 compliance ( The HTML5 test - How well does your browser support HTML5? )
"We cannot distinguish between a browser that supports a particular feature and a browser that lies about supporting that feature."
I encourage you to test this for yourself. Go to the Microsoft testdrive website that I linked to above. Run the first few tests on your Z10 and your desktop browser. The desktop browser, with a lower html5 compliance score, will run those test without crashing and with better results. This test, which you can do in five minutes, will show just how much we can trust the Z10 self-reported compliance.



Again, I encourage you to try out the same webpage whose link I provided above. Record a video like I did and post it here.

I'm not claiming that my tests are infallible, any corroborating or contradicting evidence is most welcome.

HTML5Test saying they can only check if the browser said if it supports something proves nothing. Neither does the MS site.

Already midnight and I'm gonna have a very busy week, but I can tell you YET another thing: The MS site you mentioned, already been there. Most of that is built to run better on MS devices (including Windows Desktop). Back on the early PB games, I visited that site and many of the things didn't work, then 2.1 came and most of then worked, then MS patched the site and most stuff stopped working.

Spin this way, spin that way.

Not gonna record videos.

But I like one thing you said: Repeatable tests. Get us a place we are not bound to anyone's bias, and let's have the CB comunity run tests. I like tests.

Here's another test to see some rather poor implementations in the browser, in this case: canvas.

CanvasMark 2013 - HTML5 Canvas 2D Rendering and JavaScript Benchmark

Pay close attention to the visual artefacts, especially in tests 4 & 8.

I think this will do. There is also http://peacekeeper.futuremark.com. Been there and I know the Z10 won't do particularly well numberwise.

But the main thing here is, as usually people forget: User experience. It may not do well in tests, but it usually does well in day-to-day use.

EDIT: Tests 5 and 8 were the bad ones here, where 5 was slow and 8 just didn't do anything (ran it on the PC first to actually know what to expect). 4 had a few boxy issues, but nothing that bad.



Chrome on my W7 scored 4853.

EDIT: Funny as it may be, the PB got 4363 (4415 2nd time) on this one... I'm adding devices as a means of comparison. After this, I ran the Z10 through a 2nd time and got 4115. Then 3455.

[mikeo007]

The MS site you mentioned, already been there. Most of that is built to run better on MS devices (including Windows Desktop).

So wouldn't it stand to reason that two devices, neither of which are Microsoft products, would be equal competition for a test that's not biased towards either of them?

[diegonei]

So wouldn't it stand to reason that two devices, neither of which are Microsoft products, would be equal competition for a test that's not biased towards either of them?

It would Mike, indeed. It would also stand that both are biased against. Llike I mentioned, I've been there on different occasions and got very different results, only explanation I could find was something withing the code changed. So I gotta stick with the "let's use sites that are not associated with phone/OS making companies".

Of which the site you linked is a good example.

EDIT: Here's another one we could use: http://browsermark.rightware.com/

EDIT: Here are my BrowserMark's results. 67% conformity to HTML5.



Chrome on my W7 scored 3693. Same 67% conformity.

PB got 2161, 65% HTML5 conformity.