Hacker News new | past | comments | ask | show | jobs | submit login
Hands-on with Vivaldi, the new Web browser for power users (arstechnica.com)
134 points by Heewg9 on March 6, 2015 | hide | past | favorite | 105 comments



And just about the first thing you see when trying to install it…

Vivaldi may collect visitor statistics. The visitor statistics may include information about the visitors IP-addresses, usage patterns, <…>

I guess I don't want your shiny new browser.


At least it's one of the first things you see!


But for "advanced" users this is a turnoff. I wonder what they are getting paid per-user, and if that amount is greater than what they'd make if they just charged for the app.


Every website collects that stuff.


Most websites don't collect that stuff across every other website.


Most don't. A surprising amount do - any page that embeds a Facebook 'like' button loaded from Facebook servers with a referrer header ... or JQuery hosted by Google or a Doubleclick advert or a reTweet button, and on and on.


Strictly: that's not "most websites", it's "many services".

Facebook "likes", Google analytics, New Relic monitoring, and any of numerous other tools can do this.

Which is why I'm loaded for bear with noscript, adblock, Privacy Badger, Ghostery, and numerous entries in my /etc/hosts file for particularly noxious / ubiquitous sites.


This is why I stay logged out of facebook and don't allow third party cookies. Trying to move off of gmail also so I am not always logged in there either. Probably not perfect but hopefully it foils a lot of attempts at tracking.


Install AdBlock, Disconnect, uninstall Flash and Java, and use a JavaScript blocker and you'll make a good dent in a website's ability to track you.


And make a good dent in your ability to browse websites.


Are you speaking from experience or out of thin air? Because that is not my experience.


I block all scripts globally, flash heavy sites do not load and I consider that a 'feature'. Frankly, surfing the web without NoScript & AdBlock is an experience I do not want and have learned to do without. IF I need drivers or need access or ecommerce I'll allow the main site to run temporarily in Sandboxie or use tools/alt approaches to get the data without 7 or 8 ad servers running. The false promise of the internet is easily eschewed when you take a pragmatic approach to it.

Also: I was looking for something on mainstream retailer sites and was amazed at the list of servers. How many TargetImg servers does it take to load a product page? Answer: At least four, plus all the other tracking/ad loading detritus. Just wow.


You re-enable JS on the sites that you visit often and I've never had an issue.


Oh, well, missed that. That's what I get for replying to an out of context snippet.


It's perfectly reasonable to expect extra data to be collected when using prerelease versions of software. Complaining about it now is premature.


That's not a tradeoff I'm comfortable with, "even" in pre-release software.

I'm not accusing you of this, but in other cases I've seen proponents of something say "This isn't a big deal, it's just starting out". Then, when people continue to complain a year from now, the defenders say "This isn't a big deal, it's always done this."

Vivaldi looks interesting, but I'd wager that there's a decent overlap between the Power Users they're after, and the group that values their privacy the most.


Privacy has nothing to do with it. If the software is not yet ready, users are expected to encounter serious bugs. Your QA department is paid to help provide you with information to diagnose and troubleshoot those bugs, but users aren't, so you need to collect that automatically.


If they explicitly say they will stop in the release version, I would accept that argument. If not...


> It's perfectly reasonable to expect extra data to be collected when using prerelease versions of software.

This is incorrect in the world I live in, and I do not want to visit a world where it is correct.


You're in the wrong world then, buddy. If the software is not yet ready, users are expected to encounter serious bugs. Your QA department is paid to help provide you with information to diagnose and troubleshoot those bugs, but users aren't, so you need to collect that automatically.


This is the worst-written article I've read in a long time. It repeats itself (forgivable), contradicts itself (unforgivable) and I'm pretty sure is factually inaccurate in places (Doesn't Firefox already show the slowest addons?)

In spite of that Vivaldi looks interesting, and as a long-time fan of Opera (the original company and the browser) I wish them well.


>Doesn't Firefox already show the slowest addons?

I don't know about Firefox, but Internet Explorer definitely warns about addons that slow down launch times and opening new windows/tabs.


The claims that there aren't new browsers is wrong as well - there's many of them, it's just that they don't have a marketing team behind them. Uzbl, luakit, surf, jumanji, midori, vimprobable, xombrero, etc.


Well, none of those are new. Breach.cc is relatively new but even that goes back a year. But these highly extensible vim-like microbrowsers you list (excl. midori) have been around for years.


FF38+ gives you a popup if it detects an addon is slowing page render: http://www.ghacks.net/2015/02/18/firefox-38-notifies-you-abo...


Note that the current stable is 36.


There are two things I really miss about Opera. Generic Content-blocking - that is not only blocking ads but any URL-pattern of interest. And also the per-site settings. For example a lot of sites use extensive javascript only for tracking, but not for functionality visible to the user, so blocking javascript on those sites truly improves loadtime on many sites. And also being able to fine tune cookie settings per site.

It's such simple functionality, but really makes a huge difference both for performance and privacy and it saddens me that Opera never will be what it once was, and probably no other browser either. :-(


Yeah, I want stuff like the F12 menu, and a good workflow for uses who whitelist sites for cookies, javascript, and flash.

Now, Firefox might be able to fill the void, but I doubt Chrome ever can. Those kinds of features are an irritant to Google, because they want their pervasive analytics, ads, and various web-apps to work out of the box.


What I want from a browser:

1. Speed 2. Privacy 3. Add-ons/features (and this is far below 1 and 2).

This browser collects all sorts of information (seemingly more than chrome based on the message from instillation), and consumes an absurd amount of memory per tab, even on plain text webpages.

There may be something promising there, but this isn't it.


This browser consumes about same amount as chrome in my tests. Maybe 1/5 more.

Firefox and Chrome also collect the same information, but it is an opt-in from what I remember.

Edit:

Chrome usage with 3 tabs: http://i.imgur.com/mGiX0So.png

Vivaldi usage with the same tabs: http://i.imgur.com/ejnNzib.png


Installed Chrome extensions are your remaining helpers in there.


> Firefox and Chrome also collect the same information

Uhm, reference? I seriously doubt that...


Well, first from https://vivaldi.com/privacy/ it's not clear to me whether the privacy thing applies to the browser or the vivaldi web site.

Second, here's google chrome's privacy policy:

https://www.google.com/chrome/browser/privacy/

You can see that they have all kinds of information being sent to google. At least as much as Vivaldi.


Well, that probably makes you a minority even among "power users," however we might define that term.


> And since Chrome arrived in 2008, the Web hasn't seen another major browser launch—until now.

Lots of new browsers have launched since 2008, and Vivaldi certainly isn't a major browser yet, if it ever will be.


I normally respect the things you say on here, but for once, I do have to cry "Citation". If you would be so kind, can you list some of the lots of new browsers that have been released?


The original quote was

> And since Chrome arrived in 2008, the Web hasn't seen another major browser launch—until now.

The contested bit is "until now". Either you have to drop the major from the classification for until now to make sense, in which case there have been tons of minor browsers released since then, or you have to classify this new browser as major, which is a stretch.


Not an unfair questions. Another post already listed some:

https://news.ycombinator.com/item?id=9159709

Sure, these aren't major browsers -- neither is Vivaldi, yet, and its purely speculation that it might be in the future.


Furthermore, there is little reason to suspect Vivaldi might become major. It brings nothing substantially new on the tech side - it's using Blink, like Chrome and Opera, and like Opera, it only differentiates itself from Chrome on the UI side.

Many browsers have differentiated only on the UI side; none have really succeeded. As another example, Flock.

The 4 actually major browsers have all brought major new innovations: Firefox was a different rendering engine than IE (and much better in pretty much every aspect, except legacy website support); Safari launched with WebKit (a new major fork of KHTML with many improvements); Chrome in turn launched using a major fork of WebKit (bringing process isolation and a completely new JS engine, v8).

Vivaldi brings nothing like that to the table. It's interesting to see new UI ideas, of course. But for the article to call it "the first new major browser in many years" makes me immediately wonder if the author doesn't own Vivaldi stock (which of course is not the case; but the point is that it seems unjustifiably positive on the product).


Firefox was an innovation compared to IE, but it was explicitly a trimmed-down version of Mozilla, which had existed in the form of Netscape going back what amount to geologic eras in terms of the history of the Web.

Firefox's innovation wasn't the rendering engine, per se, but everything surrounding it, including add-ons.


As a loyal user of Opera 5-12 (and a grudging user of the new Opera, which I find marginally faster/less full of surprises than Chrome), I love that the article reads like a list of old Opera 12 features that I miss.

At the same time, I'm deeply worried that they believe they have to rebuild everything, including the mail client, to get to their first official release. Opera had a long pattern of rough but innovative releases that took a while to stabilize, and I can't help but believe that with them emphasizing feature completeness first, it's going to be rough for quite some time.

Given the high bar set by modern browsers I'm not sure users--power users especially--will put up with the associated quirks long enough for Vivaldi to gain traction.


Still it proves that Opera 15 could have been done differently while still remaining Chromium-based.


Where's the source? What's the license? I couldn't find this information, and downloads seem to be binary-only.


From some quick searching, apparently it's closed-source, but built on top of Chromium's rendering engine (Blink).


Wow… why would anyone ever want to use this browser? Honestly. If it were open source, it would at least have a chance.


To be fair to Vivaldi, Chrome isn't open source either, and lots of people use it.

Chrome is of course mostly open source, except for some proprietary bits. The same is basically true of Vivaldi (and Opera), they all add a relatively small amount of proprietary code on top of the massive open source Chromium codebase.


AFAIK the only differences between Chromium and Chrome are:

* Preinstalled proprietary plugins (Widevine, Flash) * More preinstalled codecs (woohoo, patents!) * Various user metrics reporting

There is a pretty big difference between that and keeping all of your changes private.


Well, for Vivaldi and Opera, "all of your changes" in this case amounts to just the UI. Neither has the capability nor motivation to modify Blink in any significant way.

Also, Opera has been contributing to Blink, so they aren't keeping all their changes private.

Vivaldi hasn't launched yet, but I would guess the same would be true of them - it just isn't rational to do otherwise (if Vivaldi fixes some small bug in Blink, why keep it as a patch they need to constantly reapply on top of Blink? Far easier to upstream it).

So aside from the UI (which is tiny compared to the entire browser), Chrome, Blink and Vivaldi are about equal in terms of being open or closed.


But everything beyond Blink/Webkit is precisely what I'm interested in viewing, contributing to, and wanting the FLOSS community to be able to audit and alter, etc. etc.

I was a pretty die-hard Opera fan for many years, so this news really piqued my interest... up until I learned the software is not libre.


And the updater (for Mac/Win).


As someone who tried using it: There's nothing about this for power users. Even with an absolutely minimal website, it consumes ~20mb per tab.


I imagine "power users" probably have more than 2GB of RAM.


Not on my good old T60, I don't. That users have a lot of RAM is a terrible excuse for bad memory management, anyway.


But also much more programs competing for it - I have 6GB on my work laptop, and I'm constantly at 95%+ (not counting cache!).


This needs to be said - upgrade your RAM. If your device doesn't support a memory upgrade, get one that does.

Upgrading my laptop to 32GB RAM was one of those things where you wish you had done it a lot earlier. No more OOM-killers, no more things grinding to a halt because the box is swapping (in fact, I disabled swap altogether), no more checking memory utilisation before starting a beefy VM.

Just smooth sailing all along.


Yeah, until the next release of Chrome and Firefox, where GMail and Facebook together will start to happily eat 16GB of your 32. The amount of memory browsers now eat up is insane.


It's worth noting that there is already an open source browser project to recreate + extend the Opera 12 UI: Otter Browser

http://otter-browser.org/


Let me know when Vivaldi implements vimperator style link selection and I'll try it. Otherwise if you have to click Firefox+vimperator or palemoon+pentadactyl will always be the true powerusers browser


>Firefox+vimperator or palemoon+pentadactyl will always be the true powerusers browser

Only if you are already a hardcore vim user and want that functionality in your browser, tons of people don't care


I was going to say exactly this. A fast browser with pentadactyl, but still the ability to use heavy SPAs like Asana without too much of a hassle.


Vim users are a niche within a niche. Not all developers use Vim or care.


Because not all developers are power users - but this browser is supposedly aimed at those of us who are.


Even as a vim devotee, I think that the suggestion that "power user" implies "vim user" is incorrect at best, and offensive at worst.


Of course. It implies "vim or emacs users".


Why can't they have stacked tabs and web page notes while using the Blink renderer? Saying that dropping Presto made those features go away doesn't make any sense to me. Its not like the Browser is built using its own renderer. The web renderer renders into the area of the tab ... the rest of the browser functionality exists outside that real-estate--and should be independent of the renderer.


I loved and used Opera as my main browser back in the day, flicking back and forth between pages with right-mouse-button gestures and reading my email in-browser before Gmail. What a fantastic browser it was! I am excited to see the Vivaldi project mature and hope to be able to evaluate it once it is more complete.


I gave Vivaldi a go because I wanted to speed up the testing of a website 'logged in' (with Chrome) and 'not logged in' (with Vivaldi). Vivaldi was notionally perfect for that with the same F12 tools as Chrome. I could delete session cookies and do my testing with it, no learning required.

So I would say I have given it a fair shot. But what browser am I using now? Chrome. Sometimes when you want to get something done you don't need to find that the maximise/minimise buttons are somewhere different, and with Ubuntu/Unity + Vivaldi they are - 90's style, in the top-right hand side.

I was actually a late convert to Chrome, for a long time I 'needed' Firebug and found the Chrome F12 tools to be inferior, or with a learning hurdle. With Vivaldi there is no leap of re-learning required, but, it doesn't have anything extra that I need for the day job. A regular incognito window does all I need for testing, with the UX chrome in the expected places. Note I also have the indispensable 'Hover Zoom' and 'Vim bindings'.


I really, really, really want to like Firebug/Firefox but it's just impossible for me. I already tried to switch 2 times this year. Once with the FFDev and once with v36 (stable) - both times I was left just as disappointed. I open the developer tools and refresh a SharePoint page and the browser hangs for close to 1 minute - this is just unacceptable. Even IE doesn't have this hiccups. I know this isn't common and is only reproducible on v36 for me but it's a huge PIA. Other than this, I'm curious why do you say the chrome developer tools are inferior? I haven't used firebug in a while but AFAIK Chrome Dev Tools has most of it's features and since it's "native" it's smother and faster.


Try it while not logged into SharePoint. My experience when SharePoint is logged its not exactly lean.


> I gave Vivaldi a go because I wanted to speed up the testing of a website 'logged in' (with Chrome) and 'not logged in' (with Vivaldi)

Most people just use incognito mode for that.


Why did they keep the built in mail client of all things? Fighting opera to get it to use my mail and RSS clients was the worst part of my experience with the browser.

If you want to make a customizable browser, give us the option to not install such unnecessary addons. The browser already seems to use as much memory as chrome, it does not need to waste any more.


This is something I don't really understand about the idea of a "power" browser. The entire reason browsers cut back on features was bloat that 99% of users never use. I guess targeting that 1% makes the bloat more reasonable, but for anyone who wants one feature and not another, it's easier to start at 0 with Chrome or Firefox and add the one thing you want.


On the other hand, the reason that Firefox got a reputation for being slow, bloated, and leaky was because 90% of the features were written by third-party developers with varying levels of competency, no quality controls, and little incentive to bring their stuff up to date.

I used Opera pre-12, and the reason I did so was because all of the features that I wanted were baked-in to the browser, which meant that they went through the same quality controls and integration testing as the rest of the browser. There was no separate step of updating add-ons, or even worse having outdated add-ons with no compatible version. Everything Just Worked.

Granted, some of the features were not as feature-complete as comparable Firefox add-ons. E.g., tab-stacking is strictly inferior to tree-style tabs because there is only one level of collapse. But everything I wanted was already included, at an acceptable-enough level that for me, it outweighed the cost of having to comparison-shop for mouse gestures or custom CSS solutions or whatever.

FWIW, Opera never felt "bloated" to me. It had crop-tons of features, but for whatever reason that didn't translate to a feeling of bloat. The browser was fast and responsive, and the pile of features never really got in my way. Every once in a while, I would read about an interesting Firefox add-on, dig into the Options menu, and find out Opera had the feature all along, whereupon I would drag it onto the UI.

Personally I like having the email built-in to the browser because it means I don't have to context-switch to see the status of my inbox. But that's a personal quirk.


I was using Opera heavily until about 6-7 years ago when the compatibility just became too bad. The feature I miss most is the keyboard shortcuts (which this seems to have) and 'fast forward'. This was a neat feature for paged sites where the browser would try to parse the site for 'next page' links, preload it and when you hit the fast forward button it would show it immediately. Worked great in all kinds of user forums and photo albums back in the day - although granted, there's much less paged content nowadays. I used to put back, forward and fast forward on the 'z,', 'x' and 'c' keys, sort of similar to vim (although I was never a vim user). It really felt like being in total control of your browser and having the web at your fingertips.


> even the ability to render a page with monospace fonts if you want.

Is there any major browser that doesn't allow you to do this?


In Opera you could quickly change these things for the current page in the menu. I don't think that is possible in Chrome or Firefox.


Not really my style (I'm a vim guy), but for power-user browsers, there's conkeror. Webkit rendering, emacs commandline.

http://conkeror.org/

(I mostly use it as a frequently-wiped-configs-and-cache bare rendering test.)


I don't personally see whats so exciting about this, it really doesn't tick any of the boxes for anything I'd want in a browser (as a 'power user'), this just reminds me how much I dislike Javascript.


I got this prompt when I opened HN in Vivaldi: http://i.imgur.com/HKMLXQD.png

When I click the ?, nothing happens. Any idea what this is?


My guess is that they want to use your password manager for Chrome to allow you to login to HN.


"The problem is bad enough that a future version of Firefox will even have a feature dedicated to letting you know which of your add-ons is slowing you down."

Does anyone know this feature of Firefox?


It's in Nightly. I've seen it a few times. It appears as a bar across the bottom of your screen telling you x addon may be causing this page to load slowly. From there you have the option to disable the addon, ignore the message, or ignore permanently. I saw it for uBlock, HTTPSEverywhere, and reddit enhancement suite, and amazonsmile redirect. In each scenario I felt it was well worth the tradeoff to run a little more slowly.


Worth noting that it's set to be hypersensitive right now for testing. Don't trust it for much of anything yet.


Good to know, I was starting to think either noscript/ublock were badly written, badly integrated or that Firefox had deep issues.



It's in current nightly builds. You get a status bar telling you that such add-on is taking to long when you load a page.


I installed it, but crashes every time I try to open it.


It's not open source enough for me. No thanks.


"first there was ie and Firefox"...

holy crap I am old ...cello anyone? :|


Their website needs Javascript to load. Ugh. I hope that's not an indicator for the rest of their philosophy.


The entire browser is built on JS/React. I'm pretty sure that is an indicator of the rest of their philosophy.


I don't understand why there are still people persisting in disabling JavaScript when the vast majority of Web sites today have significant functionality degradation or just plain won't work without it.


> I don't understand why there are still people persisting in disabling JavaScript when the vast majority of Web sites today have significant functionality degradation or just plain won't work without it.

Install NoScript and browse for a couple days. Permanently whitelist what you have to on sites you frequent and trust.

You'll find that pages load light-years faster, browser hangs/crashes are non-existent, annoying advertisements are a thing of the past, and actually, most web sites work fine without Javascript. Even though some of these websites have literally dozens (I suspect sometimes even hundreds: they sometimes chain each other and the full list takes up my entire screen and scrolls for a while) of scripts from different domains.

And then you get used to this web, and all of a sudden there's you run across what should be just a plain-text article - not a web application - and you discover that you have to enable Javascript just to see any text and it feels a little absurd. This is definitely the exception, and not the rule. My whitelist, after years of browsing, is pretty small.


As one of those people, allow me to help you understand.

A lot of websites are much quicker to load with NoScript active in my experience. It's quite an eye opener going somewhere like The Guardian's website, or Salon, and seeing the amount of 3rd party js that loads on every page. If one of those hosts is slow to respond, or down, it can often slow the whole pageload.

I also dislike executing fb, twitter, etc.. widgets all over the place, as well as giving over cpu time to chartbeat, scorecardresearch, and the like.

The tradeoff is that I have to right-click and selectively enable some js from site.com, othersitehost.com and aws-mumbo-jumbo-belonging-to-site.com for some sites that I care enough to read. Sure, the first time I load a site it's more work, but regular sites get necessary js whitelisted.

Personally, I'm happy with this tradeoff - I don't go around recommending it, but it works for me. HTH :)


The vast majority of web sites are crap. I'd really rather default disable javascript and opt-in if I decide I want to allow their code to run on my machine.


This. It's not really about tracking. Most of the websites embed tons of JS for no good reason, just because developers are lazy fashionistas and designers wouldn't be able to justify their pay if they didn't add tons of unneccessary bells and whistles.

There are web applications that need JavaScript, and I'm fine with that. There's lot of cool stuff JS could be used for, and it's not used - check out Bret Victor's explorable explanations. But instead doing something useful with this power, web designers and developers just use JS to make "shitty skeuomorphic bastardizations of what should be text communicating a fucking message" - as the quote from a very insightful article goes[0].

[0] - http://motherfuckingwebsite.com/


If it needs JavaScript to display properly, then it's probably not worth my time. Anyone concerned with privacy and/or security should use noscript.


Nobody in their right mind is going to practically double the amount of work they do on a project to support people disabling JavaScript (a tiny minority who also knows how to undo this change if they are so inclined) unless they are absolutely huge.

In any case, there are other ways to track you without the use of scripts.


Ironically, if you are reading this thread, you'll find that most people think websites have significant functionality degradation with JavaScript enabled.

This is the new web first world, where everyone is obligated to maintain finely-tuned block lists for various trackers and hostile JavaScript, run adblockers that prefilter sites and click on various odd-looking rectangles to get that plugin to run. The web is a joke.


It's not true that "everyone is obligated" to do those things; people who want to only selectively enable scripts and install extensions to achieve that goal are obligated to.


It's 2015 man, get off lynx..


I agree, w3m is much better. More modern. lynx sometimes has trouble making connections on my machine.

Sent from my w3m


Have you compared w3m to "links" (links2 on Debian/Ubuntu)? links is my preferred TTY browser, but it seems to be neglected by most users.


I compared. w3m is better than links2, elinks better than w3m.


And what do you think about eww, the new browser embedded in Emacs 24? I personally like it.




Consider applying for YC's Spring batch! Applications are open till Feb 11.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: