|
Plorkyeran posted:I use Nightly and have like 20 addons installed, and FF crashes something like once a week for me. When that happens, I have to wait all of two minutes for it to restart. Stability is a complete non-issue for me, so why would I be happy about sacrificing functionality for it? I get maybe one crash every two weeks but I constantly open too many youtube tabs and drive up the memory usage to the point where I need to restart Firefox. I don't think it's a memory leak or anything, even though I do close the tabs, because the same amount of tabs spread out over a week won't cause issues. Firefox seems to be pretty dumb about realizing that it's running into the memory limits for a 32 bit process and isn't aggressive enough with its garbage collection. Would be fixed if they got off their rear end and made their 64 bit version, which people have wanted for years and every other major browser has.
|
#
?
Aug 23, 2015 02:13
|
|
|
|
| # ? May 28, 2024 11:53 |
|
|
Desuwa posted:I get maybe one crash every two weeks but I constantly open too many youtube tabs and drive up the memory usage to the point where I need to restart Firefox. I don't think it's a memory leak or anything, even though I do close the tabs, because the same amount of tabs spread out over a week won't cause issues. Firefox seems to be pretty dumb about realizing that it's running into the memory limits for a 32 bit process and isn't aggressive enough with its garbage collection. They're targeting it for the next release.
|
|
#
?
Aug 23, 2015 03:17
|
|
|
Not holding my breath with how much its been pushed back, delayed, or de-prioritized.
|
|
#
?
Aug 23, 2015 03:50
|
|
|
So does that mean that the comment in the OP about all 64bit versions having lovely JS performance is now out of date and should be removed?
|
|
#
?
Aug 23, 2015 03:51
|
|
|
Lum posted:So does that mean that the comment in the OP about all 64bit versions having lovely JS performance is now out of date and should be removed?
|
|
#
?
Aug 23, 2015 04:27
|
|
|
Palemoon lurches on but I really don't know why anyone would actually use it these days unless they really want to be limited to about 2 dozen barely maintained forks of ancient extensions. And last I checked awful Javascript performance because a fairly major upgrade to that offended them somehow.
dis astranagant fucked around with this message at 04:52 on Aug 23, 2015 |
|
#
?
Aug 23, 2015 04:45
|
|
|
tete009 is still maintained and up to date. Been using it for a while and it feels a bit less janky, but hard to quantify. Might try waterfox 40 as a stopgap until x64 Fx41 comes out.
|
|
#
?
Aug 23, 2015 05:09
|
|
|
Desuwa posted:Sometimes it's just not liking change for the sake of change 
|
|
#
?
Aug 23, 2015 12:40
|
|
|
dis astranagant posted:Palemoon lurches on but I really don't know why anyone would actually use it these days unless they really want to be limited to about 2 dozen barely maintained forks of ancient extensions. And last I checked awful Javascript performance because a fairly major upgrade to that offended them somehow.  The main issue with any third-party builds is that they significantly lag the official ones when security updates are deployed, and that's a pretty big deal.
|
|
#
?
Aug 23, 2015 13:26
|
|
|
Alereon posted:Ah I see, I saw the 25.6 version number and assumed it was abandoned, great plan rolling their own version numbers to be unique snowflakes Palemoon's worse than that. Starting sometime around the launch of Australis they decided to start cherry picking every patch that came through and more or less turned into separate development on a now 2 year old ESR release. I liked them back when they were just Windows builds targeting modern hardware instead of the then standard early 90s processor targets. By the time I realized what they'd become that was irrelevant and they were throwing out 90% of patches as "untested", rendering them super slow and probably full of old exploits they didn't like the fixes for.
|
|
#
?
Aug 23, 2015 13:46
|
|
|
dis astranagant posted:Palemoon's worse than that. Starting sometime around the launch of Australis they decided to start cherry picking every patch that came through and more or less turned into separate development on a now 2 year old ESR release. It's basically a fork at this point. I'll let others be the judge of whether it's a lovely fork or not. They even have palemoon exclusive addons now, and berate AMO for just claiming that this addon is incompatible, even if there's an archived version in there that works. Because it's AMO's job to support someone else's lovely fork.
|
|
#
?
Aug 23, 2015 15:33
|
|
|
Lum posted:I'll let others be the judge of whether it's a lovely fork or not. [...] to support someone else's lovely fork. I hardly even knew Palemoon and Waterfox kept trucking along. What's their big selling point anyway nowadays? "It's firefox as you remember it 3 years ago!"? Because that's not really selling anything in my mind.
|
|
#
?
Aug 23, 2015 19:03
|
|
|
Marinmo posted:Think this one's in the bag Johnson. You sound like you've never even seen Waterfox. Waterfox is all about having a 64 bit build without running nightlies.
|
|
#
?
Aug 23, 2015 19:05
|
|
|
Nintendo Kid posted:You sound like you've never even seen Waterfox. Waterfox is all about having a 64 bit build without running nightlies. Waterfox is basically stable Firefox compiled as 64bit with icc, and a different default search engine that claims to be ethical.
|
|
#
?
Aug 23, 2015 21:54
|
|
|
The Merkinman posted:Change I don't like = change for the sake of change I'd count curved tabs as change for the sake of change. Well maybe change for the sake of matching Chrome, which is worse. No usability improvement with that.
|
|
#
?
Aug 23, 2015 22:20
|
|
|
The actual reason for curved tabs was to have a common design element that could be used across all platforms (Windows, Linux, Mac, and Android). While it's similar to Chrome's tab shape in the sense that both are roughly trapezoidal, they're still different enough that I can tell just by looking at the tab shapes what browser someone is using. Since browsers are moving towards showing as little chrome as possible, tab shapes are one of the few areas where you can easily show your "brand". You could argue that this isn't a "usability" improvement, but that's not to say that Australis in general wasn't an improvement. One of the primary goals of Australis was to simplify the tab code and improve performance (especially when animating the tab bar). That was the main reason Australis took so long to ship: the new implementation had to actually be better than the old one. In any case, I'm not sure I understand why the tab shape matters so much to people. I don't spend a whole lot of time looking at the tabs themselves. (Granted, I remember being really pissy when they changed the icons for Firefox 2.0, but then I realized I had no actual reason to care aside from the brief cognitive load of acclimating myself to the new icons.) I do think Mozilla could stand to schedule its UI changes better, though. I'd rather see significant changes all at once 1-2 times a year than a trickle of changes every 6 weeks.
|
|
#
?
Aug 24, 2015 00:30
|
|
|
The new tab shape takes up more space on the bar, meaning that when you have a bunch of tabs open there's less text on them, or less tabs for the same amount of text. It also looks pretty bad when put on the bottom where tabs belong.
|
|
#
?
Aug 24, 2015 00:44
|
|
|
Ew, a tabs-on-bottom person. What's next, top-replying? *throws glassware at wall* Something-something Fitts's Law on screen boundaries Avenging Dentist fucked around with this message at 02:11 on Aug 24, 2015 |
|
#
?
Aug 24, 2015 00:49
|
|
|
Alereon posted:Yep, I've changed that section to be a note that official 64-bit builds are going to be available with Firefox 41. All the third-party builds are dead except Waterfox, and there isn't really much reason to point people to software without security patches. Cyberfox still sees plenty of updates. Current looks to be 40.0.2, so it's not cutting-edge, but it updates pretty frequently.
|
|
#
?
Aug 24, 2015 01:02
|
|
|
Avenging Dentist posted:Something-something Fitts's Law on screen boundaries If you're going to bring up Fitt's Law then explain the stupid hamburger. There's a straight usability regression. The old menu button had infinite effective width and height where the hamburger has neither. I don't use the menu button much, usually less than once a day, but I gave the hamburger menu a try to squeeze one more tab's width out of my tab bar and I just kept opening my first pinned tab.
|
|
#
?
Aug 24, 2015 02:01
|
|
|
Avenging Dentist posted:Ew, a tabs-on-bottom person Agreed. Tabs belong on the side, especially in these days of widescreen monitors and websites targeting mobile.
|
|
#
?
Aug 24, 2015 02:04
|
|
|
Desuwa posted:I don't use the menu button much, usually less than once a day... I think you've answered your own question here: UI that's used less often doesn't need to be as easily-accessible. While it'd be nice to have something in the top-left corner, I don't think the hamburger is high-enough priority to go there.
|
|
#
?
Aug 24, 2015 02:15
|
|
|
Fangs404 posted:Those of you that you CTR - what exactly about the current look of FF don't you like? I really like the improvements and simplifications over the last 5-10 versions they've made to the UI. My biggest gripe is not being able to move the hamburger button, and having a star button permanently attached to the default bookmarks button. That's literally all I use CTR for.
|
|
#
?
Aug 24, 2015 02:41
|
|
|
Avenging Dentist posted:I think you've answered your own question here: UI that's used less often doesn't need to be as easily-accessible. While it'd be nice to have something in the top-left corner, I don't think the hamburger is high-enough priority to go there. At the same time the menu button had been there for years and no browser developers cater to people who fill up their tab bar in the first place. The change was to be more like Chrome, down to the iconography, and it showed. I'll actually give the menu button on the toolbar another shot but it won't be a hamburger. Same for removing the new tab button and seeing if I can't force myself to use the keyboard shortcuts more. e: Also with australis they could have rewritten the underlying code and still had it look like Firefox, instead they chose to make it look like Chrome. Desuwa fucked around with this message at 03:49 on Aug 24, 2015 |
|
#
?
Aug 24, 2015 03:03
|
|
|
Avenging Dentist posted:Ew, a tabs-on-bottom person.
|
|
#
?
Aug 24, 2015 03:58
|
|
|
FRINGE posted:Tabs belong just above the window the represent. Below the URL and above the webpage.
|
|
#
?
Aug 24, 2015 04:15
|
|
|
The Merkinman posted:Why above the window they represent but not the URL they represent? Yeah, the default behavior (tabs above the URL) makes more sense UI-speaking because the URL changes based on the tab you're on. Tabbed interfaces inherently represent a parent-child tree-like relationship, and the URL is a child of the tab, just like the content.
|
|
#
?
Aug 24, 2015 04:46
|
|
|
If Im madly clicking back and forth working between pages the page and the tab are all I care about. The URL bar is a keyboard-based device, the mouse interface (tabs) should be near the interactive area (webpage). edit also the URL bar creates a visual break between the current tab highlighting and the actual page.
|
|
#
?
Aug 24, 2015 04:51
|
|
|
The metaphor I'm most comfortable with has the titlebar, the menus, the URL bar and search bar, and then the bookmark bar. These are elements of the UI that do not change. Tabs are variable by design, moving about, being opened and closed, and occasionally even being torn off to spawn new windows if I sneeze at the wrong time. I just don't read them as some skeumorph for file folder or binder tabs. Having been trained to ignore the very top of the screen by decades of Windows UI standards doesn't help matters much either. When I use Thunderbird and a tab gets opened by accident, quite often I won't even notice at first.
|
|
#
?
Aug 24, 2015 05:12
|
|
|
FRINGE posted:If Im madly clicking back and forth working between pages the page and the tab are all I care about. That's exactly why putting tabs at the top is better. Fitts's Law models the speed at which you can navigate a pointer to a UI element (among other things), and putting an element at a screen boundary effectively gives it infinite size, since you can't overshoot. This even extends to UI elements near a screen edge, since you can easily "bounce" back from the edge of the screen to get to it.
|
|
#
?
Aug 24, 2015 06:17
|
|
|
Ooh. Okay, I can see that. That's clever thinking.
|
|
#
?
Aug 24, 2015 06:20
|
|
|
I find "Fitt's Law" highly questionable in this case. You still need to carefully navigate to a specific horizontal location no matter what, and "flinging" the mouse vertically while still carefully moving to a horizontal location is actually more awkward than just pointing to a specific part of the screen (which is what you're doing in virtually every other situation on a computer, barring something like active hotspots in corners). Breaking the standard model of the titlebar in Windows is disruptive, and (for me, at least) outweighs any "Fitt's Law" benefits. The most common UI elements I click on are tabs and bookmarks, and I want less used elements farther up, ending with a standardized title bar that works the same as it does in every other application, full stop. I have major problems with any application that modifies the title bar, or otherwise has nonstandard UI that fiddles with where the click-and-draggable regions are. Chrome and the way it does tabs is neat in a sense, and I use Chrome as a secondary browser for windowed video (easier than hiding the bookmark bars and other stuff in Firefox), but it's not useful as a primary browser, for me, considering the number of tabs I use, the way I use them, and the way I want my application windows to look when taking up the full vertical height of the screen (as in, having a standardized title bar). I also really dislike the conceptual placement of tabs on top above so many other UI elements. I'll grant that conceptually having the URL bar below the tab would make sense, but every other element, like the back/forward, reload, etc. buttons, along with the bookmarks, should be parents to the tabs, which make sense as immediate parents to the window content. Besides, the URL bar is essentially nothing more than a "click here to google" bar, and in terms of space saving it makes more sense to put it alongside the other rarely used buttons (like back, etc), so ultimately that's the conceptual break I'm comfortable making, rather than figuring out a way to get it under the tabs. I mean if you're really going for conceptual purity, stick it in the status bar at the bottom, as a display-only element just telling you the url you're on (when not hovering over a link), condense everything else (back, pretty much, and then the bookmarks) in a single line at the top, and have the status bar turn into a combination search/url thing on new tab. Say something along the lines of "that's also where search within a page goes" and you're done. Yay! I do want to stress that I'm really talking about what works for me. I understand the general explanations, and I'm really sympathetic to arguments about saving vertical space, I just don't have as much of a premium on vertical space on the screens and resolutions I work at. It's irritating when people use the "Fitt's Law" explanation to say that it's objectively better. It's also really irritating that they're using a specific weird interpretation/subsection of Fitt's Law, the idea that edges are of infinite value. That's not how mouses actually work. Corners are of infinite value, but edges clip off vertical/horizontal movement without arresting the other direction, causing you to slide along them when "flung towards." It's a false idea, unless the regions are particularly wide along the edge it's no better than having a UI element of a certain size near the edge and still having to click it precisely. zachol fucked around with this message at 07:17 on Aug 24, 2015 |
|
#
?
Aug 24, 2015 07:13
|
|
|
zachol posted:I find "Fitt's Law" highly questionable in this case. You still need to carefully navigate to a specific horizontal location no matter what, and "flinging" the mouse vertically while still carefully moving to a horizontal location is actually more awkward than just pointing to a specific part of the screen (which is what you're doing in virtually every other situation on a computer, barring something like active hotspots in corners). That's not the case at all for me. I find that Fitts's Law accurately explains the behavior I use to select tabs in Firefox. Maybe it's different on a touchpad or something, but I don't really like touchpads to begin with, so I try to avoid them. zachol posted:Breaking the standard model of the titlebar in Windows is disruptive, and (for me, at least) outweighs any "Fitt's Law" benefits. I agree for Windows XP, but even Internet Explorer messes with window layout. Windows 7's UI is all over the place.
|
|
#
?
Aug 24, 2015 07:26
|
|
|
Avenging Dentist posted:I agree for Windows XP, but even Internet Explorer messes with window layout. Windows 7's UI is all over the place. yeah well they're dumb and i hate them
|
|
#
?
Aug 24, 2015 07:33
|
|
|
zachol posted:I do want to stress that I'm really talking about what works for me. I understand the general explanations, and I'm really sympathetic to arguments about saving vertical space, I just don't have as much of a premium on vertical space on the screens and resolutions I work at. It's irritating when people use the "Fitt's Law" explanation to say that it's objectively better. It's not at all wrong to say buttons on the edge of the screen have infinite width/height. UX studies have consistently found that elements at the edge of the screen can be accessed faster, even if the users don't consciously notice it. Here's one study I found quickly, and if it wasn't such a pain to dig through PDFs I imagine I could find others (summary on page 10 specifically mentions elements on the edges): http://insitu.lri.fr/~chapuis/publications/RR1480.pdf It's faster to hit a one pixel line on the edge of a screen than it is to hit a considerably larger rectangular target of the same height that's one pixel from the edge, whether or not it's in the corner. The corners are the fastest targets, though, and the reason is obvious; with a corner you can overshoot in both dimensions where with an edge you still have to be on target in one. That's still an improvement over having to be accurate in both dimensions to hit something not touching an edge. You can choose aesthetics over efficiency but you're probably wrong about it not being faster for yourself. I'm not going to dig around for a study that looks at people who are outliers in regards to accessing elements on the edge of a screen but I suspect there are vanishingly few.
|
|
#
?
Aug 24, 2015 07:57
|
|
|
Staring at my screen for a while and flicking my mouse around, I'm realizing I didn't think this out properly. I'm sorry about that. However, I think the Fitt's Law stuff about screen edges doesn't actually apply anyway, since with Chrome and Firefox, the tabs aren't actually at the edge of the screen, they're a bit away from it. In fact, not even the title bar is strictly at the edge, when I flick my mouse to the top I hit the resizing zone. Turning on safe mode to disable addons and looking at what I assume is the "clean" version of Firefox, you've still got a region at the top for the dragging region, which sort of isn't a title bar anymore, and the very top is still a resizing zone. If you're flinging to the top, you're going to have to backtrack a little, requiring that careful control and losing the edge benefit. It does work in fullscreen mode, which is neat I guess. Acually I like this fullscreen popup behavior a lot, I can see the Fitt's Law thing at work, and I suppose it makes sense to preserve that appearance in the non-fullscreen UI, and makes sense to have that as the default UI for new installations. However, when dealing with windows, not fullscreens, the way that resizing borders works in Windows makes the application of Fitt's Law to screen edges difficult. You'd have to break that standard to replace it with something else.
|
|
#
?
Aug 24, 2015 09:08
|
|
|
Avenging Dentist posted:That's exactly why putting tabs at the top is better. Fitts's Law models the speed at which you can navigate a pointer to a UI element (among other things), and putting an element at a screen boundary effectively gives it infinite size, since you can't overshoot. This even extends to UI elements near a screen edge, since you can easily "bounce" back from the edge of the screen to get to it. Fitt's Law doesn't account for a decade or more of FPS gaming teaching you to maintain precise mouse control at all times. 
|
|
#
?
Aug 24, 2015 09:45
|
|
|
Desuwa posted:It's not at all wrong to say buttons on the edge of the screen have infinite width/height. UX studies have consistently found that elements at the edge of the screen can be accessed faster, even if the users don't consciously notice it. Here's one study I found quickly, and if it wasn't such a pain to dig through PDFs I imagine I could find others (summary on page 10 specifically mentions elements on the edges): http://insitu.lri.fr/~chapuis/publications/RR1480.pdf From my personal anecdotal observations, people who use their desktop browsers in fullscreen mode are much rarer then competitive fps-players with perfect mouse controll.
|
|
#
?
Aug 24, 2015 09:58
|
|
|
tonberrytoby posted:From my personal anecdotal observations, people who use their desktop browsers in fullscreen mode are much rarer then competitive fps-players with perfect mouse controll. Fullscreen (F11) is probably only for weirdos and people doing presentations/kiosks, but I'm pretty sure that the majority of Windows users browse with a maximized browser window and almost no MacOS users do so. I don't have up-to-date statistics, though. MacOS always encouraged non-maximized windows with it's top menu bar optimized for Fitt's Law though.
|
|
#
?
Aug 24, 2015 10:11
|
|
|
|
| # ? May 28, 2024 11:53 |
|
|
isndl posted:Fitt's Law doesn't account for a decade or more of FPS gaming teaching you to maintain precise mouse control at all times. Having precise mouse control makes you faster at clicking arbitrary points than other people clicking arbitrary points. It will still be faster to click a large button than a small one for any given person. Back in college I had to make a program to test this and run it on a bunch of people. FPS players still show the same overall trend, even if overall speed is greater. It's irrelevant to the question of where tabs go, of courses, since that's a matter of conceptual hierarchy, and most folks don't run maximized. Edit: I thought about it a bit more, and it actually wouldn't surprise me if most people with smaller screens or ones running below, say, 1080p actually run maximized. It does mean there's no good reason the menu button should be in a corner. I'm just not sure there's any element in a web browser more deserving of that location. Is there any reason to make the hamburger immovable, though? I could see making it impossible to remove so that you can't remove it, though that's a weak argument given that you can right click anywhere on the navigation bar and select customize; surely anyone who couldn't figure that out wouldn't customize in the first place. I literally can't think of a single reason to not let users move it at all. Blue Footed Booby fucked around with this message at 13:26 on Aug 24, 2015 |
|
#
?
Aug 24, 2015 13:24
|
|
