"Always after a defeat and a respite, the Shadow takes another shape and grows again." -- Gandalf
Chrome works on every platform. It hogs memory but its fast. Chromium & v8 are open. This is the kind of things that gave us Electron, nodejs and everything that is built on top of it. I appreciate Google for working on it.
Regarding privacy, I totally agree. Would be nice if the most popular web browser wasn't developed by the world's biggest ad company. I still think they do a descent job of isolating the two orgs. I can technically install extensions that stop much of the ad invasions.
An what Chrome is giving out is "Tier1" search results only fed to their browser (Thus search results are more accurate using Chrome.... and now Google Earth is "Chrome Only" which uses WebGL and the latest tests show Firefox is still over 3x as fast with WebGL content then Chrome.
Electron is HORRIBLE, enabling people to write apps in Javascript while using all your PC"s resources is insane. "Etcher" a program written in electron who's sole purpose is simply an "ISO USB Writer" comes with a payload of 180mb's on disk and over 200MB's of RAM and runs like an old dog with cancer... along with the other electron apps. This could have been written with something we had for decades with little overhead and small payload, it's called Python...
Except anything MIPS-based. Or Power. Or in fact anything that isn't x86 or ARM.
And it's not just a matter of compiling it for those platforms. There's a bunch of architecture-specific porting that would have to be done (e.g. you _have_ to implement a V8 backend; there is no platform-independent way to run V8 just with a C++ compiler).
Not sure about Chromium support on those platforms, but V8 support is pretty impressive.
So you're saying that Chrome is but a Sauron to IE's Morgoth? ;-)
I've run into several other complex sites that fail on Firefox. It's sad, because I've used it for years. I'm using it right now. But my default just switched to Chrome because I started having too many Firefox issues.
Once I became aware of this, I've been trying to be more diligent about thoroughly testing things on Firefox.
I wonder how many other developers are in a similar situation, where Chrome is their default browser and/or their main debugging environment. Part of the problem for me is that I find the Chrome dev tools superior, and that makes it so much easier to just forget about the rest (not that I'm justifying my behavior, btw).
I still use Firefox as default, both for developing and for general web browsing. It and my set of extensions fit my preferences too nicely and have no equivalent in Chrome. I use Chrome at work primarily for Google Hangouts / Meet, the occasional debug session, or just to have another session. (Trying to get into Chrome's Profiles feature too.) At home I just use Chromium from time to time, mostly because my computer is starting to age and I notice the performance difference for certain things.
(Also, I think a lot of people discount how good Edge's Dev Tools have gotten. There too my corporate mandated environment is mostly stuck with Windows 7 and an intentionally broken IE 11 due to Oracle and using their terrible software internally.)
Upon googling, I discovered that drop-down menus have been an issue in chrome (even not using vbox). I'm using zurb foundation for the menu js/css, fwiw.
This hasn't happened to me in a while, except for occasionally government or bank sites which require "IE or Chrome". In many cases, spoofing the user agent works just fine for those. I agree it's bad, though.
The one exception I've noticed is Yubikey (U2F) support for Google services. Firefox has an add-on that provides Yubikey support, but last I checked, Google blocked access to those (I believe even if you spoof the user-agent).
spoofing the user agent works fine
By all means, politely warn a user that "your browser is not tested". It's getting to feel like a marketing driven decision, where pages just about say "our site is so powerful, we only work with the greatest browser ever, Chrome, so come back when you have it".
A local airbnb competitor currently does this.
[0] https://philipwalton.com/articles/loading-polyfills-only-whe...
God I miss the old Opera....
I'm fairly conscientious about this myself since I'm working on plotting data, and the dumb client-side number crunching involved is actually pretty good at eating CPU cycles.
Most plotting libraries want to show off how smooth and incredible their animations are. What I really want to know however is: does your library keep updating the canvas at 60FPS, or does it only refresh when the data does and idle otherwise?
I think some of it is also that the web has numerous things about it that are fundamentally expensive operations, going back to things like "the default table sizing algorithm reacts to the flow of the content within it, which also depends on how the table decides to format it". It's not hard to create a pure HTML page that has no interesting images or scripts or anything, but still is fundamentally slow to render. (You probably wouldn't want to write it by hand, but I've accidentally written programs that output such pages over the years.)
I know there are all sorts of flags etc, but consider old Opera, where switching the status bar between none, simple and advanced was right there in one of the main menus. That was a good start, that stuff would be compact and super useful by now if we'd just keep going.
Sure, it doesn't explain to average users why the webpages they view might be slow, but average users don't care.
The average smoker probably doesn't want to hear smoking is unhealthy. Does that mean doctors should adjust their advice respectively? At what point does "professional" really mean nothing other than "gets money for it, like a carpenter or a thief or a drug dealer might"?
We don't even have the right to "just give people what they want without any judgement on our part", but we certainly don't have the right to ignore those with legitimate concerns because ignorant or apathetic people are more in numbers. That goes for everything, everywhere. That goes to how you are supposed to look out for little siblings when parents are away, and it goes for expert knowledge or intellect.
But what is any user supposed to do with a blinkenlight telling them what they already mostly know: that a website is bloated/slow/eating their machine slowly? If there was an alternative website, maybe they'd already be using it. If they thought complaining to the site's owners about it, maybe they already had or are already aware that they'd be shouting into a careless corporate void. That mostly just leaves uselessly blaming their browser for a blinkenlight that tells them something they already know and can't care about.
"Microsoft has lost over 300 million browser users in 2016, mostly to Chrome, tracking site shows"
Not going to take very long at this rate. Appears to be accelerating. Personally just use Chrome as for me extremely stable and when tried Edge it was not stable. I do a lot of surfing and often times have a lot of tabs open and can not remember the last time a tab crashed.
As for privacy you could use the open source chromium, there's a fork somewhere which has all the Google removed.
Chrome using less resources than Firefox and IE was their sales pitch.
Now they've become what they made fun of, while Firefox is smooth as a baby's butt.
Chrome is still a huge improvement compared to the browsers it was competing against. Chrome changed the market and now the other browsers are competing in the world Chrome created. So while Chrome might fall behind in some areas now, it's naive to say that it's become what they made fun of.
It's nice there working on it, but why didn't they ever care before?
I've always assumed people talking about having hundreds of tabs open just don't understand how to properly use a browser. My grandmother, for example, usually has 100 or so open by the time I get a call about her having computer problems.
There's no logical reason I can come up with for doing this instead of using bookmarks.
Then throughout the course of the day I end up looking up API docs, get linked to blog posts, news articles, and YouTube videos, and read articles which themselves have relevant links to follow. Most of these I just open in a background tab to check out later in the day. These accumulate until I have time to go through and quickly review them. Those that I want to read and don't have the time currently go to Pocket. The rest get read or summarily closed out. I find bookmarks to be a terrible way to triage tabs.
This workflow works for me (and evidently others). It's faster than bookmarking. It's less prone to failure, in my experience (I've suffered bookmark corruption more than once). And a modern computer ought to handle many background tabs just fine. Moreover, if browsers aren't expected to be used in this fashion, they really should set an upper-limit on the number of tabs that can be opened.
Hopefully this gives you some perspective on alternative use cases. It sounds like your workflow works out well for you. I've tried it and couldn't get it to stick. If that means I don't know how to use a browser, so be it. At this point, there's enough of us (your grandmother included) that maybe the browser vendors should just find a way to cope with it better.
Each browser instance is tabbed completely across, I keep them open until I read the page fully, and then save it in keep to keep forever.
By Friday I can have hundreds of tabs that I go through and clean up. Web apps are a huge pain to constantly log in.
I run Korora with 24GB RAM and an I 7, Chrome is never a system hog for me, and most of the time it surprises me how well it handles my use.
I have hundreds of tabs open at a time. Instead of searching for something, then going to the Google page, clicking and waiting for it to load, in Firefox I search for what I want in the bar, it presents the tab as a result and opens it instantaneously.
In addition with the vertical tab bar extensions I can see a list of about 40-50 tabs open at a time, using the additional horizontal space monitors provide that web pages don't use to keep an easily visible list of tabs.
I open pages that interest me, I might read them later like I did this discussion or just drop them. Add in open tickets, reference pages, the build server, youtube, etc. and the number grows over time.
> My grandmother, for example, usually has 100 or so open by the time I get a call about her having computer problems.
Maybe she should use Firefox instead of Chrome?
> There's no logical reason I can come up with for doing this instead of using bookmarks.
I did this when I started, by now I only use bookmarks for high interest pages, no point in bookmarking everything.
So has cVim, but that has no relation to any of this.
Watch out for Opera's add block, it works through a proxy to compress data, thus they know your every move online too...
You spend too much time working lurking around HN sites and young web developers.
That's not true. Microsoft has now gotten into the spying business, and is infamous for the Windows 10 telemetry. They're basically copying Google.
Firefox and Safari are the only ones that come from companies that aren't notorious for wanting to know everything about you. And Firefox doesn't try to get you to spend scads of money on massively overpriced but mediocre hardware that locks you into their ecosystem.
Firefox has its warts, but it's the only choice that really makes sense if you care about privacy and freedom and avoiding vendor lock-in.
Sure, error reporting feeding in to a QA database is one thing. But is there the capability to target Win10 OS ads to, say, folks with old video cards? I'd be very surprised if someone in Redmond didn't think of that.
(I highly doubt that there is any truth to this claim)
> One of the most persistent misunderstandings about Safe Browsing is the idea that the browser needs to send all visited URLs to Google in order to verify whether or not they are safe.
> While this was an option in version 1 of the Safe Browsing protocol (as disclosed in their privacy policy at the time), support for this "enhanced mode" was removed in Firefox 3 and the version 1 server was decommissioned in late 2011 in favor of version 2 of the Safe Browsing API which doesn't offer this type of real-time lookup.
Firefox 3 - That was 9 years ago.
https://feeding.cloud.geek.nz/posts/how-safe-browsing-works-...
May be.
I see this as truly the case of making the pie bigger.
All devices include 2 Billion Android devices. That's a bigger pie that Chrome owns now, but FireFox can target in future.
Are you suggesting that Chrome gathers data about you? Because unless you tick that box (which they show pretty prominently) it doesn't. It doesn't by default in most linux distribution packages.
> Unlike IE
I don't know if you've been following, but Microsoft is now the king of knowing everything about you. They record things about their customers' computer activity which should horrify anyone. Sometimes they don't respect user selections either, even all the way up to Enterprise editions (where it is often mission-critical not to send competitive information to Microsoft by accident in a core dump), which is infuriating.
> In my opinion chrome isn't doing a good job either. It's a massive energy hog and waste more CPU than it needs to.
My impression is that at this point, most unique performance problems in Chrome are either an inherent cost of multi-process, a mediocre implementation choice in that model, or a performance tradeoff toward better application latency at the cost of heavy initialization. Chrome could display many pages more quickly if they ignored the GPU, but they use it across the board so that they don't have to restart into "GPU mode" when they realize there is a lot of compositing on the page. Chrome has converged toward other browsers recently, they'll now run multiple tabs on the same process as long as they share a FQDN (sites that host together, crash together), I suspect they do this to save memory.
If we're talking about runtime speed of real web apps and sites, Chrome has everyone matched or beat.
> The time to celebrate victory was a few years ago. Now it's starting to look like the new boss is the same as the old, maybe worse.
The problem with IE6 was not Microsoft, or IE6 itself. Microsoft did not win by literally forcing people to use IE6. The problem was, and probably always will be, greedy unscrupulous web developers (and their managers) who want all the cool new toys at any cost. Microsoft was doing the cool, "html5, bro!" type browser innovation that google is doing now, and developers (and their managers) lapped it up. People forget that Microsoft made box-model: border-box, XMLHttpRequest, favicons, <ruby>, and bi-directional text on the web. They did this all in IE5, this put IE6 ahead, and people loved it too. Microsoft did the right thing and didn't break compat for honest customers who just wanted their webpage to work, so IE5 quirks are the way of the web.
The problem is not that nobody likes the boss, the problem is that everyone likes the boss.
No, I didn't forget. But you forgot to say that MS did all that with draft specifications or even no spec at all (XMLHttpRequest), just to beat everyone to market, then refused to correct their implementation once the standard was revised and agreed by others. And they sprinkled ActiveX on top, for good measure.
> developers (and their managers) lapped it up
Disagree. Developers were the ones that pushed Mozilla and then Firefox (and then Chrome) as soon as they could.
> Microsoft did the right thing and didn't break compat
Microsoft did the right thing for their own bank account: they smashed the competition with bundling then left IE to flounder, even obliterating their dedicated team, because they had reached their objective, which was to dominate the web so that they could sell what they really cared about: ActiveX and other Windows-only technologies.
> the problem is that everyone likes the boss.
No, the problem is that people are lazy. IE won with OEM bundling on Windows. Chrome is winning with OEM bundling on Android. As long as the default is good enough, people won't switch, especially 15 years ago when downloads took a degree of effort (waiting several minutes, restarting after failure etc etc) and now on mobile where it is awkward and/or completely forbidden to switch browser. This is basically what the article says as well: they couldn't push a browser, they had to push an OS with a browser bundled. If people don't switch, developers can't build for alternative browsers, because their managers won't allow the additional time and effort.
No, the DOM put into IE was legitimately better than that in the old NN. MS won that war because their browser was BETTER, period.
They then sat on their laurels and the rest of the world passed them by, so now they're still trying to play catchup.
But at the time? No, anyone who had any experience in the DOM of IE vs NN would hands down push IE. It was just better in every way.
Where did I mention Netscape? I didn't.
IE5 is from 1999, IE6 from 2001, and they were undoubtedly better than Navigator; but the first 0.x releases of Mozilla with the new Gecko engine are from late 2000/early 2001, and were better than IE (although the suite was slow and bloated). Firefox was branched out in 2002 and took off very quickly because it was a great engine without the bloat of full Mozilla. That's why people pushed it (or rather Phoenix) right off the bat.
If you were pushing IE6 over Mozilla or Firefox in 2001/2002, you weren't paying attention. Navigator all but died in 1999.
When developers were pushing IE was when it was IE vs NN.
"Microsoft has lost over 300 million browser users in 2016, mostly to Chrome, tracking site shows"
No. Microsoft had a non-standard box model which was an utter pain in the ass, regardless of their box model being more sensible than everybody else's in theory[0] (and ultimately standardised as an option circa 2010) having to code for a single standard documented box-model is way the fuck easier than coding for two different box models.
Also it's box-sizing not box-model.
[0] because in practice MSIE's layout engine was a buggy pile of shit
The sane way of doing boxes pre-dated the "standard", which in hindsight appears as though it was specifically crafted to spite Microsoft.
It was a silly time to be a developer.
I definitely run into sites that only work on chrome and not in any other browser.
In my opinion chrome isn't doing a good job either. It's a massive energy hog and waste more CPU than it needs to. Unlike IE and FireFox and Safari it comes from a company that is notorious for wanting to know everything about you.
The time to celebrate victory was a few years ago. Now it's starting to look like the new boss is the same as the old, maybe worse.