What's new in Firefox 10 (shipping this week) | Hacker News Reader

What's new in Firefox 10 (shipping this week)

mozilla.org

109 pointsck274 comments

> Full Screen APIs allow you to build a web application that runs full screen

Finally YouTube HTML5 videos can play full screen and not just full browser window.

Am I the only one old enough to believe to remember that IE had this functionality (I refer to the "full screen on click on something in the page" functionality) at least more than 10 years ago and it sucked big time, as poor users would just click somewhere and then they would end on some malicious page which would "replace" their menus, taksbars etc with some nefarious code?

Why should it be different this time?

Because mozilla has given it some thought https://wiki.mozilla.org/Gecko:FullScreenAPI

How does this compare to Chrome's kiosk mode?

Also, Changes in Firefox 10 for Web Developers

https://developer.mozilla.org/en/Firefox_10_for_developers

Now that I've learned to think of their major version numbers instead as minor (ie. this could be 5.10) I have no problem with their "rapid release schedule".

I don't think that makes much sense (nor has the entire version number debate ever made sense). You'll get as much work as was completed in the 6 weeks window. This might be a bunch of smaller changes. This might be some important stuff. It's just less clumped together because each version covers a fixed, smaller time period.

A good comparison is Ubuntu. Is Ubuntu 12.04 a bigger increase than Ubuntu 11.10 was to 11.04? The major version number is bigger, after all.

     Is Ubuntu 12.04 a bigger increase than 
     Ubuntu 11.10 was to 11.04?

On 11.10 I've had and still have problems with wifi, video, audio in, audio out, webcam and freezes on boot. Also the last time I felt a laptop so sluggish was when I was trying to work with Windows XP on a 486 DX 66Mhz, even though my laptop is a dual-core T4500 @ 2.30GHz with 4 GB of RAM.

So considering how 11.10 completely fucked my laptop and nothing works properly anymore, spending countless of cycles going through every 11.10 bug report ever created and trying fixes in the hope that something works, I sure hope that 12.04 is an "increase", otherwise I swear to God that I'm switching to anything else I can get my hands on. And in the meantime I just have to live with "sudo reboot".

IMHO 11.10 should have been called "Ubuntu Millennium", as that would have made more sense.

Linux tends to break hardware drivers occasionally as things get updated and they don't have open source drivers or hardware to test on. Especially when the OS is built from testing releases. Best thing to hope for on Ubuntu (and Mint) is partition /home by itself and run a clean install when you need to upgrade distros.

Only upgrade to the latest version half way through a cycle. Ie, in January and July. That way, you can see all the problems that everyone else has had and make an informed decision on whether to upgrade. This month would be the time that I should be upgrading from 11.04 to 11.10. I'm not going to though because I've seen all the problems that other people have had with it. I'm hoping I can upgrade to 12.04 in July as I've heared they concentrating on reducing power usage, so hopefully it will make my laptop last longer. I could ugprade with everybody else in April, but then I'll be taking the same risks as everyone else. It's only a three month wait.

Ubuntu version numbers are strictly time-based. 12.04 is the April, 2012 release. 11.10 is the October, 2011 release.

As with FF, the version number gives you no indication of the scope of change between releases. FF releases are planned "at 6-16 week intervals" according to Wikipedia.

I'm not so much a fan of incrementing major with each rev, but it's just a number.

Firefox releases every 6 weeks. Sounds strictly time-based to me. (Not clear where Wikipedia gets the 16 weeks from - the article they give as the source doesn't mention it at all)

Thanks for the correction.

Firefox releases are strictly time-based. Schedule at https://wiki.mozilla.org/RapidRelease/Calendar

I disagree. I think this is a case where Mozilla did the right thing and Linus got it wrong. Linus should have called the version after 2.6.39 "40" instead of "3.0". If there's no semantic reason for two or more fields, just go with a single field. Humans instinctively realize that the difference between "40" and "41" is probably minor so there's no need to emphasize this through the use of a minor digit. Firefox doesn't have this advantage yet, but will soon.

There is a lot of infrastructure at this point that breaks if the Linux version number is not "x.y.z" and in fact, a lot of scripts had to be fixed in the bump from "2.6.39" to "3.0".

They sound like bugs to me.

Calling them bugs doesn't make them less in need of fixing or easier to fix.

Why do you even think about version numbers? Just always be on the current release. The numbers keep getting bigger so the average user stops and considers how far out of date he is. The average tech person should be happy with version # 3.1415926535.

So the average user considers how far out of date he is? Oh! I thought it was just a marketing thing. If this is making users upgrade then that's totally different and I am all for it.

Google Chrome originated in December 2008 and is now at version 16. At the current rate, Google Chrome version 37 will be released in 2016, and Google Chrome version 58 will be released around 2020.

If had actually taken the time to understand that Chrome is being released in an approximately 6 week interval you would know that Chrome 37 will be released far earlier (sometimes at the beginning of 2014) assuming that the schedule remains unchanged.

I believe the average TeX person is.

The forward button is now hidden until you navigate back.

The cost of inconsistency and confusion is not worth the 40px of extra address bar space.

I've been using the Firefox nightly channel for a while, and I had never even noticed this change til it was pointed out to me.

I think you're overestimating the opportunity for confusion, here... :)

-edit- I'm using OS X, perhaps the change is somehow more obvious on other systems?

> the opportunity for confusion

I like this phrase. It makes me aware of bad practices. Studying anti-design patterns.

I think that "Studing Anti-Design Patterns" is a great name for a band... =)

On OS X at least, this looks really good and I haven't found it a bother at all. Then again, I'm probably not the average computer user.

As it is only hidden when there is no forward-history, and the user wouldn't be looking for it at that time, I still believe it's a good change.

when there is no forward-history, the user wouldn't be looking for it

The only way I can tell whether or not I have any forward-history is by checking whether or not the forward button is disabled. If the button is altogether missing, I'm going to wonder for a split second if I did something wrong the last time I tried to customize my toolbar.

Well, I guess I'll get used to it after a few days.

Still, I wonder if there's some usability testing data comparing the lack of a button with a disabled button. A disabled button conveys different information from the lack of a button. Every web browser is trying to be minimalistic these days, but there's got to be a point where any further minimalization hurts usability rather than helping it.

They did some user studies and found that a fair number of folk never use the forward button at all.

Here's one study, not sure if it is the most recent: https://heatmap.mozillalabs.com/

40% of users never touched the forward button even once. And those who did use it, didn't use it very often.

Several of the very low% UI elements (like RSS or Go) were removed completely from the default setup.

40% of users never touched the forward button even once.

Another way to interpret that data would be "60% of users at least sometimes use the forward button."

The back-forward pair of arrows conveys a very intuitive idea for most people. In the absence of a "forward" button, I wonder how many FF10 users will accidentally click "refresh" thinking that it's the "forward" button. After all, the arrow in the "refresh" button points in the opposite direction of the arrow in the "back" button, and now the two buttons are side by side. It's not difficult to imagine them as a pair. Yeah, one arrow is straight and the other one is curly, but I often think of "forward" as "undo back", and many programs represent undo/redo with curly arrows.

The refresh button has moved since that diagram. It, along with the stop button, now lives on the right edge of the location bar, far away from the back button.

Ah, I'd forgotten that. That would invalidate my last point.

IIRC, I moved the refresh button right back to its original location as soon as the change happened, a couple of versions ago.

Sure, but the forward button hasn't been removed -- you can still use it, but only when there's actually a forward stack.

I think the ratio of back to forward button uses was something like 100:3 in that study.

(I rarely use either button -- I used to use keyboard shortcuts, and now use a trackpad swipe.)

I would just like to point out that your point and the parent's point are both valid and non exclusive.

As a example, one can navigate back and forth by gesture or shortcut keys, and never have to use the back and forth buttons. Still, the display of the buttons to conveys the presence or not of back and forth history can be useful.

You're absolutely right.

Fortunately, this is Firefox we're talking about. I'm sure there will be an about:config option to disable the new behavior if you don't like it. Just like you can change browser.urlbar.trimURLs in Firefox 7+ to restore "http" at the beginning of URLs.

Even less users use the huge orange FIREFOX button that uses up vertical real estate, but they didn't hide/remove/move that? I see no reason it should be where it is, as big as it is and as bright as it is. Put it where chrome puts it.

makes more sense to have it where it is. at least its "somewhat" complying to the rest of the ui chrome/IE put it in a relatively arbitrary location.

I don't understand your reasoning..?

Yes. This constant urge to remove any browser chrome possible is getting more and more irritating. It's basically a textbook UI anti-pattern that some fool thinks is a good idea and some other fools are copying.

I have a nice, large monitor that runs at a nice, high resolution. In fact, I have several. Not even my laptops are short of screen space these days. Some of this is high-end kit, but entry-level monitors are widescreen and run at decent resolutions these days too.

Space for an extra icon in a consistent position is nothing in this context. However, the irritation from having UI elements or page content move around under-cursor is significant, and the irritation from having to click multiple times or click-and-hold to access functions that used to be one-click is also great, particularly if you first have to guess which of the entirely meaningless icon thingies might hide the secret menu with the item you're looking for, or remember which browser you're using so you know whether the icon you want is on the left or the right of the other unlabelled random icon thingies and whichever text box bar thingies we're drawing this week.

I'm all for innovation, but we have platform UI standards for a reason. Breaking them because you think your software is more important than everyone else's isn't clever. Being the company that also makes the OS and wrote the platform UI standards is no excuse, either.

I think we have two groups of people to blame for this phenomenon:

(A) "Novices" who get confused when presented with more than a handful of buttons, because they never bother to figure out what those buttons do. BTW, what's a browser? You mean Google? No, I use Bing.

(B) "Experts" who don't care whether or not there are buttons in the UI, because they know where everything is hidden, and they use keyboard shortcuts anyway. Meanwhile, everything Apple does must be the One True Way.

When Group B makes software for Group A, you get the results you're describing. Caught in the middle are a large number of moderately proficient users who rely on explicit visual cues (buttons, arrows, menus) to perform tasks.

<.<

>.>

People actually use those buttons?

This is an odd thing to complain about. Sure it's a UI faux pas, but it's nothing compared the the creative placement of cancel buttons that's been introduced in 4+. It took me a full day to figure out why 90% of the files I downloaded were nowhere to be found.

I fixed it with this

https://addons.mozilla.org/en-US/firefox/addon/firefox-3-the...

and this

https://addons.mozilla.org/en-US/firefox/addon/backforward-d...

And now I am happily back in Firefox 3 / 4 days. Looks perfect.

In recent versions of Firefox, the back-forward dropdown menu can be activated with a right click over the back-forward buttons. No need for an add-on for that.

I think it's wonderful. I'd been doing this since the days of Firefox 2.x, using a custom userChrome.css file that hid the back and forward buttons, and combined the stop and reload buttons. But ever since I discovered Vimperator (I've now moved on to Pentadactyl), I've hidden all the chrome buttons, since they're entirely unnecessary.

What's the difference between Pentadactyl and Vimperator?

Pentadactyl -> Vimperator <=> XEmacs -> GNU Emacs. Its a political fork and PD maintainers claim its more actively maintained and developed.

In my experience, Pentadactyl is more actively developed. The people who started PD were the two most active developers of Vimperator before the fork. PD has since diverged enough from Vimperator that the code bases are almost entirely incompatible.

For what it's worth, Nielsen just released an alertbox today about "chrome" and the user experience.

http://www.useit.com/alertbox/ui-chrome.html

The general gist is that /too much/ chrome is bad, and a good way to fix it could be to temporarily hide parts of the chrome, as long as it's consistent in how to find it again and that there is a simple and reliable way to re-enable it.

It seems that Mozilla nailed the consistency (in this case, the forward button will always come back when the user goes back), but I'm wary of the "simple and reliable" method of re-enabling it. From what I understand, you technically can't make the forward button show up until you go back, correct? You can't always reliably go "back", especially if you're on the first page when opening a new tab. I'd say that Firefox is violating design guidelines (or whatever you'd want to call the research just put out by Nielsen today).

yeah, but they seem to love that type of thing.

Lets hide the http:// prefix... but we have to show it on https connections so really we have accomplished nothing.

Unresolved: This version of Firefox will not work on Macintosh hardware with Power PC CPUs

Seems the folks building TenFourFox are pulling it off...

http://tenfourfox.blogspot.com/2012/01/100-rc-finally-availa...

But with TenFourFox you can't have flash.

Or JavaScript JIT compilation.

Yes, by disabling various features until the remainder works.

Mozilla just feels that shipping an effectively crippled Firefox on a particular platform is not a good idea; for one thing web developers would scream bloody murder because of their browser-sniffing habits...

The forward button is now hidden until you navigate back.

I've been using the "Aurora" channel, so I've had this feature for a while. It does take some getting used to, since the whole address bar now jumps around when you push the Back button, and then jumps back when you click Forward enough times. On the plus side, the address bar usually has more room now.

Hmm, that sounds annoying. Couldn't they find a more elegant solution than having toolbar elements jump around all the time? I'm not sure if ~30px of extra horizontal space is worth all the jumping around.

It's pretty smooth, so it's not a huge deal.

Edit: Plus, it only shows up when you change pages, and everything (favicon, SSL, site info, etc) is going to jump around anyway.

Huh, I don't see that here on Aurora. It's just greyed out and the layout stays static.

Here's an example: http://seanpalmer.smugmug.com/photos/i-bpVFjBV/0/O/i-bpVFjBV...

When I click a link, it shows the URL without a Forward button. If I click the Back button, the Forward button appears and pushes everything over. When I click Forward, the Forward button is grayed out, but as soon as I move my cursor away, it disappears and again looks like the first.

Ah, thanks. I put my bookmark button on the left, and in-between the urlbar and buttons, which precludes this from happening.

That would explain it :) A similar effect happens with the Stop and Refresh buttons. If you put them next to each other, they become a single element. And if you put them both at the end of the address bar, they all merge together. But if you move them apart, they are separate buttons.

My observation is that using the small buttons disables this feature. And I use the small buttons anyway (I'm on nightly).

Some users may experience a crash when moving bookmarks (681795).

Seems pretty strange to proclaim a crash as something to be proud of in the 'whats new' list.

Probably they mean this bug went away (thus the fixed tag) but then the text should say 'Moving bookmarks no longer crashes'.

Weirdly, there's actually a bug filed for that: https://bugzilla.mozilla.org/show_bug.cgi?id=701653

I understood that this is also the release that makes all add-ons compatible by default.

It does. I think the release notes are slightly out-of-date and are currently being regenerated. Check back tomorrow!

Some day, the new <bdi> tag is going to save me a huge headache. Hear this prophecy, and know it to be true: weird hairy Unicode problems always happen, though usually not immediately.

Excited for the CSS Transforms!

FF 10 Release Candidate is here:

http://ftp.mozilla.org/pub/mozilla.org/firefox/nightly/10.0-...

Been using it 24+ hours now with zero problems (including a ton of extensions and custom theme).

Or just wait for the release late tomorrow or Wednesday I guess.

not related to this release per se, but could anyone give a 'helicopter view' of why Mozilla seems to be releasing new versions so quickly?

It seems like a year ago and I was on v3.6, and now it's version 10. I'm kind of getting upgrade fatigue and often struggle to see any major differences, except occasional performance changes (for better and worse). Just hoping somebody has an insight into the strategy and why the numbers are churning so quickly.

sorry, replying to myself - I didn't realise that this had a name (Rapid Release Schedule) and that there's already plenty of googlable information. Ignore...

The release notes format is quite "pretty" but very simplified and I can't seem to find a more detailed version?

some of it

https://wiki.mozilla.org/Platform/Features/Firefox10Bugs

In the past they have provided a full list of bugs fixed. My guess is that they will tomorrow when it's released.

This should be pretty close though: https://bugzilla.mozilla.org/buglist.cgi?op_sys=All&quer...

So, essentially...nothing? Major version, my shiny metal butt.

Reply on news.ycombinator.com