Vivaldi – A new browser for our friends
vivaldi.com
vivaldi.com
Its not a new rendering engine.
So basically its just a different UI skin on chrome. That's cute, but its not a new browser.
It's the same browser with a slightly new UI.
Count me as skeptical and unexcited.
I am similarly unexcited by the 'it's browser' webkits views in android and ios, for exactly the same reason; they're dime a dozen, and lack any compelling reason to switch or use.
(Servo, by comparison, is a new rendering engine, with new features that make it extremely interesting)
But then I saw the part that said the browser's UI was Built on Web Technology, and then I read your comment, and my dreams were crushed forever.
Disappointing.
Seriously, I can't understand how they expect this to be any faster when everything on the homepage (which is still mostly content-free) says otherwise.
In ten years, people will be looking at the next next big thing because it has none of the cruft that <Servo> has.
(Not saying that it isn't important that people work on new, different engines as well in the long run)
Chrome on iOS is rubbish.
...because the Chrome guys are daft? Nope. ...because it's Apple's playground, and they don't want to play, so they've deliberately ('not deliberately') made it so anyone using their engine is a 2nd class citizen, and can only build a poor user experience.
(https://code.google.com/p/chromium/issues/detail?id=423444 for the meta bug if you're interested)
You can use UIWebView, the older interface. The biggest problem is that its JavaScript engine is slower than Safari's, and does not JIT compile code (according to Apple, this is for security reasons). I've also seen complaints about scrolling performance.
You can use WKWebView, the newer interface. Its JavaScript engine does JIT, and it boasts 60fps scrolling. Unfortunately, it's missing several features detailed in the bug, including a way to clear and manage cookies, custom URL handlers, POST request bodies missing in certain cases, etc.
One would hope that the WkWebView bugs and omissions will be addressed in future iOS versions, but today is not possible to make an iOS browser that works as well as Safari.
What makes it rubbish? I remember using it when it was released and I switched to it as my main browser instantly. At the time I felt it was better than Safari, if not only for the top bar being smaller and/or hidden, something Safari did not do at the time. It also had the "request desktop" which is invaluable at times.
Currently I use Safari but it's because I haven't bothered since I use the iPhone browser for almost nothing.
and its not Chrome's fault. Apple do not allow any other web browser on iOS, so Chrome is just WebKit except you don't get the same javascript engine as Safari does.
I stubbornly continue to use it because I want my history and bookmarks to match my desktop.
I also use my devices jailbroken, and there are fixes for the Javascript engine. That being said, I mostly used it without them, and even when I did I noticed no difference. Again, probably because I use the iOS browsers very little.
This is very similar to what Windows Phone is doing as well. You can't download FF or Chrome on their phones, just other poor IE clones. Granted, IE has come a long way and is pretty decent on Windows Phone, but if I could use one browser across all my devices, I'd be a bit happier.
Windows started the war when they locked out native third-party code development on the Windows Phone 7 platform. Then in subsequent findings, Chrome lamented they can't get enough privileges with Metro app development to make it work and Mozilla knows MS doesn't want to further their platform - so a sort of stalemate at this point, and a loss to Windows Phone users.
Sources:
Mozilla: No plans to release Firefox for Windows Phone 7 (March 2010)
http://www.zdnet.com/article/mozilla-no-plans-to-release-fir...!
Google Has ‘No Current Plans’ to Bring Chrome to Windows Phone (August 2014)
http://www.omgchrome.com/no-current-plans-to-bring-google-ch...
As a developer, I don't have to think about browser specific coding.
As an user, I can use same technology with favourite UI skin and additional features.
I have like 20 reasons why I miss the old Opera and I currently don't have any browser I feel comfortable with. All of them have to do with the UI. None has to do with the engine.
I was actually struck by its honest at first and wanted to try it.
http://www.phoronix.com/scan.php?page=news_item&px=MTgzNDA
A re-made Firefox in Rust could be not just much faster but also much more secure, even than Chrome, if combining Rust's safety features with Electrosis/a similar sandbox and multiprocess system.
Doing HTML and JavaScript are commodities. Given the existence of standards, there's not a lot of room for innovative behavior. Any developments (faster, new HTML standards, etc) won't impact the market unless they're in one of the major browsers. Competition is good, but there's only room in the market for a handful of rendering engines. Maybe 3-5, tops.
Conversely, UI and human interaction with the browser is a place where there's room for an unbounded number of players. This is where the competition can actually be happening. And it's where I think the competition should be happening.
The idea of a browser as something that just renders HTML and executes JavaScript is kinda lame. I use Opera because I think it's also the browser's responsibility to include tab organization and search functionality and mouse gestures and flipping CSS in and out and synchronization. (And I like my mail integrated, but that's a personal thing.) And I want it to Just Work, instead of having to manage a half-dozen extensions.
I am guardedly optimistic about Vivaldi, because it's what I actually out of a browser. (Specifically, an updated version of Opera 12, which Opera 15+ is not.)
> Given the existence of standards, there's not a lot of
> room for innovative behavior.
Servo is a research project for a future version of Firefox. That's the definition of a major browser. No web developers want a new browser with incompatible features to test or build against, but Mozilla/Google/Microsoft experimenting for future versions is expected.
The original point is that independent browsers using established libraries and adding new UI concepts is great for users and doesn't make extra work for web developers. I'm looking forward to what they make of it!
Then why have I begun to see "This website only works in Google Chrome, upgrade today!" messages all over the web?
I suspect a "This site only works with chrome!" message only applies to older versions of Internet Explorer.
Crying wolf is so boring these days.
edit: Also i find it hilarious that you think Google would risk rejecting all users who aren't using Chrome. Talk about a great way to give your Search competition a legitimate case.
edit2: The downvoting is cute, and i hope it's because of my tone and not me calling out lies. "Anything but Chrome!" he says - absurd. By all means, downvote my tone, but i should hope outright lying gets similar treatment.
It shows on old Opera (Presto), new Opera (Blink) and IE 11, so it's not just old versions of IE.
I see it often on Firefox, I have the latest Firefox on linux (35.0.1). It's a blue bar along the top of many Google properties that are decently cross browser. Go to Inbox without chrome and you'll see the "only works with chrome" crap.
The biggest difference is probably NaCl in Chrome right now.
Anyway, I think the TL;DR here is, "A new browser for our friends". To me, this implies they're building a new browser for the everyman, not a bunch of curmudgeon-y developers on HackerNews.
People outside of HN couldn't care less about new JS / rendering engines. They want to Google things, check their email / InstaTwitterVineBooks and generally just kinda meander around the web. To that person, new interface features is really the only thing they see. Honestly this is kinda refreshing to see in a weird way.
This new browser isn't just replacing Chrome's images with its own. It has new UI code, apparently using web standards. In other words, it's about as far from a skin as you can get.
When you replace images it's a skin. When you write code it isn't - it's a new UI.
It looks like this browser is a continuation of classic Opera, which pioneered a lot of stuff we take for granted today:
- Speed Dial — the start page you get in every browser these days
- mail client with filtered views, later known as "folders like in GMail"
- page modifications with user-supplied JS and CSS (later known as Greasemonkey and basis for browser extensions, which ironically Opera didn't have for a long time)
- address bar history based on full-text search of visited pages
- bookmark sync across devices
- pages rendered by a proxy server (like Amazon Silk browser)
And they've tried plenty of things that didn't catch on: built-in Torrent client, built-in web server (like node.js) and Widgets.
There's plenty of UI R&D to be done.
So I'll keep an eye on this for sure. Even though there are lots of missing features and couple of bugs that I noticed right out of the box. I hope they'll get to the point where they can maintain their own version of Blink, to get rid of some stupid choices they made (text selection for example).
Relevant (open-source) project: http://otter-browser.org/
Who even finds "select tab to the right" useful??
The placement of newly opened tabs / ctrl+opened links is on the other hand something I always find wrong in most browsers. It's seem there's no logic pattern in that.
By default, many browsers also arrange tabs in an order I find not entirely optimal. Tree Style Tabs / Tab Mix Plus help here.
Also, my biggest issue is with the inability to change this behavior. So I don't think there is one right way to do things (as there obviously isn't). I think such option should be included in settings, or at least let it be possible via extension (as is in FF). By the way, request for this option is one of the top voted issues in chromium (like in top 10 of all time I think), yet it was closed as WONTFIX without any comment from developers.
> or you could use a mouse
I try to not use the mouse, even while browsing the web. One of the reasons I liked Opera was the excellent spatial navigation feature, which sadly has no good equivalent in other browsers currently. Vivaldi claims to reimplement it, so that's interesting, but at this point I'm not going to risk more lock-in by switching from an open-source browser to a closed-source one.
+1 for affordances and maintaining the visual order of tabs.
On a related note, Shift+Cmd+[] doesn't work in Vivaldi for tab switching, unlike every other browser on OS X. Dealbreaker for me.
As for your first remark, why not both? MRU is most useful for cycling between two (or similarly low number) tabs. It's not possible with "go right" function. You have (and "always" had) other shortcuts for spatial switching, which makes sense in some situations. So you can (and do) use those. Why remove functionality for key combo that works like that in other environments? Also, you have the "index" right there in your OS as well.
it's like having random switching
EDIT: Sadly Vivaldi doesn't seem to support CTRL+PGUP / CTRL+PGDN yet.
By default it was on ctrl+3/4, and I don't think it's a wise choice as ctrl+number is good for either selecting first 10 tabs (chrome, ff), or for opening a speed dial entry (Opera=<12). I would actually like both of these functions, perhaps on alt+number and ctrl+number.
edit: Just tried Vivaldi and they've disabled both of those shortcuts that I use ALL THE TIME. Makes it more or less unusable.
As is Chrome for me because they disabled ctrl+tab :)
Either way, I think we can agree, that shortcuts should be something that user can configure.
Edit: Some more info here: http://thenextweb.com/apps/2015/01/27/meet-vivaldi-new-brows...
Seems as if it uses the Chrome engine. I wonder about the differences to Opera then.
Edit: Found also this on the homepage (somewhat hidden under the "Web technology" tab):
We use JavaScript and React to create the user interface — with the help of Node.js, Browserify and a long list of NPM modules. Vivaldi is the web built with the web.
So it's like http://breach.cc/ (https://news.ycombinator.com/item?id=8952152) then?
There are also a lot of libraries listed that are licensed under the Mozilla Public License, which is incompatible with the GNU GPL. Distributing this therefore might not actually be legal, regardless of whether or not it's released under a FOSS license, since those two license families have conflicting requirements for derived works (including works which use those libraries).
YMMV, IANAL, etc., but this thing's running the risk of legal technicalities killing it in the crib.
One suggestion: 'A new browser for our friends' is nowhere a good introduction. It simply fails to explain to me the reason why I should care this in the first place.
I think it's the first useragent string to contain 7 different browser/engine names.
It's ugly, as any user-agent string is. The uglier beast, to my knowledge, is this one:
Mozilla/5.0 (Mobile; Windows Phone 8.1; Android 4.0; ARM; Trident/7.0; Touch; rv:11.0; IEMobile/11.0; NOKIA; Lumia 520) like iPhone OS 7_0_3 Mac OS X AppleWebKit/537 (KHTML, like Gecko) Mobile Safari/537
Windows Phone 8.1 Internet Explorer.
The more things you put in your UA string, the harder it'll be to fall into the "this is IE" bucket.
Microsoft is trying to avoid UA-based blocking (e.g. Google once used this do block the Google Maps web app in Windows Phones) and styling (some sites only send the mobile version to webkit mobile browsers). Microsoft effectively masks its browser by changing the UA string to something which is detected as mobile Webkit, and this is (IMHO) wrong. Even if the site is "functional", a site owner should be able to opt to not support Windows Phones.
Suppose you are responsible for a site which carries a heavy and valuable brand. Suppose you don't have/don't want to allocate the resources needed to correctly develop and test your site in Windows Phone. You don't know how the site will run: you won't know about bugs, you can't guarantee a solid user experience and you can't guarantee performance. Ergo, you might affect the brand you carry. Although not ideal, I think it is valid for someone to say "I want to block my site in some device/OS, because I can't guarantee the quality and it might be bad for my brand".
I concede (and believe) that the end user should be able to circumvent these mechanisms by, e.g., changing the UA string. However, I don't agree with Microsoft approach of masquerading, by default, as Android and/or iOS browsers. Pardon my rant, but it's just Microsoft being Microsoft: bending standards at will and playing desperate when not it control of a given market.
"Title: Vivaldi End User License Agreement
7. Without limiting the foregoing, you are neither allowed to (a) adapt, alter, translate, embed into any other product or otherwise create derivative works of, or otherwise modify the Software ; (b) separate the component programs of the Software for use on different computers; (c) reverse engineer, decompile, disassemble or otherwise attempt to derive the source code for the Software, except as permitted by applicable law;"
And even if they could theoretically sue, there's no reason at all to think that they would. This is the former Opera CEO. He's not that kind of guy.
It's not as bad as Firefox though.
Maybe as matures it will introduce it's own rendering features and then your point will have more credence.
[1]: http://en.wikipedia.org/wiki/.local [2]: http://tools.ietf.org/html/rfc6762
I can understand why they don't touch the top of the screen: the tab stacking shows up above the tab. My suggestion would be to expand the tab to the top of the window, and then just make the tab stacking show up inside at the top of the tab. If not that, at least make the tab's bounding box touch the edge of the screen and leave the tab stacking on top of it.
I found a few more issues while typing this up: 1. The browser seemed to think my CTRL key was stuck or something. Typing in my password on imgur made it switch tabs and zoom out multiple times when I went to type some of the numbers in my password. 2. Copying the image URL from the address bar after I uploaded it didn't copy the protocol. The protocol was hidden since it's http, but that adds an extra step of typing "http://" when I wanted to paste it here.
The browser looks like it has promise. I like the look-and-feel. I like that the address bar is the progress bar when the page is loading. Being able to add notes and screenshots also seems like it would be incredibly useful.
Overall, I'd give the browser another shot once you've made some more progress.
I could learn to use CTRL+F4 to close tabs, but I find that slower than just middle clicking the tabs.
I'm still of the opinion that the space at the top makes sense. I sometimes drag windows to the top to maximize, and like the inverse operation of dragging from the top to restore -- behaves consistently like any other regular Windows application that doesn't mess around with their titlebars.
I doubt that they're going to offer an alternative mode because the developers seem more concerned with making their own custom behavior instead of just fitting in with my OS. They did a lot of work to break the functionality of a native window.
Sadly, you can mess yourself up if you go exploring in there, see a slider, wonder what it does and give it a go. On a non-HiDPI display anyway!
I wish this trend would end. The few applications I tried out that advertised being built on web technologies were not very well integrated with the OS (UX wise), huge in binary size, very memory hungry and noticeably slower than native alternatives. It's not a selling point, imo.
Especially not when you consider how big a dependency webkit/gecko is and how much bugs and attack surface you're bringing in and are responsible for patching in your end product.
I don't understand what the use-case is. Why expend all that energy to make a custom native UI platform when all you end up doing is breaking the standards for my native OS? On Windows, Internet Explorer is the only browser that just gives a simple native window. Chrome and Firefox both cause issues because of their custom drawn windows. I don't see any reason to use Vivaldi if they're going to follow the same path.
I wonder what their business model is. How does one earn money off a browser nowadays? They seem to focus on being a complete browser suite, old-Opera style. Do they plan to sell it as an app? Sell web services (mail etc)? Convert all encountered amazon links into affiliate links? (har har)
And I got an error when trying to register on their forums (edit: solved through IRC), oh well. I wonder if they even considered doing something similar to now-defunct Opera Unite. That thing rocked.
https://www.dropbox.com/sh/ljm9h48qn2x3v33/AAAmdrTw2ISsKKQNR...
oh for fucks sake it's goddamn node.js again. Stop this shit. I don't need a node runtime taking up 400MB on top of an already memory hungry browser. I don't need a browser using and abusing fucking javascript to run.
Looks good though, and features seem interesting.
Chrome and FF had a decent amount of memory leaks "despite" being written in C++
Blink-in-JavaScript is a mechanism to enable Blink developers to implement DOM features in JavaScript (instead of C++).
Aside from the fact that I don't need a goddamn 400MB runtime running all the time, the major problem is that node.js (and javascript, by extension) are awful tools for anything desktop related. (Server related too, but that's more a personal opinion that node.js is cancer (hi ted)).
Performance has never been up to par with C/C++ (no, testing a 100 line script is not a valid bench. No, comparing it against awful code is not a valid bench either. Which kinda invalidates 90% of node benchmarks.). But sure, let's trade a bit of performance for ease of use, I'm all for it. Except when that tradeoff implies using javascript, whose warts are known by anyone sensible. Having a single threaded architecture for a browser, which should basically be running one process/thread per tab is an error we shouldn't be doing anymore in 2015. I sure love one tab crashing my entire browser(hi firefox). Which I managed to do quickly with Vivaldi, but I can understand that it's still a beta.
Also, let's ignore the OS's native UI libraries because atom-shell/whatever-desktop-lib-they-use chose to render the entire thing as a DOM because reasons.
>Chrome and FF had a decent amount of memory leaks "despite" being written in C++
Nowhere did I hold Chrome and Firefox in high regards when it comes to memory usage. Chrome seems to think that allocating 4GB of memory for 20ish tabs is a good idea. Firefox is a bit better but has this slight tendency to leak memory like crazy.
tl;dr: browsers suck but that's not a reason to write one in fucking javascript
awaiting the nodejs defense force
As much as I like the idea of writing a browser using web technologies, most attempts so far have been pretty lacklustre. I'll wait until it's finished to pass final judgement on Vivaldi, but I don't really want the old Opera UI and features on top of the browser engine that the current Opera uses.
Also, I'm waiting for someone to just make a nice plain web browser. I don't want a mail client built in, I already have one of those. I don't want panels everywhere that I can't hide, nor do I really want a note taker. All I want in a browser is the minimal amount of UI around a fast browser engine that has lots of green boxes on caniuse.com.
Definitely keep it installed for now, probably won't use it daily.
1 (cycles bt. tabs?), 2, 3 (next tab), 4 (prev tab), z (back), x (forward)
`g thisissearch.com` would be interpreted as Google search query, not url. I think this is good behaviour.
Open Preferences->Search->One-click Search Engines, make sure google is checked, double-click "google" in the keyword column, change it to just "g".
If you want to keep the old "google" term or outright disable Google as a built-in search engine you can still add "g" as a keyword search. Go to Google, right click the search box, select "Add a Keyword for this Search..." and select "g" as your keyword. This works for search forms on most sites that aren't doing insane stuff with javascript or mandating a security token and it generally remembers other form settings as well.
The sad truth of browsers is that most people use whatever comes on their machine. A lot of times that would be IE - lots of people I know/companies in the UK still use IE out of a lack of knowing any better. Some more savvy users will use Chrome/FF but that's because they realise it has a lot of benefits over IE or are shamed into not using IE.
This doesn't really offer any 'wow' features that entice me as a web developer to try it, so your average user is even less likely to try it. And really, when making a browser, your target needs to be the masses because otherwise no web dev is going to support your browser if it requires even the tiniest of special treatment in their code.
Clearly not, as statistics show that IE is no longer the most popular browser in any country in Europe.
The "Opera community" is, or was, a thing. Opera spent years cultivating an actual community around their users, with forums and chat and social networking features galore. Then they shut it down last year.
The Vivaldi folks have also started Vivaldi.net, which appears to be an attempt to keep that community going elsewhere now that Opera has essentially axed everything that made people love them in favor of being an also-ran Chromium fork.
The My Opera site was what made people love Opera? Uh... No. Opera 12 had tens of millions of users, and hardly a fraction of those visited that site.
Having mail built in is not useful for me, but I wonder if it will support other email providers eventually (gmail etc)?
But, just fyi, you can switch off the color changing. At least in this preview release.
I'm assuming this on the basis that they features they are looking to implement were part of the old Opera featureset, which included a mail client.
But some of us actually miss those UI, so this is good news. Although I'm still disappointed by Vivaldi atm.. not the Opera 9-like I was hoping for =/
Someone is targeting Opera-like experience and damn, I like it.
I am not sure about liking it being another incarnation of Chrome inside though.
So I guess one could say there's a link between Opera and Vivaldi (no pun intended).
IE tried to do that years ago by coloring tabs, but then you ended with a rainbow in your face instead of really (abstracted away from you) grouping.
I also wonder why they are implementing an integrated mail client. It seems a waste of developer time which could be spent on something more important.
That said, the browser is pretty fast. Chances that I'll use it: maybe, but not as primary browser. I'm using FF with some vital addons as my primary browser (Firebug, NoScript, AdBlock, SelfDestructingCookies) and Opera/Blink as wordprocessor and spreadsheet on Google Drive, to separate concerns (customer mandated docs). I can't see me leaving FF for this (I didn't for Chrome) but it could end up replacing Opera if they keep it minimal.
Seems like a cross between Opera and Chrome. Will have to keep an eye out for this.
Is this finally a replacement for Opera?
* Why not by default ?
* Why is Ctrl+Shift+Tab and Ctrl+Tab working in reverse? Even when I change my preference to "tab order".
* It's not customizable enough, I want something that works like Tree Style Tab.
I love the future
Everyone used to hate Flash, but it's perceived as normal when the pages are bundled a good chunk of rendering/layout engine to run inside your browser engine. Can't track you and serve you advertisements with plain hypertext documents, duh.
No need to let your biases leak out when noscri.
Think about a "better web".
µMatrix/Policeman, do not allow requests to fonts.googleapis.com (IIRC fonts are served as CSSes from this host, but actual font files are served from the other hostname), done.