A browser where the security updates aren't married to "interface enhancements".
Just give me the security updates for the version I have, because I don't want your designer's latest idea of what a browser should look like, and unwanted new features forced down my throat with the security updates.
Choice about what new features to install - much like Windows updates, would be nice.
At the very least provide ways to keep things working and looking exactly how they are at the moment. Needing an add-on to bring back the status bar in Firefox is a joke.
See the history as a tree, not as a list, so branch points can be preserved and you can navigate back over everywhere you've already been.
A detailed development roadmap.
Privileged browser Apps and Extensions working on mobile. C'mon, it's 2015! ( no PhoneGap required )
An automation API.
The same really committed team who optimized V8.
Source on GitHub ( not necessarily all of it mind, some can be privileged access, tho a place to post issues that is no more code.google or crbugs, just, why? ).
For an automation API, would the RemoteDebug APIs³ work, or perhaps the command-line⁴ API for Firefox? See node-firefox⁵ for an example of using the RemoteDebug API to automate some functions in Firefox.
What makes you perceive the V8 team as more committed than other teams?
You can use your GitHub login to sign in to Mozilla's Bugzilla⁶, is that not enough of an integration for you?
I watched a couple of video presentations V8 did about optimization. It was really coherent. And their skills were really advanced. Also, visceral experience of the speed and stability of Chrome.
Also visceral experience deving Chrome apps and extensions ( man, I love their clear APIs ) versus doing the same in Moz with XUL ( omg, die ).
Browsers are as much about branding as anything, and I'm never going to come to the Mozilla camp. I don't believe a browser works to be a political cause, nor do I believe Mozilla is anything like the narrative they have about it. I mean what the hell is the committee of the Mozilla foundation doing? Wasting time with meetings about the grandiose purpose as a "movement", instead of making a better browser.
They pretend they're so "righteous" and yet all that pedantic dogmatic nitpicking and differentiating themselves is actually costing a lot of efficiency, which doesn't work for people who browse or for people who code.
Aside from that, I like the way Chrome does JS, and I don't like the way Moz does. Custom Elements, c'mon!
The masses have voted anyway. Chrome FTW.
Your answer was clear, it just seems we're on different sides of the browser divide. I wonder why there's so much loyalty? The browser is such an important piece of software.
A dev roadmap is just a general program requirement. Not specific to web browsers but good.
And you're right about mobile and extensions. There's no need to restrict when you don't have to! For a web browser on iOS, sadly you have to use Apple's engine but that's workable with js injection for extensions. And of course on Android there's no problem. I like that iOS is opening a little (I have an iPhone) but I wish it was more open like android, that way jail breaking would not be required for a lot of features. The one good thing now is that they're letting you side load apps from source with Xcode 7 without paying, but it's some michigas to go through. Mobile platforms are weird.
This may be not exactly what you are looking for but there is a session management tool for firefox. It allows you to save/restore the tabs (and their history) of a session.
For automation API you probably want a way to record actions and replay them like Excel macros? If you look for scriptable automation for Chrome then http://phantomjs.org/ might be a step in that direction.
I want to be able to stay logged into Google/Facebook/Linkedin for convenience without other browser tabs knowing anything about it...or isomorphically, I want to be able to browse the internet while logged into those services without the pages I am browsing leaking data back to those sites. I'll take the hit of downloading 100k of jquery on my 20mbit pipe a few extra times.
I want my browser to sandbox my information to the same degree I can achieve by running separate browsers or setting up special browsing VM's.
I'm not sure what you mean. If this involves "sign in to Chromium" then that seems to involve signing in to Google, and that sort of defeats the purpose.
Incidentally, Chromium is what I use for Facebook...and only Facebook, on this particular computer. And signing into Chromium with my Google account seems to me to be a big information leak.
Chrome/chromium has the distinction of user profiles[1]. You can have separate windows have separate profiles (but not separate tabs). This has nothing to do with Google accounts. Each profile has their own 'jail' of chrome resources (cookies/history/tabs/etc). This is really useful when you need multiple persisting sessions while doing web dev or whatever else.
In the meantime I can recommend the RequestPolicy[1] add-on for Firefox. It's not exactly what you're asking for; it blocks cross-domain requests entirely instead. Since it uses a whitelist you'll have to add exceptions for the sites you use (images and stylesheets are often placed on a different domain). This takes a bit of time but I think it's worth it.
- mouse gestures and switching between tabs with right-click + mouse wheel up/down or something similarly effortless
- tab stacking,
- and last but not least: if you want to make money off me, instead of sneakily tracing where I go and what I do and partnering with Google, ask me for it on first start and maybe with a discreet "donate/subscribe" button on the start page / speed dial.
Opera Unite would be a killer app with support for CloudFlare, so the pages would remain online when our computers were off, and auto sync of everything when the computer comes back online.
Also, with something like wordpress with themes and all the other silly things that made wordpress popular.
Chromium developer tools are very good, and the smartphone preview functionality is killer, but for really obscure JavaScript issues only the Dragonfly error messages made sense. They were clear an to the point every single time.
I remember the JavaScript that ran in Kestrel bug free, would run everywhere bug free.
Now, about mouse gestures and switching between tabs with right -click + mouse wheel, you can do that just now with Vivaldi. It only needs synchronization and supporting extension icons to replace Chrome in my computers.
Yeah, I use Vivaldi as my tertiary browser (or rather sometimes Vivaldi, sometimes Pale Moon, sometimes SeaMonkey) and I'm keeping close eye on it. It's still running on top of Google's code though, isn't it? :/
As for Dragonfly, it has "Resources". Chromium developer tools have "Sources", but I can't even tell if those things from Dragonfly's Resources are there. Mobile preview isn't very important for me as I use these tools for analysis and data extraction instead of actually developing, heh.
Anyway, a lot of the features that are critical for me are still missing and even if that was just the image properties windows, I won't uninstall O12. But Vivaldi is also missing (from those things I listed) other than Dragonfly also loading feedback, proper config (though it has vivaldi:flags, which seems to have extended from the last time I was there), structured history, the window bar doesn't say the page title, downloads are basic, ... all maybe tiny details to some, but they're huge quality of life improvements for me daily.
Cookieplux[1] is a plugin I made for Firefox, It is still waiting for approval from Moziall. But I have released the source.
So you can build it yourself.
I think Firefox + Vimperator is pretty decent by itself, it could be slightly better.
* Even better extension support -- enable writing extensions in any language: expose a file system based API for a full Unix philosophy experience. uzbl has something like this but it's not nearly as ideal as Vimperator is for my use case.
* As others mentioned, lose mandatory UI downgrades and other crap Mozilla's been pushing us lately: give us a platform to customize instead. (Practically running a Firefox compiled from sources helps with some of the issues people have, but sadly there will be a day they overhaul the UI in such a manner it hinders the usability. And because of the security stuff I just can't go back and use an old one forever.)
* Yeah, better bookmarking. Firefox could use a better system, and Vimperator doesn't solve it either. Full-text search among history + bookmarks would be pretty good.
The well exposed API for external apps could solve it all, save for crap UI. For instance, the bookmarks could be crawled to an org-mode document where I could annotate, categorize and tags things to my heart's content.
I second better bookmarking, a lot better than what is currently offered. For me it is the easiest and fastest way to keep up with places I have visited and want to return to.
I have bookmarks from back as far as Netscape Communicator.
Moving between browsers on multiple machines and keeping bookmarks the same on all is impossible other than by pinning a link to a saved bookmarks.html file. Bookmarks should comprise at least 15-20% of all browser activity. Restoring archived .json and .html bookmarks files never seems to work unless you have a small collection. There has to be a better way. Maybe I will just pin a text file and add URLs to that each time I need to save a location.
- works on X, the terminal and Android (yeah, that's a tall order, but it sure would be nice to have the same keystrokes available in X and on the terminal, and the same information available everywhere)
- securely shares passwords across multiple hosts (emphasis on 'securely,' unlike Chrome and Firefox)
- blocks ads
- blocks JavaScript, but makes it very easy to selectively enable it
- supports <script type="text/python">, <script type="text/common-lisp"> and <script type="text/smalltalk"> (hey, you did ask for ideal!)
- written and extensible in Common Lisp (elisp or Python would be minimally acceptable): the goal here is to be able to have 'the emacs of web browsers'
Write down their site / password info. They'll never know you have their login data, no access logs or warnings.
Note that the people freaking out the most about this are incredibly uncreative, because they think this is the only way an unattended computer can be powned, usually combined with weird beliefs about "Security" being a boolean value. Obviously, if you have physical access, you stick a keylogger on there, steal the whole DB of passwords at the binary file level, take over the whole operating system, etc. Also for extra comedy the people most likely to be outraged stereotypically have the same password for all saved sites (LOL) so you really only need to write down one password for that user, and also stereotypically its a variation of "1Password" or their kids name, etc.
I didn't know that the chrome://settings/passwords page existed until I read your comment. I actually just checked it out, and when I clicked the "Show" button on the password box, Chrome made me re-authenticate with my Windows password before it would show the password.
I don't know if this is a new feature, only available on Windows, etc., but it seems like this may be less of a concern now.
I did find it weird that I had to re-authenticate with my Windows password and not the password that Chrome is syncing passwords with, but it's better than nothing.
Sites that are heavily automated / packaged / locked down will not be up to date.
I can verify it works fine on linux 44.0.2403.107.
I have access to a 41.0.2272.101 on windows that is extremely heavily locked down and centrally distributed but I don't use that out on the internet, it would be non-trivial to test.
How exactly is any browser supposed to autofill passwords without storing them in plaintext? If you want them encrypted, you need to put in a password on every startup, which can be easily done by using truecrypt containers or fde.
Its development should be managed by a group who are not concerned with generating money from advertising/tracking or any other privacy reducing endeavors. A group who put user privacy and security at the forefront of every decision made.
I agree. There should be three main aspects of a web browser's development team's plan.
1. Stability and Features
The browser should be stable and free of major bugs, reasonable feature requests should be fulfilled when fairly trivial or found to be important. Programmers should be knowledgable in the language the browser is in. Personally I'd write it in C++/wxWidgets, using Blink as the engine since that's the most popular and compatible renderer it seems.
2. Privacy and Security.
The user's ability to control their information should be key, and it should be kept locally unless the user explicitly chooses to send it off.
3. Customizability
The user should be able to make changes to how the browser operates through extensions, userscripts, userstyles, etc.
Modularity.
I would like to see a browser that is designed as an engine, meaning that no UI/UX, or 'browsing' features are included. Much like a raw linux kernel in philosophy.
I would like this ideal browser to have interoperability with various JS engines, HTML and CSS renderers, as well as the standardized specifications for the underlying network protocols. It should be 'safe' by being written in languages where safety can be guaranteed ( with the exception of human error of course ).
In essence, if an open source core for a browser existed, maybe we could all work together and build the ideal browser.
"Ideal" isn't one thing. Allowing for personal customizations while following closely the standards make the base for implementing many "ideal flavors" of browsing experience.
I'm having a really hard time working out whether or not I agree with this idea. On the one hand my inner developer loves the idea. The notion that everything is a module, with well documented and robust APIs for communication between modules, is perfect. The browser could be all things to all people. But on the other hand there would be absolutely no additional value for a typical user. 99% of people (generalising, don't know the real number, but it's high) do no customisation to their browser. They don't even know they can.
A modular browser would only benefit developers, and we can already cope with delving deep in to the internals of many browsers if necessary. So is there really any point in a modular browser? Who would it be for?
Allow me to present an analogy between operating systems and browsers.
You can install Arch/Debian Linux or Ubuntu, and there is no collision there. You build your OS the way you want it or you run an automated installation and soon you are doing work, browsing, watching movies, etc.
My point is that if Debian wasn't what it is, you wouldn't have Ubuntu which offers additional value to a typical user, and be stuck with Windows forever. Choice is of value.
Do you want to be stuck with Firefox, Chrome etc. knowing how hard they adapt to real needs and ignore bug fixing in favor of new features?
A browser is not your 'typical' application. Browsers are in my opinion true virtual machines and should be treated as such.
I like the OS analogy, and it highlights the potential benefits well, but we also need to consider the downside.
It would load a huge amount of complexity on web developers. Rather than 5 browsers with a couple of versions of each, you'd need to start testing against a vast matrix of renderers, CSS engines, JS engines, chrome (as in browser chrome, not Chrome) plugins, etc, with versions of each and every one.
Testing software on Linux is hard enough that the economics mean an overwhelming majority of software manufacturers don't bother, or they do but they only support a very limited range of versions, or they release unsupported software that it's up to you to get working. That's definitely not something I believe we want to do for the web.
Well fortunately in the web we have already got standards committees. As long as those standards are supported there should be no problem.
Unfortunately current mainstream browsers don't comply with the standards to the fullest, although I admit compliance is orders of magnitude better than years ago.
Aren't Chrome only apps and web sites harming the web?
I agree with testing Linux software completely. It's maddening how much the notion of testing is ignored.
Well it's a win-win situation, build the ideal browser, then you have already built the "ideal" OS ;)
Yeah, it provides all the features of a browser without the UI for them. You embed the WKWebView (not WKView, I mistyped) in your browser app and call methods on it to load URLs, go back and forward, and so on.
Mozilla made a language in order to build their new browser engine, servo [1].
That language is Rust [2].
The reasoning was that we have been relying for too long on systems languages where memory safety [3] related bugs have caused a lot of exploits [C, C++].
In big codebases, like the browsers tend to be, bugs have many places to hide, and such bugs are very hard to spot. You can't rely on contributors being extra careful continuously, because everyone makes mistakes. So by using a programming language with a garbage collector or memory safety guarantees, like Rust, and enforcing extreme modularity, critical bugs are much easier to be spotted.
I wish it were easier to select text with the mouse.
On "complicated" web pages, tiny changes in the position of the mouse while dragging (selecting text) can result in large changes in the selection (the parts of the page highlighted in light blue). Sometimes a tiny change in mouse position even causes a whole 'nother column of text to be highlighted.
It sometimes takes many tries simply to select the contiguous words I want to copy. It sometimes takes planning and learning from experience. I would prefer for selecting text with the mouse to be so easy that I can always keep all of my planning and learning faculties focused on the ideas discussed on the web page.
Even on pages (such as the page where this comment will appear) where selecting text with the mouse is relatively straightforward, I would prefer to be able to select a contiguous series of lines (or paragraphs if individual lines is too hard) by dragging in the left margin the way that it is possible to select a contiguous series of lines in TextMate 2 or in Sublime Text 3 by dragging in the gutter (the left margin where the line numbers appear). (Neither Firefox nor Chrome lets me do that with HN's pages.)
In general, I would prefer for my browser to be optimized for "active reading" or "reading broadly understood" (reading plus related activities like copying a short passage from a web page and pasting it into a text editor).
I would probably make heavy use of such a "browser optimized for reading" even if it barfed on "complicated" web pages (i.e., failed to render all or part of pages that use a difficult-to-implement modern web API) provided there were an "Open in Firefox" or "Open in Chrome" item on the right-click menu whenever the mouse is over a link -- or some similar easy-to-implement convenience.
In other words, I think I'd use it and love it even if it was unusable with most pages on the modern web as long as it provides a good experience on certain sites with relatively simple pages, like HN and Wikipedia.
From a previous HN post - I want a browser standard for an easily optimised subset of html technologies. To conform to this standard, pages would be restricted in the ways they can manipulate the DOM, have a simpler DOM, use only a small fraction of CSS properties, and not use some JS features e.g. eval, delete.
We can use the asm.js model for opting in. Browsers that support it run the pages super fast, other browsers run them just as fast as usual.
A browser engine supporting just this standard would be considerably smaller, more embeddable, and a nicer base for current webkit based apps (e.g. spotify, steam, or atom). It might also help apps that want to use something like webviews for embedding content but need to be careful with memory / performance.
No antifeatures / downgrades. I am aware that this is a fuzzy concept. However, a few specifics:
* Think very hard before adding things to make sure that they actually need to be in the core browser. Looking at you, FF Hello.
* Don't baby the user. Everything should be configurable. And existing options should not be changed on upgrades. And general solutions are (always) better than centralized services. Looking at you, Firefox "You cannot install unsigned extensions, period".
* Don't introduce "features" that hurt the end user. Looking at you, Pocket "you now are bound by the terms of Pocket's user agreement when you run FF".
* Don't try to do cat-and-mouse games. Looking at you, Firefox "we must try to prevent malware on the user's computer from taking over FF".
* Don't assume that everyone has the newest and best computer. Looking at you, Firefox "electrolysis tripling memory usage is fine, right?".
* Don't couple security updates with anything else. Looking at you, Firefox "you must downgrade to australis or else your browser will be unsecure aah!".
Sorry, but I'm going to call this arbitrary FF hate.
With the exception of the Pocket and Hello arguments, Chrome has all of these problems. Electrolysis, the malware stuff, etc... all comes from Chrome stuff!
It's not arbitrary FF hate. It's frustration that the browser that I used because it didn't have any of these problems has decided to emulate the antifeatures of a browser that I don't use because it does have these problems.
As you said, it all comes from Chrome. Well... Guess why I don't use Chrome. And guess why I've switched away from mainline FF.
It's bloat. Memory use, startup time, attack surface, terms of use (although this is more with Pocket than Hello), number of things that need to be configured on install, etc.
I have no problem with Hello as a concept. I have every problem with it being part of the core browser as opposed to being an addon.
I know this is crazy but I want there to be a service where I can subscribe to a number of websites, and every week I get a printed hardcopy of all the content that I can leaf through. Eco-greenie-ologists be damned, I want my degradable, ephemeral paper to leaf through while lounging in a rocker chair drinking ice tea. Civilized. Now that's a browser.
Solves the security issue right there, aint nobody tracking me now. I'd like to see 1337 kidz xploit that paper. Escalate that privilege , from my rocking chair! And let me choose from a number of layouts. The smell of that fresh printers ink, nothing like it!
Bits and scroll bars and cookies. No taste, it's just trash.
You're asking for a individually customized newspaper with no ads or classifieds to support it, that probably needs custom editing every "issue" since no automatic formatter will work 100% of the time. I don't think 5 quid is so unreasonable.
Although the service did fold so maybe it was too much.
Ads are okay. Classy print ads are classy. 5 quid is way too much. Regardless of how the product is seen internally, 5 quid for what is perceived by the customer as a commodity ( a newspaper ) is too much.
What an interesting idea -- if I wasn't reading this webpage, I'd say it was a solution looking for a problem. But honestly I would love to try out something like this purely for novelty. It's like the definition of a horse-shaped car. (Or is that a car-shaped horse?)
Yeah, browser manufacturer, I know you want me to use HSTS and you want me not to browse to HTTPS sites without certificates and all that. That's fine. I can understand that (for example) Google wants to be the only org that can track where I'm browsing, not my ISP or the NSA.
But for the love of dog, give me an option to turn all that nannying shit off as and when I want to. My computer, my connection, I'll choose how much information I want to share with people. Not you.
To the best of my knowledge and reading of the provided link there's currently no way to disable "weak ephemeral Diffie-Hellman key in Server Key Exchange handshake message. (Error code: ssl_error_weak_server_ephemeral_dh_key)" type errors.
Its impacting me on an intranet-ish site that doesn't need https level security but the admins set it up incompetently. If they had set up https correctly, that would be a waste of time but it would be OK. If they had run the whole thing on http that would have been OK because there is no (... known ...) sensitive data on that site. But no they had to do a halfway job.
There may be market space for a bifurcation. Merely being able to render html doesn't mean all html rendering has to be done by one app, much like .ps or .pdf does not work best with the "one true app to rule them all". I could see a market with a iron clad virtualized one virtual (OS?) image per corrupt and insecure domain, and a hippie flower child browser for intranets and local files that blindly trusts everyone but by design won't talk to non RFC1918 ip addresses or maybe it blindly trusts the user not to do anything too stupid.
Per-tab CPU and memory profiling and volume control.
Never remove old history items when revisiting a page, but leave the history intact.
Treat tabs more like temporary bookmarks (and thus allow bookmark actions to be used with them like search (full-text), (batch) re-naming, arranging in folders, sorting).
Notes that can be attached to tabs and to websites.
A history of closed tabs.
Scripting support with REPL that allows to quickly process content of multiple tabs at once.
Transparent access to cookies and localStorage.
Something like Privacy Badger included.
Place dialog boxes always near the cursor, not in a tiny bar or in a pop-up near the top.
Bandwidth stats during page load.
Speed: No full-screen transition animations, lazy tab loading, smart tab unloading, ability to quickly and temporarily disable img, embed, video, script etc.
>"Treat tabs more like temporary bookmarks (and thus allow bookmark actions to be used with them like search (full-text), (batch) re-naming, arranging in folders, sorting)."
That's an awesome idea, does any browser have that?
Tree-style tabs and the bookmark model to go with it (a bookmark can have a parent, so you can restore subtrees of tabs, each with its history intact).[1]
Scroll below bottom of page as necessary when navigating to anchor links.[2]
Put focus in urlbar when switching to a tab unless the user has already moved it onto the page.
Continue to allow bookmark keywords for custom searches from urlbar (e.g. "w fun" to search wikipedia for "fun"). This still works in Chrome and Firefox but it gets harder to configure with every release.
History should be sortable by most recent view, where closing or switching to a tab counts as a view. Makes separate "recently closed tabs" menu superfluous.
Last time I checked ad blocking on Android was quite difficult to set up. It requires root, it runs as a background process, and I need to set up a proxy for each wifi connection. When the adblock process died, the internet connection would stop working completely.
I just want an ad blocker in the browser, like on the desktop.
Ad blocking on Android system wide using something like AdAway ranges from easy to impossible depending on the phone but you can install the same ad blocking extensions on mobile Firefox that you have on the desktop without any trouble.
> Last time I checked ad blocking on Android was quite difficult to set up.
Nope, just install Firefox, go to Tools:Add-ons, find AdBlock Plus or uBlock Origin, install it, and never see an ad again. No rooting, no proxying, no goofiness.
Only works in the browser, but it works well enough for me.
Why are ads addressed as browser related? If the question would be about "the perfect internet as medium" then yes, I would love to see an internet without ads.
Built-in ad blocker would just create a challenge for marketers to create a new solution and I doubt that they would come up with something that we would like to see.
I wouldn't want any browser to have built-in ad blocking. The best situation from my point of view is if majority of users gets to see the ads, while I don't. Who's going to pay for ads anymore if browsers come with adblocking by default?
"Who's going to pay for ads anymore if browsers come with adblocking by default?"
Why is this my problem? Nobody pays for buggy whips any more because we don't use horse and buggies. Times move on.
Apologies for sounding flip. I understand lots of folks have business models based on ads. This is most unfortunate, but it's no reason to screw over the consumer and not give them something they want simply because other people get paid money for it.
The consumer that wants ad-blocking already downloads an ad-blocker and gets exactly what they want. Making ad-blocking the default option for a mainstream browser means that people who don't dislike ads sufficiently to find a plugin that blocks them no longer subsidise the web for those that do.
It's your problem when you start encountering paywalls where previously only [blocked] ads existed.
That was a long time ago and even mailing lists and BBS had ads.
You can see it already on sites like Reddit and Buzzfeed, users block ads so the only solution is paid content hidden as user content. Be happy that, for now, it is more lucrative to separate the ads from the content, because the alternative is even worse.
In some ways, this last shred of free journalism we all enjoy online is because of those big ugly ads. Without them, no journalism, just illusion.
What else would you like it measured by? Paid subscriptions? Because we all know how that goes.
The breadth of online journalism we see today is precisely because of those ads everyone claims to hate. It allows anyone to quickly join the pot and lets the market decide their value. Without ads we'd be stuck with the 4 fuhrer model we had for the last 100 years.
Because a journalist needs a minimum of $50 a day, it's a hit based business, and readers are fickle. Everything you say must be in complete agreement with their world view to get a donation. One mistake and they're gone.
Also, it's been tried before and has failed every time.
A subscription based donation system might still work though.
Okay, so, let me premise my answer with the acknowledgement that probably /very/ few people want what I do.
1. Terminal-only (^Z is a thing that I want)
2. Capable of some basic CSS support (e.g., colors, basic text decorations; whatever can be simply translated to ANSI escape codes)
3. Capable of doing <div>-based page layouts fairly consistently.
4. Just enough Javascript support to login to a couple of webpages I like and nothing more
5. Some basic support for cookies and bookmarks (though, a session save would be an acceptable alternative for me)
6. Basic support for privacy/security features (think DNT header, HSTS, etc.)
7. (optional) libsixel integration for image/video output
Having the above presents me with the ability to leave GUI web-browsers behind entirely. Ads and the like would essentially be immaterial if JS and images are toggleable and everything would be glorious. Of course, no one appears to want this kind of thing enough for it to have been created yet. So I will eventually probably break down and make it myself :P
It's true that elinks covers a fair bit of what I want. But it does not have libsixel integration, does not have enough JS for me to login to the websites I'd like (afaik, perhaps that has changed since I last tried it) and does not handle <div>-layouting very well.
Admittedly, those three things are the hardest ones in the list.
Mostly security, privacy and usability related stuff. Here's a quite long list:
* Browser written in a way that prevents most exploits.
* Built-in Javascript and cross-domain request blocking.
* Single-process sandboxed browser where only the active tab is allowed to run.
* Tab content written to disk so it can be restored in the event of a crash or running out of battery.
* Ability to "swap out" tabs to disk so they consume no memory.
* Ability to freeze all javascript code on a tab temporarily.
* Deep searching of history (i.e. page content).
* Created by a privacy-friendly organization/people who aren't trying to spy on you.
* Avoid doing things automatically (download files to a folder, search for misspelled URLs, send URL to Google when typing it in, etc). Don't have keyboard shortcuts that are hard to undo (also, don't have single-key letter/number shortcuts at all).
* Don't have an over-engineered build process. The browser should be possible to build (not necessarily fast) on an old or low-end computer.
The reader browser should show you the "web" in a kindle-like readable experience, just as it does today.
However, the application browser, although it does receive a URL, should not have any HTML page as a default view. Instead, a JavaScript should be first loaded, and from there, following the permissions it requests, it can open new windows, notifications, or any other desktop feature needed, which websites that just need to be read do not need.
I can imagine an "application browser" being essentially a virtualized runtime browser - you point to the 'url' of an application, it downloads the runtime and source code and environment (if you don't already have it) and everything gets sandboxed and compiled on the fly, and runs as if it were native code. Of course, nothing would touch your actual filesystem, and you could get rid of an application as easily as you clean your browser cache today. This would also have the advantage of being language agnostic.
This is already technically kind of sort of possible with any number of sites that use compile-to-js languages, but these are invariably limited by the capabilities javascript and HTML. I suppose the ultimate version of this would look like browseable software images like Docker or something.
Reader browser: I would argue that it should be a reader/writer browser, where authoring, changing and interlinking documents is just as trivial as browsing them.
Application browser: but then why limit it to JavaScript?
I agree with the r/w web, and that was the original vision, but it's not really practical at this point because nobody supports such editing in a standard fashion
Interesting concept, but it seems needlessly complicated. What's to prevent websites like, say, the Guardian from declaring themselves as applications?
The reason why I originally switched from Firefox to Chrome was because Chrome was minimalistic and very, very fast - a chrome for the web.
Now I'm seeing Chrome become bloated, and slow. On top of this, it's now trying very hard to identify me by asking me to sign in with my Google account.
I want my next browser to make me forget that I'm using a browser in the first place, and also to not care about my identity.
Why not Opera? The new Opera is Chromium with what I think is a better design[1] that respects your OS[2]. You don't have to sign in to sync, but you can if you want. It's not open, but they're not an ad company. And finally, it's said to be faster than Chrome.
[1] I prefer Firefox so I use the FxOpera theme
[2] Well, it tries its best on Linux (unlike Chrome), but it's near perfect on Windows and Mac OS
Personally I don't use Opera because it doesn't let you customize mouse gestures. They consider it too geeky a feature :) http://i.imgur.com/fDSjYZq.png
Yes, because it manipulates the system in some way beyond copying the app to app, it's often hard to deinstall without traces, hard to keep two versions side by side etc.
So if the pain of the installer is larger than the added benefit of using Opera vs. Chrome, then yes.
Sounds like you might want to give Firefox a try again then – it’s much better now than in the 4.x days. Or if you’re on a Mac Safari is pretty slim (although slightly lagging in JS abilities).
In chrome, you can make a shortcut for a profile by providing a flag --profile-directory="YourProfileName"
It will be created if it doesn't exist and you can even set it to point to a TEMP folder for throwaways.
Ah, didn't realize that. Thinking about it, I've only used this 'hack' to have two different gmail accounts open at the same time (one in normal, one in incognito).
POST support from the URL bar. Let me type in POST bodies and edit Content-Type/User-Agent/etc headers by hand.
Show TLS negotiation in detail by default (which ciphers are supported? which was selected?).
Let me edit DNS blacklists & whitelists in text. Let me turn off the blacklist for trusted domains, so they can show ads and make money. Let me disable JS except for trusted domains.
Render .md/.markdown files directly. Support some extras (GitHub-flavored, for example).
Store my settings & bookmarks in git repos. Let me snapshot my bookmarks manually/periodically and see side-by-side diffs.
Perhaps as a large plugin, let me pull up magnet links to fetch & serve documents over BitTorrent.
When a link is opened in a new tab, that tab should keep the history of the tab it was opened from.
There used to be a Firefox extension for that, but that now doesn't work anymore because of changes in Firefox. Chrome closed a feature request for that with WONTFIX (https://code.google.com/p/chromium/issues/detail?id=1639).
EDIT: Oh, and also a simple user-friendly way to block autoplay of all audio and video elements. Maybe a "quiet mode" switch or something like that.
I would love to have a bookmarks file that is text, not SQLite, so I can use rsync or git to keep a master copy on a server and have all my computers share it.
Same with a passwords file, except make it encrypted with a master key.
Are you worried about privacy? Are bookmarks an attack vector of any kind? Or am I just naturally paranoid about these things?
I've always been wary of who can see my bookmarks, but not for any particular reason. It kind of reminds me how US libraries are vigilant to shield your checkout history from government searches.[0]
Yes, I care about privacy. That's why I want my shared bookmarks to live on my own server, not on delicious or Google Drive or whatever.
If they are encrypted before they leave the client so much the better. I would consider that a hard requirement for the passwords, so might as well do it for the bookmarks too.
Scalable tab management.
Being able to (name) group them without restrictions and being natively implemented on the browser without separate screens etc.
An open webpage file format which is not markup. Something of the like of YAML, or something simpler to read for a human.
Or maybe compiled HTML (like microsoft .CHM), except open (duh).
Reason ? Parsing XML or HTML is very CPU-intensive, and cheap smartphone's small CPUs (which have a very low L2 cache) have a very hard time dealing with it, not to mention classic desktop browser memory footprints. If not even regex can do it, why the hell not ? I'm sure the memory footprint of firefox and chrome is incompressible because of the raw nature of HTML. And for the life of me, if nobody can read a HTLM page generated anymore, what's the point of it ? Why not just compile it ?
I guess I could submit it to the IETF, to be honest. I'm not even a software engineer so basically I'm not sure if my idea is right or not, but I'll always find weird that the app-version of websites perform so much better than their HTML version.
In the internet explorer days, this was not possible because firefox and chrome were not mainstream, but I'm sure now is a good opportunity. It could eventually even replace or substitute .DOC and PDF files.
To not make me jump through 7 hoops until I can enter a self-signed https site and verify its fingerprint. Meanwhile they don't complain about http sites (which is reasonable), I don't know if they complain if you fill out a form with a password field on a http site though.
It would be nice if they showed a nice visual fingerprint too, although it's hard to prove their safety.
- No tabs, that's the job of the window manager. I get that most window managers are too underpowered to replicate browser tabs. Stop using those window managers I guess. Note that this gets you tab isolation for free, since every "tab" is just a completely different browser process. It hardly wastes any memory, too, because of dynamic linking.
- No features. I want to be able to set the url, search, have back/forward buttons, zoom the text. That's it.
- Extensible in a Unixy way. Press a key and execute a program with some hooks to modify the browser.
So far suckless's surf browser fits the bill. I currently use surf with a patch that lats me search duckduckgo from the dmenu url field.
- Support for privacy enhancing extensions. NoScript, µBlock, PrivacyBadger, HTTPS Everywhere, etc. I'm trying to see how these could work as a component separate the browser and I'm coming up blank on how to present the UI nicely. Currently I use privoxy for this, but it's subadequate.
How is the lack of "memory waste" caused by dynamic linking, if I may ask? I thought multiple copies of the same binary already shared text whether dynamically or statically linked?
Also, agreed: surf is lovely, except for the lack of adblock/selective noscript. And the cookie jar was pretty rudimentary last I checked.
Google Chrome is pretty close. Some extra things I'd like:
* Fully open source, with clear and open licensing.
* Extensions with a permissions system that works at multiple levels of granularity. So I should be able to trust an extension but deny it some specific things (without breaking it.)
* A headless mode, with python/ruby/other bindings, so I can use it as an automation library.
1. I totally agree with the open source part. Although I use a closed source OS often in everyday life (OS X, so at least it's a *nix) I'm a big fan of Linux (which I use on the side, and way before Windows) although personally I like the MIT and BSD type licenses over the (L)GPL
2. Granularity is a great thing to have as well. Extensions could maybe be broken into multiple sections, so you could have a- Permission to inject into page 2- permission to access saved passwords c- permission to access local files etc etc etc, and then you could set its local file root or what sites it can run in.
3. A headless browser sounds pretty great but with the nonstandard essence of web design I wonder how one would use it usefully in automation.
In answer to 3 - ideally in a way similar to http://www.nightmarejs.org/, well worth looking into if you've never played with a headless browser before
Apart from customizability, security and decoupling of UI changes from API/perf/security fixes:
- Option to stop all audio/video in all tabs
- Lazy session restore
- Reliable "save for offline viewing"
- Decoupling of "read later" from bookmarks and a separate interface for both
Going wild:
- Automagic unloading, archiving and clustering of opened tabs, perhaps with an auto-extraction of TL;DR from long articles
I very often open lots of pages (from HN etc.) but have no time to read them all at the time of opening. They linger as background tabs until the background tabs count is so big so I have to close them all and probably never come back.
Would be cool if the browser was able to somehow unload and archive unused tabs (to free some memory), and perhaps automatically group them into several categories: say, "angularjs", "bash", "youtube", "news" etc. and provide a good interface to return to them later.
However due to sheer amount of links I follow, my problem might be unsolvable :)
Zotero [0], with browser connector (extension), or alternatively a bookmarklet.
Originally intended as a research tool to manage sources, hence the added benefit of being able to export collections as BibTeX.
1. A per-website privacy mode. I don't want fb tracking me across all my private-mode tabs.
2. A separated out password store where the store prompts me for each password request and I have to allow it. The store would work with more than just browser passwords.
3. Better cookie control: I need no persistent cookies, apart from whatever Google uses for 2FA.
* Something other than javascript as a compilation target when using something else for those of us allergic to JS (scalajs.) I've only looked at it briefly but WebAssembly looks like a good candidate for this
* Less feature bloat, remove as many features as possible from the browser core and leave any non essential feature to plugins.
Basically just Chrome with Tree Style Tab, a ctrl-tab that does LRU tab switching, and an option to delay tab loading on startup until the tab is clicked.
I'd also like the browser to ensure that websites can't read my identity when I'm browsing in an unrelated tab. Something like a variant of RequestPolicy.
A content filter plugin, that allows me to block content semantically. Every year I feel the net gets swamped by certain events which affects my browsing experience. Halloween and the Superbowl are fairly regular, or the football world championship. People on facebook, imgur, twitter or 9gag and so on then just can't shut up about it. I would like to take notice of a new image of Pluto and then click "seen" and have the similar pictures blocked on imgur, twitter and news sites.
I would like to have this as a browser feature that I can control and easily change, rather than just a plugin. Adblock Plus would be a similar idea but I wish I could automatically teach it to block even more.
Vertical Tree Style Tabs (aka, Hierarchical Tabs) where opening a new Tab will make it a child tab of the original Tab.
Virtualized Tabs that unloads completely when it is no longer visible on the Tab Bar to save memory and CPU processing.
NoScript built in.
Ability to write Extension in any Language, aka, the Extension API is exposable to all Languages.
Tabs opened in the background should not start playing Flash automatically. Only once the Tab is selected, does Flash play, and continue to play in the background after moving to another Tab. Should be able to right click on the Tab and Allow Flash to Continue Playing in the Background.
Most importantly is speed and low CPU usage (running a hundred Javascripts per webpage literally kills speed).
My ideal web browser wouldn't be a browser at all, but instead a bunch of "do one thing and do it well" components that I can piece together to do things and swap in/out as necessary. I'd like to be able to swap rendering engines, Javascript interpreters, UI components, plugins, extensions, etc. in and out with maximum modularity and flexibility.
I'd also love for there to be a Unix-ish interface for browser plugins/extensions that isn't dependent on a specific language (or, if it is, for that language to be something C-compatible for maximum FFI compatibility). The more I can do without touching a single line of JS, the better.
MOBILE: when I zoom a page the browser should treat this the same as if I change the window size on a desktop browser, and reflow the text. Mobile browsing is a really lousy horrible experience because there are some conflicting philosophies.
No. It should actually limit horizontal reflow width inside element to screen width. That way you can always read a passage of text by using only vertical scrolling.
In other words, a block element's internal flow width should be restricted to screen width, so that a row of inline elements inside that block can never be wider than screen width.
In other words, text columns should flow at screen width.
Just doing what desktop browsers do would squish the text in a very narrow space. Not desirable.
For me, this is the single most important feature in a mobile (phone) browser.
-Easier text selection on e.g. websites where all the text is links or weirdly placed dives, basically a mode where anything that's text or looks like text can be selected and copied as plain text. And somehow don't surprisingly select all text of the website just because you started dragging from the right side of some sentence.
-Less continuous disk usage for caching and stuff, batch all disk writes together every now and then or so
-Something to block the new type of popup, the type that brings some thing to the foreground and shades the background
-Something to block all the annoying European cookie warnings, I don't need them, I already know websites use cookies
- Search through them better (by indexing their content and subject, autolabeling while allowing custom labels). (As a side-note, star-searching in Firefox is an undiscoverable feature; the new tab page should offer a way to do that.)
- Give a visual organization à la Desktop / iOS app layout. Currently, the new tab page gives the six most common sites, which subconsciously forces me to forget about the seventh most common site I visit.
- Give a neat access to offline webapps. Right now, going to one is inconvenient.
Most of all, include tab freezing by default (to remove the memory, stop the JS, etc. without having to close the tab)
Less memory consumption is often at odds with less processor usage. For example, a browser can consume a lot less memory if it doesn't cache decoded images, etc. But then the engine needs to decode same images over and over again.
My ideal web browser should be able to identify if a page really needs JavaScript enabled or not, and disables it if all the page does with JavaScript is load Facebook/Google/Twitter's iframes, advertisements, shows animations and popups etc. which have nothing to do with the content. Disables all autoplaying content, doesn't load images like large banners and footers automatically.
Kinda like switching to that "Reader mode" of Firefox automatically based on some AI.
I'd like a strong "project" concept: a project could be a set of multi-tab windows, there'd be fast switching between projects and fast tab selection based on projects. Most computers nowadays have a lot of RAM and people use lots of tabs. It seems wrong that one resorts to public browser extensions to find ways to navigate through the tabs.
Possibly also tree-like relationships between tabs so that subtrees of tabs can be operated on as a unit.
Keyboard shortcuts, single-key shortcuts for everything.
A good bookmark manager.
Customisable till the very end, through a simple to understand menu instead of the about:config way.
Minimal, no mail or RSS client, just the vanilla browser.
Superb in-memory caching for instant page navigation.
In fact, the old Opera had nearly all these features before they went and messed it all up, making it a chrome clone in a different skin. I was a paid user since the 2.x days.
should be simple and easy to navigate, understand what i might do.
Even if i open many tabs shouldnt freeze or crash, Thats one of the most important thing i care about, Mostly firefox does that, keeps crashing if you open many tabs.
Would be great if allows to sync with mobile, like if i enjoy a website and display right away on my mobile, instead of using extension it should the out of the box. We are living on a mobile world now.
As such a heavy user of bookmarks that I pay cash money for a 3rd party bookmark manager: literally any improvements would be great. The built-in bookmark managers of browsers - especially mobile ones - are incredibly anemic.
EDIT: The bookmark manager I pay for is called Linkman - http://www.outertech.com/en/bookmark-manager. I first bought it around 10 years ago and still use an old version, so unfortunately I'm neither up-to-date with its present capabilities nor the current landscape of 3rd party bookmark managers.
I book mark often. Unless you have rigid discipline your bookmarks will end up as a huge confusing list. Current bookmark managers feel like an afterthough. <tinfoil>Google don't want you to use bookmarks. They want you to search so they can mine data and serve ads</tinfoil>.
I'd pay money for a book mark manager that could organise my list of about 2,000 bookmarks. That software could have a "bulk bookmark import" feature so that users can share some set of bookmarks. "Here is my list of best pages about $ANIME_SERIES".
A bookmark manager that looked a bit like Pinterest, and that has options to locally save some content, is something I'd pay money for.
It's hard for large browsers to improve bookmarks because people use them in such varied ways. Changing bookmarks to work more like Pinterest may work well for you, but it may completely break someone else's workflow.
One solution would be to let you install third-party bookmark managers which would replace the built-in bookmark UI. That seems even more ripe for abuse than features like custom toolbars, though, which are already heavily abused in current browsers.
Using Firefox and bookmarking heavily with giving tags to my bookmarks. I would like to have a bookmark simple search language that allows me to combine the tags and narrow down the results, for example:
give me all bookmarks that are tagged "tools" AND "unix" AND "visualization" AND (NOT HAVING "monitoring")
There is an add-on called "tagsieve" that I used heavily, it had a nice visual possibility to filter the bookmarks with a tag cloud - though it didn't have a search language. But throughout the years often it was not compatible with new Firefox version (just see that it is installable again).
- full libre gpl3'ed codebase including plugins and extensions.
- immediate disclosure of all security problems (in the hope they get fixed promptly). the socalled "responsible" delayed disclosures annoy me to no end. let every exploit be a 0day exploit and leave it to the user to decide, at least it'll be an informed user!
Ideal web browser was here already, Opera between ~8 and 12 had everything from technical and user experience point of view. I could configure almost every option globally _and_ on top of that change behaviour per page. I could do whatever I wanted with UI. I was in the driver seat at all times.
Try NetSurf, if you haven't already. It's not perfect but it's a bit faster than Dillo is. (The build process feels a little wobbly to get used to but it honestly takes 3-5 minutes to build.)
I used Dillo between 209 and 2012 on an old 800MHz AMD box. It was really nice... being able to resume from hibernation, have 800 tabs open, and have a hyper-responsive machine. :D
- It must be ultra lightweight: fast when starting up, fast when handling load of tabs, just fast.
- Doing _one single thing_ exceptionally well: enable the user to browse. Built-in modularity means add-ons/plug-ins/etc can optionally be added to expand the browser capabilities.
So yeah: IMO a browser really should be just the browser with a good Addon-On API. That's why I would add a third point to your list:
- The Browser Team should provide (optional) Core-Plugins which are easy to install and in turn provide basic but useful features. For instance those mentioned in other comments here (e.g. uBlock, MAFF-Support, Mouse-Gestures, ...)
Yes, customizability and security (including privacy) are my top concerns.
Which is why I first and foremost want my ideal browser to not be owned or beholden to a major corporation that will abuse it for it's own purposes, like Google, Apple or Microsoft.
Firefox does something like that. If you start typing the name of the site you want, it will find the tab you already have open and give you a "switch to this tab" option.
Also, I feel like inspiration from the early web browsers should be taken. Tim Berners-Lee's book "Weaving the Web" fascinated me and I think that we can learn from both the current and the oldest designs of browsers.
I'd like its source code to be extremely hackable and easy to understand / build / debug, quite unlike Mozilla's present day codebase (which I personally find very hard to figure out).
Long term cookie saving only with opt-in by default - Firefox still will give every user the google cookie by default, handing millions of users out to google tracking.
A browser that knows and can retrieve quickly what information from the web I want, or likely require, at that point in time and without any expressed guidance from myself.
A browser without all this new stupid checks of "malware sites", HTTPS complaining and so on. I want to use it as a program, not as an secure antivirus tool
I would like an ability to snapshot save the web page, with all the dynamically loaded stuff (content, JS, Flash..). Just like some emulator systems have.
Sadly this would break a lot of websites and apps, since a lot of scripts are served from the Google (and other) CDNs these days. A whitelist of sorts could help though.
The problem with a whitelist is that tracking and malicious websites would just move their scripts to these CDNs.
It would certainly break compatibility with some websites, but so did removing flash or silverlight support, or introducing popup blockers before that. And with http2, scripts hosted on third party CDN are bound to disappear.
Yes. You could still serve ads from third party domains but these ads wouldn't be able to run javascript. This would be a protection in term of security and privacy. With no cookies or javascript I do not see how a tracking company could track users accross websites. And unless I miss something it would be a more constructive alternative to ad blockers.
[edit] in fact an alternative to blocking third party cookies is to restrict third party cookies to the domain-visited/third-party-domain, such that third parties can still use cookies to track you on the domain visited but that cookie can't track you accross websites.
This thing relies heavily on javascript, flash and silverlight. Both flash and silverlight require javascript. So unless I missed something, without scripts the only thing device fingerprinting can harvest is pretty much the user agent, which is pretty much useless for fingerprinting.
I've wondered about where UI and UX is going for a few years now. I think it's one of those areas where, if it weren't for the fact that corporate involvement would categorically be too politically incorrect, good (GOOD) ground-up UI/UX is a completely un-tapped field.
A few months ago I was thinking about the old MinWin improvements (to Windows' basic core services), how Windows 8.1/10 are (arguably) smaller and lighter than Vista, etc... and it clicked: Microsoft have realized they're no longer the "hardware industry killer app," if you will, and have moved out of the way... to let the Web take its place.
In my opinion, Web browsers are basically mind-bogglingly complex rube goldberg machines, based on conflicting standards that are also rube goldberg machines, and the whole scene is basically meta-rube-goldberg-ception.
A lot of the complexity and conflict is mostly due to the history and tradition behind the technologies we're using: on the one hand, all you want is arbitrary sandboxed code execution and sane UI; on the other hand, sites start delivering content exclusively via JS and everyone gets mad. And in the meantime, who audits their web browsers? What was that about "safe" code execution, where the browser protects you from "nasty" code? We hear about the browsers themselves doing the wrong thing by users almost every week.
We need to fight this. I've come to the conclusion that the only way this can be done is to simply make the effort and do the hard graft to port a bunch of existing applications to a new coherent, practical modality/paradigm, dump the whole thing on an unsuspecting world, and wait to see what happens next. It might catch on. It might not. Get feedback, try again. Keep trying until it goes viral enough to make a dent and improve people's lives.
There are so many interesting and fascinating UI experiments out there, but nobody seems to want to commit to actually practically trying out a few of them in real applications :( and I think that's what's holding everything back - everyone's waiting for everyone else to make the first move.
Here's where I'd start: everything should be able to send messages to everything else, to work around the language problem; everything should be an object with arbitrarily settable properties - everything from the files in the filesystem to the windows on the screen; with the appropriate security context surrounding everything, you could do crazy things like programmatically share the tags you attach to the files you share over P2P, or tag images with text descriptions of their contents; the console should be inherently textual but in such a way that it is exclusively touch-gesture-drivable; I strongly believe in the anti-mac/post-mac UX paradigm; I believe that it's impossible to create the "ultimate" solution that will work for everyone, and that all I can ever hope to do is nudge everything in what I think is a good general direction. And the only sane way to release an implementation like this is to put it into the public domain.
Agh, this probably reads like a rant, and I guess it is. I'm just not sure where to start, although I do have some ideas. Incidentally, I frequently use "asmqb7" on the web, including for my Gmail account.
Just give me the security updates for the version I have, because I don't want your designer's latest idea of what a browser should look like, and unwanted new features forced down my throat with the security updates.
Choice about what new features to install - much like Windows updates, would be nice.
At the very least provide ways to keep things working and looking exactly how they are at the moment. Needing an add-on to bring back the status bar in Firefox is a joke.