What Do You Want On Future Browsers?

Coach Wei writes "An industry wishlist for future browsers has been collected and developed by OpenAjax Alliance. Using wiki as an open collaboration tool, the feature list now lists 37 separate feature requests, covering a wide range of technology areas, such as security, Comet, multimedia, CSS, interactivity, and performance. The goal is to inform the browser vendors about what the Ajax developer community feels are most important for the next round of browsers (i.e., FF4, IE9, Safari4, and Opera10) and to provide supplemental details relative to the feature requests. Currently, the top three voted features are: 2D Drawing/Vector Graphics, The Two HTTP Connection Limit Issue, and HTML DOM Operation Performance In General . OpenAjax Alliance is calling for everyone to vote for his/her favorite features. The alliance also strongly encourages people to comment on the wiki pages for each of the existing features and to add any important new features that are not yet on the list."

On a related note, an anonymous reader writes "The Tao of Mac has put up pretty interesting list of five things that are still wrong with browsers these days, and I have to wonder — with things like AIR starting to be accepted by developers, do we still need the browser at all?"

Not just support for SVG, but mixed SVG/XHTML

[mrchaotica]

So browsers other than IE support (to varying degrees) referencing SVG drawings using the <img> or <object> tags. But that doesn't go far enough, IMHO; since both SVG and XHTML are both XML, I'd like to be able to embed either within the other, e.g. by putting a SVG polygon or circle on a webpage (surrounded by HTML), with another field of HTML embedded inside it.

I want what most users want.

[Lumpy]

More speed and less bloat.

Make it launch in 1 second and run for years without consuming much ram as well as render the page and all text FIRST before loading graphics and other crap.

I am tired of the bloated dead fish that browsers have become.

What do _I_ want?

[dotancohen]

What do _I_ want? HTML and CSS compliance. That's it. Get that done first then worry about the 'features'.

Comment removed

[account_deleted]

Comment removed based on user account deletion

Am I the only one who doesn't mind that much?

[gparent]

I do enjoy a minimum browsing quality. However, personally, all of the competing browsers currently on the market do what I ask them to. Yes, this includes IE7. Microsoft has vastly improved their browser and I applaud them for it. However, I think there's a point where feature packing has its limit. I guess you could compare it to Microsoft adding tons of bloat to XP and making Vista instead of fixing the outstanding issues of XP. I believe there's a point where browsers are just fine, and extra features would be superfluous. I thought Firefox 2 had attained that point until Firefox 3 came out, with its many performance improvements. At this point I only think that bug fixes and even more performance improvements are necessary. Vector graphics? No thanks. My work computer already has enough trouble loading Toms hardware and slashdot properly as it is.

Re:What do _I_ want?

[hackstraw]

What about <MATH>

Re:What do _I_ want?

[hackstraw]

Shouldn't reply to myself, but also what about media besides images and text?

I don't mean plugins, but a standard.

Re:What do _I_ want?

[Rob T Firefly]

What do _I_ want? HTML and CSS compliance. That's it. Get that done first then worry about the 'features'.

The problem with that equation is, the non-compliant crap still has major sway over the market since Average Joe Luser has it already installed on his new Windows box. You need to get the compliant browser into the average home, and the only way to do that is to give Average Joe the bells and whistles he wants and do it better than that pile of crap MSIE. The non-geeks need a reason to switch beyond "it follows some invisible rules you don't know or care about."

Re:Upload progress bar

[Ignorant Aardvark]

Two more things I'd like to see: native support for vector graphics (in the form of SVG) and native support for video (in the form of the <video/> tag and a Free codec such as Ogg Theora). The latter is actually already written, but Mozilla isn't going live with it yet because of patent fears from certain large companies.

How nice it would be to have integrated video support directly in the browser, though. No need for all of the hackish solutions, such as anything Flash-based, that have grown up around this gaping capability hole in the original spec. Make embedding videos into a webpage as easy as embedding text. That would be an amazing feature for a future browser.

Boobies!

[DarthVain]

Seriously though how about some decent security for a change. It would be nice to have a browser that doesn't let malware pown you system with a million vulnerabilities or so. Integrate an adware/spyware protection system.

That and boobies.

and tabs, and decent memory management. Speed is good also. Sharks with frikin' lasers...

Make it possible to select multiple files

[siDDis]

and not just one single file when I want to upload. I really hate to go that java/activex way to solve this issue today.

Re:Why only 2D Vectors?

[TubeSteak]

Give me 3D vector graphics, and let me play Battlezone in the browser!

3D vector graphics sounds nice, but (and no offense) I'd rather there was less convergence of the browser and the desktop environment.

Browsers are inherently buggy and exploitable, or include technologies that are. Until security is locked down tight, IMHO, we should not be moving to a place where the browser does more.
 
/If it isn't clear, I'm also not a fan of browser based webapps.

Re:Why only 2D Vectors?

[AKAImBatman]

A game like Battlezone is actually well served by 2D vector drawing. All you have to do is do a quick rasterization of the vertexes (x2d = x3d/z3d, y2d = y3d/z3d), then pass the result to the 2D vector routines. Rendering engine done.

While I can't view the site right now, COMET support sounds like one of the more interesting feature requirements. The only thing that I don't get is (and maybe this is explained on the currently-slashdotted site), isn't this solved by Server-Sent DOM Events [whatwg.org]? That effectively provides a smooth and scalable form of COMET support. Of course, only Opera supports it at the moment, so maybe that's the problem...

a rich-text editing standard

[brunascle]

browser based rich-text editing is a huge mess. of the browsers that claim to support it, there's very few functions that work universally, and everything else has to be hacked together. one of the 4 major browsers, up until the latest version, couldnt even create hyperlinks!

we need a standard desperately, and we needed it years ago.

Henry Ford

[Lank]

"If I had asked my customers what they wanted, they would have said a faster horse."

Maybe we should be thinking what do we want _beyond_ a web browser?

I want my broswer to well, browse the web.

[shypht]

I don't want it to read my email, or be my RSS reader. I don't want it to be an image editor, or a word processor, or MP3 player or media library. I would like it to be standards compliant, render web pages quickly, not consume loads of ram, and be stable. If I want any of the various 'features' as above, I'll take them in a plugin-format, or through a web application programmed to standards that can accomplish that task. Or, use a stand alone program for it. I want my applications to specialize in a few things and do them VERY well, I dont want 'jack of all trades, master of none' applications that implement dozens of features (most I dont want/use anyways), that don't do them very well, and add to overall bloat/instability in the application.

There are so many things I want

[mikael_j]

IMO the most important things for browsers in the near future is the following:

• XHTML and CSS compatibility - To save us all a lot of trouble.
• Memory footprint - It needs to be smaller.
• Stability - When I've got fifteen tabs open I don't want something in one of those tabs to crash the browser.
• Some form of page rendering where browsers are able to render page layout and text without waiting for larger images and such, perhaps by figuring out how to just fetch the dimensions of images from the server somehow.
• Properly sandboxed plugins - I want to be able to let flash run but limit the resources available to it, same for javscript and java applets..

If all this could be done then I'd be pretty happy with the state of web browsers and would stop complaining...

/Mikael

Re:stability?

[Darkness404]

This isn't 2002, browsers should be above that.

Sure the browser can be, but Flash is a plugin, not a browser and a poorly-written plugin for any platform other then Windows. So think of Flash as a program running in the background that display's the contents in your browser window. Can a program crash? Yep. So can Flash crash and make your browser slow? Yep.

Re:Upload progress bar

[CastrTroy]

Couldn't you just measure the amount of data sent out over the connection? If you only count the stuff that the server has sent back the ACK packets for, you could probably get a pretty good indication of the progress of the upload. It wouldn't represent the size of the file on the actual server, but it would be a really good indicator. I think part of the problem is that it requires going a little bit more low level than generic posting code that the browser would usually call, but there's no reason it couldn't be done.

Re:Upload progress bar

[Bogtha]

the reason for the lack of upload progress is because it's a limitation of the HTTP protocol itself. In order to upload you have to send the data in one big POST request and there's no way, via HTTP, to poll the results on the server.

You don't need to poll the results and it's not a shortcoming of HTTP. You know how much data you have sent, and you know that the server has received it because of the TCP acks.

So don't blame the browsers solely. To get this feature implemented would require modifications to the servers too. So the best way to get this feature implemented in all browsers (in a widely-accepted, standard fashion) is to call for an addition to the HTTP protocol.

No, it really is the fault of the browser vendors and nobody else. You don't need an addition to the HTTP protocol, in fact such a thing is pointless because it's already handled at a lower level of the networking stack.

My wishlist item: OpenPGP trust model

[Sloppy]

Built in support (i.e. enabled by default for millions of users) for OpenPGP trust model [gnu.org] for SSL certs. Kill the CA oligarchy by giving them serious competition, where an identity can be certed by any number of CAs, partially trusted through a WoT, etc.

Re:I want my broswer to well, browse the web.

[arth1]

If RSS rocks you, then by all means install a plugin for RSS. Don't force RSS on everyone, including those of us who have no interest in it at all.

Re:I want what most users want.

[dgatwood]

So true. Heck, I'd be happy if we could just get rid of all the web designers who build bloated Flash-based websites when simple HTML and a handful of graphics would look just as good and work much better....

Re:A Mute Button

[LMacG]

I agree. And I want that button to send a Taser(tm)-like shock to the developer who thought I'd want any sound at all to play automatically.

The user must be in charge

[Animats]

The user must be in charge. Not the remote site. Not any "toolbars". Specifically,

• All "toolbars", "branding", codecs, DRM keys, and other installed browser helper objects must show as clearly identified items that can be easily disabled, restored to their initial state, or removed completely.
• Nothing is ever downloaded to any place other than the browser cache without explicit interaction from the user. This specifically includes codecs and DRM code.
• Pages cannot disable menus or menu items. The "back" button always works, although pages are permitted to notice that they were reached via the "back" button.
• If the user chooses to disable popups, all popups must be disabled.
• All pop-ups must be on top. No "pop-unders".
• Pop-ups are treated as subordinate pages of the page from which they were launched. When the parent page closes, so must the pop-up.
• Ad-blocking support should conceal from the remote site that the ad is being blocked.
• Windows that are not on top should be limited in their resource consumption when they have active content running.

You get the idea. When it's user vs. website or user vs. toolbar, the user wins.

Re:Site Filter

[kRITek]

With experts-exchange.com just scroll down to the bottom of the page to see the content. Or use the cached version Google has.

Syncronization Primitives

[c0d3r]

Most ajax developers (NOT USING SOME FANCY/LIMITING FRAMEWORK) will run into basic synchronization problems that will cause major problems. Basic critical sections and thread safety primitives are needed. The closest I've found is an implementation of the bakery algorithm. Many of these issues can be solved with synchronous ajax calls, but for true asynchronisity, you'll need these primitives.

Re:I want what most users want.

[roc97007]

> ...as well as render the page and all text FIRST before loading graphics and other crap.

Didn't Mosaic do this? I wonder how we lost this feature.

> I am tired of the bloated dead fish that browsers have become.

Copy that.

#1

[Thaelon]

The #1 thing I want out of Firefox is threading.

Even IE has a separate thread for flash objects or other tabs.

It turns the FF browsing experience into one that is usually slower than IE and infinitely more frustrating when the browser is too busy rendering stuff in the background to listen to the user trying to use it.

Re:The user must be in charge

[nyctopterus]

I think allowing popups, or indeed letting pages control window size, position or arrangement at all was a colossal blunder. It's completely unnecessary today. I say remove it. Or, to maintain some sort of backward compatibility, have new windows appear within the bounds of the original. Pages should only control their own space, not control my browser.

Re:stability?

[morcego]

Besides the obvious answers ("Laziness" and "Microsoft did it to us"), there is the issue of complexity.

These days, systems are so complex that many times it is simply faster to reinstall.

I don't like this any more than you do. If you don't find the cause, there is a good chance you will have the same problem again.

Re:Why only 2D Vectors?

[TubeSteak]

Windows operating systems are inherently buggy and exploitable, or include technologies that are. Until security is locked down tight, IMHO, we should not be moving to a place where the Windows operating system does more.

Fixed.

Since you're so clever, please tell us:
Through what path do the vast majority of Windows OS exploits travel to reach the desktop?
A) Web Browsers
B) Desktop Programs that connect to the internet
C) Portable Media (CDs, DVDs, USB Drives, etc)
D) Other (Please explain)

Re:stability?

[m50d]

For example I can compile Firefox -O3 (or get a Swiftweasel binary) and it will run at a fast speed on lower-end hardware, Opera being binary-only doesn't allow this.

Ahahahahahahahahahahahahahaha. Go on, make my day. Do it, and do some benchmarks, or heck, just try actually using them both. I guarantee you Opera will blow your firefox out of the water, speed optimizations or none. There's only so much a compiler can do.

Number 2, it used to be adware and how can I really trust a browser that used to be adware, something that my browser is the first line of defense in combating it?

It wasn't adware in the way it's commonly used nowadays; it had one banner ad at the top of the browser, all revealed very obviously up front, and that was it. As for why you can trust its anti-adware capabilities, again, look at the results. And look at Opera's security record, and compare it to firefox or anything you like.

Also, even though it isn't adware, there could still be bits of the adware code in the source slowing it down,

There could be. But it runs a lot faster than firefox anyway, so until someone releases a slightly faster version, why does that matter?