Google Has A Solution For Internet Explorer: Turn It Into Chrome | Hacker News Reader

Google Has A Solution For Internet Explorer: Turn It Into Chrome

techcrunch.com

245 pointsjasonlbaptiste92 comments

I think someone should write internet worm that upgrades IE on the computers it infects, spreads for after a while then closes hole through which it spreads (so the computer won't get infected again) and then delete itself. The fact that such worm does not exist yet after all years of IE6 suffering might mean that worm writing hackers are assholes.

... Or that most worms are written by Microsoft?

No one legitimate is going to do this because of the legal risks. No one illegitimate is going to do this because it's probably not as profitable as building a botnet.

Alas, it seems that the intersection of "brilliant coders", "criminals", and "philanthropists" is a vanishingly small set.

Do you imply that people who ignore some laws do things only for profit?

No, he didn't.

Nope... Google might pay companies like dell and hp to sneak this baby into pre-installations. Woo. And I am sure there will be a google javascript thingy which detects IE and gives you a message about download this and supporting pages will magically run soooo much faster and look sooo much more awesome.

That's only half the problem. The other part of the problem are the large number of companies that force their employees to use IE6 (or if they're lucky IE7), combined with the large number of in house web apps that only work with those browsers.

You can't convince admins to install IE7 if it breaks IE6 pages, but you can convince them to install Flash. This is an opt-in plug-in on the server side. It seems like there would be far less opposition to installing this by admins if Google packages and markets it correctly.

Those sort of worms tend to cause more problems than they solve. The nimda-counter worm (or something else) comes to mind. The people who need IE6 will revert to IE6 anyway.

I know that such worm could accidentally be harmful. But IE6 is also very harmful. Security issues in IE can attribute for large fraction of infections with malicious software.

Drugs can have side effects in some percentage of treated patients but if they help more than hurt they are still used, especially for diseases where no good treatment is available. Just like in case of IE6.

Probably only small fraction of current IE6 users needs IE6 and would revert to it in case other than reinstalling windows after total system failure and thus involuntarily bringing IE6 back on board.

I was working on a project to do this, and I had made considerable progress. But when I decided to show it around to "some places" the reaction was overwhelming towards not to do it that I became discouraged and stopped.

Perhaps now the reaction has changed, as you seemed to receive a positive response. Or perhaps this is a selected community.

Many people in this field do not want something like this to become around though because it is such an easy target.

The other problem is many of IE6 are corporations and corporate machines. This will not help them, only other users.

I will re-investigate.

I guess there is no good way to weeding out IE6 from corporate environments. Maybe MS should just revoke their licences to use IE6 and sue them?

It's interesting question in general how to weed harmful piece of software out of ecosystem.

raise a fund to pay the license fee to upgrade everyone still using windows 2000 or less that are indelibly tied to IE 6?

That is disheartening to hear sorrow (wow the name matches). Honestly this just made my eyes glitter with hope that we can finally drop all IE support. Its an internal web-app and some terminals our users use must have IE6 because of compatibility issues with other crummy web apps (upgraded to 7 recently) which means we just need to test in google chrome and Firefox and were almost guaranteed to work in safari and guaranteed to work in actually all major browsers :). Its perfect! God I love google sometimes.

This is entertainingly insidious. Can anyone think of similar instances where a flexible plugin/extension system was subverted to basically turn one product into a competing product?

It's entertaining but it's not insidious. It's a nice bit of additional PR for the "IE Sucks" movement, but no one will use this. The people who are using IE6, by and large, don't have the ability to install stuff on their machines. Assuming any of them know/care what browser they are running. Most of the time when I answer support emails with "What browser are you running?" the response is "I don't know-- how do I check that?".

I agree that this won't make any difference to people who are totally clueless about browsers and browser compatibility, but extending the idea might make life easier for others; if you could run IE6 in "IE-mode" for a whitelist of sites that you knew required compatibility, "Chrome mode" for all other sites, and toggle between them on demand, wouldn't that be pretty useful? Assuming that you have any need for IE at all, of course.

Of course, when you think about that scenario, it starts to seem insidious the other way around. People might be less and less motivated to write cross-compatible websites if it becomes easier and easier to work around incompatibility.

Are all users allowed to install browser plugins? That seems like the motivation for this. Otherwise, it seems pretty useless. IE 6 is the main problem, and the problem exists due to a lack of administrator privileges.

I wouldn't consider IE6 the "main problem." In fact, IE -- at large -- or the Microsoft, IE team should be considered the main problem. Even IE8 is horribly far behind in CSS, HTML5, JS, et al implementations -- the very thing this plugin seeks to circumvent, in more than JUST IE6.

I agree with the notion that it is still useless to those IE users that don't have privileges to install addons/plugins. However, perhaps in those IT shops where administrators are required to keep around legacy browsers for legacy plugins / software, this is a attainable middle-ground for the administrators to give their users a better foot to stand on.

Really? It looks like a site has to opt-in in order to use Chrome Frame, so you could almost look at it like Flash.

Consider the following scenario: YouTube releases a suite of new features, but they require some new functionality from HTML5. This is fine for Firefox, Safari, Opera, etc., but if you're using IE it'll prompt you to install Chrome Frame, just like it prompts you to install Flash. I could see it getting to the point where people instinctively do this just as they do with Flash.

Hell, maybe it could even be used to subvert Flash itself. Suddenly IE users everywhere are viewing <video>-based YouTube videos.

Window is bundled with flash. So is Mac OS X. That's why it's so prevalent. So most people never have to actually install flash.

I don't see any reason to believe this particular plugin will take a different trajectory than Gears, which has almost no adoption.

Flash got it's biggest install jumps when MySpace required it for uploading pictures.

You're leaving out the fundamental detail -- Myspace was forcing upgrades to Flash 9, and the impetus was active Myspace worms that were exploiting security holes in Flash.

That's a very good point about it being bundled already. However, has Gears really been pushed that much by Google? I mean, even as an advanced user, I know that it exists and that various web apps use it, but it seems to be in the background and not all that advertised. I don't even use it myself.

I think a new browser engine offers a lot more opportunity for more tangible, up front enhancements to a user's web experience than client-side offline storage would, at least in an immediate sense.

Offline gmail uses gears.

But Gears had as many useful features, and it benefits people who actually care about their experience rather than people who arguably don't. Why do we believe Google's going to push this in ways they never did for Gears?

The difference is that gears enables new uses, whereas chrome frame can be used to make existing web sites better (canvas, faster JS, full CSS) and make web development easier. It benefits far, far more people at a much lower incremental cost than gears.

Most sites that have Gears support don't force the user to install Gears to use the site. There is no pop-up telling the users that they should get Gears to run the site - Google docs just has a nice little link at the top that says "Offline". There is no impetus to click it, and I bet many users don't even realize this feature exists. When the plug-in becomes required to use a site, and the user is presented with a pop-up telling him so, I think that changes things quite a bit.

OEMs bundle the ActiveX version of flash with Windows, so it works in IE by default for almost all users.

For Flash to work in any Windows browser not based on IE, somebody has to install the separate NSAPI flash plugin.

Yes, but you realize that this is a self defeating argument, right? People who aren't using IE had to install the plugin themselves. Doesn't impact people who do use IE, the target market for this Google plugin.

I did recognize the irony: users who've installed browsers have needed to install plugins, users who haven't installed browsers haven't needed to install plugins.

The extremely wide spread of IE toolbars and other ActiveX plugins shows that Google will garner installs. They'll probably also have the OEMs they're paying to bundle Chrome install Chrome Frame too.

Hell, they might even stuff it into the Google Toolbar for IE -- which has tens of millions of mostly paid-for installs.

Gears didn't provide users anything special, that's why it never caught on. But in this case, it provides users with a better web browsing experience. Our site doesn't support IE6 at all, in fact we just send those users to a web page saying that we'll support you guys in the "near" future. We could use their meta tag if the users are using IE6 and prompt them to install the Google plugin so that they could still use our website using IE6.

...don't have the ability to install stuff on their machines

But their IT departments can install it. Those IT departments usually want to upgrade because they have to write internal web apps and they feel the pain, too. But they can't upgrade due to various legacy apps that are tightly coupled to IE6. This lets them have their cake and eat it, too.

In my experience working in IT, the cost of MSIE6 is implicit and invisible and there's even some fear about security and legal problems related to free software. These companies are still targeting MSIE6 first and sometimes only in completely new applications; it's support for other browsers that is explicitly considered as an optional expense.

Good point, and the install isn't exactly as snappy as the ActiveX-based plugins such as flash. Here's a blog post that details the steps.

http://rickonrails.wordpress.com/2009/09/22/chrome-frame-ain...

Nice.

The only problem is that most of the people who still use IE6 are in locked-down corporate environments where they can't upgrade their browsers. They most likely won't be able to install this plugin either.

You have to wonder if Google have got some sort of killer app they want to launch but IE is the problem — the sort of killer app that corporates and institutions will want access too.

I don't see that they did this to raise anti IE feeling (unless they get off on preaching to the choir).

Don't see that WAVE necessarily fits the bill but it does need proper HTML5 support ...

What do you reckon?

Well, the wave people have already said they're stopping development for IE and will just be forcing the chrome frame: http://googlewavedev.blogspot.com/2009/09/google-wave-in-int...

As long as wave takes off it'll drive adoption.

If your job has your system locked down that much, most likely they don't want you to be surfing the web, wave, or whatever.

yepp. requires admin rights to install, which actually is quite surprising as chrome itself doesn't.

Nice idea. Unfortunately it requires running an .exe. In my work environment, running an unapproved executable is strictly forbidden. IE is also locked down pretty tight so I doubt that it would work even if I could run the installer. I can't even upgrade the Flash plugin, how bad is THAT?

> running an unapproved executable is strictly forbidden

By technology or words in a policy manual? If it's the latter, most employees can claim they had no idea they were "running a program"

haha, just the image of telling someone that "i had no idea it was a program" makes me laugh.

I can only imagine the support headaches that this will create. On one hand, the user insists (correctly) that he's using Internet Explorer. On the other hand, the observed behavior will be as if he's using Chrome.

Create a web page for support requests. Check the User Agent in addition to what the user reports.

Yes, but which part of "IE" is sending the User-Agent header? Does IE send User-Agent for the initial request and Chrome Frame sends User-Agent for all subsequent requests (after it detects the <meta> tag)?

I presume every IE request will have "chromeframe" in it, which conditionally serving the meta tag[1] would require.

[1] http://code.google.com/chrome/chromeframe/developers_guide.h...

it's only using Chrome's engine if there is a explicit meta tag on the site, so all old IE6 only internal webapps will be fine.

I wonder if this will work inside IE Tab in Firefox. So far as I can tell, it should (and without the performance hit of running VirtualBox inside VMware, at that).

Yo dawg, I know you like to browse, so I put a browser in the browser in your browser, so you can browse while you browse while you browse.

lmao

So uh, does this kill browser targeting?

http://www.alistapart.com/articles/beyonddoctype

Or can you just do something like content="IE=8;chrome=1" ?

Looks like you can target the user agent:

"Google Chrome Frame reports that it is available by extending the host's User-Agent header to add the string chromeframe. You can use server-side detection to look for this token and determine whether Google Chrome Frame can be used for a page. If Google Chrome Frame is present, you can insert the required meta tag; if not, you can redirect users to a page that explains how to install Google Chrome Frame."

http://code.google.com/chrome/chromeframe/developers_guide.h...

Oh boy. Another browser to support. Thanks Google!

Don't support browsers, support standards!

Isn't a valid part of "supporting standards" declining to support what ignores standards?

Yes, but you have to pick your battles. Unless you run an already ginormously successful site which can afford a substantial loss of traffic, then being dogmatic about web standards will only hurt you and do nothing to drive people away from non-standards compliant browsers.

... not sure anything is likely to drive people away from standards compliant browsers, maybe just sites that won't render in less capable browsers.

I reckon the user would blame a poorly rendered site on the author/owner rather than the browser.

Standards are the goal, and it's never been closer. However, interoperability is more important. Writing markup, code and styles to the standards and provide at least an accessible level of operability to the less capable user agents.

It's the same version of Webkit ued in standard chrom, according to the article. If you're already supporting chrome, it should only take one more line of html to support chromeframe.

This sounds like something like we'd complain about if Microsoft did it.

no no, if Microsoft had the browser, that passes Acid and Google had some broken mess you can't even render a diagonal line with, but somehow the vast majority of clueless users would use the Google Browser,.. then.. that would be a perfectly fine move of Microsoft.

So... does this mean IE finally has a chance of having a nice javascript engine?

For those folks who have nightmares over JavaScript's DOM performance on IE, this may be a remedy if it includes V8.

"Is our site running slow for you? Try this magic new plugin and the performance will improve!"

It does indeed include V8 and a big shiny 'Speed Up' button might actually be one plausible way to get people to install the required plug-in.

I like this. It will be much easier to convince people to install a small plugin than it would be to explain to them how to upgrade their browser. My app don't support IE6 for the back-end so no real way for us to loose by using this.

So does this just modify IE, or does it also affect aps that render HTML using MSHTML? This would be twice as hilarious/awesome if it made Outlook suddenly render HTML-formatted e-mail to standards.

Outlook 2007 uses Word's HTML rendering (no meaningful CSS2 support!), not IE's Trident.

I presume Google's goal is to reduce the percentage of IE6 users down to levels where consumer-oriented sites don't think it's worth supporting anymore? Today, IE6 is propped up by personal use of corporate machines during the workday. I worked on a consumer site where the percentage of IE6 use was higher during the workday than on nights/weekends. Various negative inputs into the IE6 ecosystem feedback loop could cause a precipitous death spiral.

I find this solution both hilarious and awesome.

Assuming no major downsides, I'd opt-in.

Right now, I'm consulting for a large company that still provisions IE6 because it has a bunch of in-house web apps that don't work right on any other browsers. Since switching thousands of users over to another browser could not happen in a single day, these legacy apps would all have to be modified to work in both IE6 and a modern browser. This can be done, but it requires coordination and resources. Furthermore, the groups responsible for the apps and those responsible for the desktop standards probably meet on the org chart at the CIO level.

Sadly, it would be really useful if Chrome (or Firefox or whatever) could accurately emulate IE6's behaviors for sites of an administrator/user's choosing. We could fairly easily make Apache insert a special "PleaseRenderLikeIE6" headers so those applications could continue to function as-is.

Yes, this is painful to write about, but it's the reality of most large corporate IT shops.

And doesn't this plugin kind of achieve this? Rather than marking sites which should be rendered in IE6, you mark them to be rendered in Chrome. So all legacy sites work without modification (using the IE6 engine), whilst modern sites can add the chrome meta tag and get rendered by Chrome.

If we could tag certain sites as legacy, we could uninstall IE6 and default users to Chrome/Firefox/whatever. Or, we could even install IE8 and use the Chrome plug-in for true IE6 compatibility -- the opposite of what Google proposes.

You could simply use the IEView plugin in Firefox to achieve the effect you are looking for. I have a few IE-only sites and it seamlessly uses IE to render those sites while in Firefox. You configure which sites automatically use that rendering engine.

You could also solve the problem in the other direction: install two browsers on every employee's computer. When that's done, you can safely migrate to a standards-compliant version of the internal web-apps.

I think there would be fear of a support nightmare. An automated email contains a link to the internal foo system. Which browser should be used? Should IE6 be the default?

You could solve that by using Google's extension, actually. If the page requires standards compliance, put in the annotation for standards compliance. Otherwise, let ie6 render as default. The employee using the webapp, ideally, wouldn't even notice.

Seems like this is only for HTML5 sites. IE 8 actually renders very well. I've yet to have something render in FF/Safari differently than IE 8. It's IE 7 and 6 that are the bane of my existence, and it's probably easier to get those users to upgrade to 8 than install a plugin.

Reminds me of the "IE Tab" plugin for firefox, which allows you to render your current tab in ie-format.

It's quite useful when you're designing websites, and also useful when your organisation has legacy apps that only display properly in ie6.

http://en.wikipedia.org/wiki/Lunascape_(web_browser) is unique in that it contains three rendering engines: Gecko (used in Mozilla Firefox), Webkit (used in Apple Safari and Google Chrome), and Trident (used in Microsoft Internet Explorer). The user can switch between layout engines seamlessly.

From reading the article and then the FAQ on the Google Chrome Frame site, I don't see how they'll get around users not having admin access to their machines. It seems to require a download/install which is the current issue preventing a large base of users from upgrading beyond IE6.

If your application is designed properly, you don't need admin rights to install - proper Windows installers are supposed to support a "This User Only" install method which shouldn't require admin rights.

True - and Chrome (the full install) already does this - it installs everything into your C:\Documents and Settings\<username>\Local Settings\Application Data\Google\Chrome folder. I think the main issue isn't this, it's that many of the corporate desktops that are still running older IE versions are no-installing-anything-at-all-locked-down.

I'm wondering if there's a population of users that have computers that are not that locked down, but don't have the savvy to install a browser. (My father's workplace gives him standard-User level access, but he's still been able to install Firefox.)

Additionally, this is probably wishful thinking, but I wonder if there's any chance IT departments would actually be somewhat friendly to this. Theoretically it shouldn't add much (if any) support load, since websites have to opt-in to Google Chrome Frame. An IT department could install it, still have all their internal apps run fine on IE and their users can have a better experience on the web without having to know or care about it.

(Of course, that assumes IT departments would actually want to give their users a better web experience to begin with...)

For distribution will they sneak it into google toolbar or offer an option to install it when you install Google toolbar?

Overall this sounds great, but there still is one huge problem; the average net user(majority) won't install this. They don't even know what a web browser is.

OK, some what I can feel Google CF is extremely insecure.

cf:http://tinyurl.com/google-bart won't display new 301/302 URL correctly on address bar and this is good for phising.

not found where source code for npchrome_tab.dll is, but RegisterNPAPIPlugin(), and UnregisterNPAPIPlugin() looks vulnerable since mixing two plugin mechanism into one is catastrophic.

Here's typelib for npchrome_tab.dll

http://initiative.yo2.cn/archives/642723#typelib

To be honest I think this is step 1, perhaps the best first step EVER MADE to move towards progress in HTML. With this plugin basically we can finally make messages on websites in the same way flash is installed to download google chrome and we have HTML5 compliance (once the spec is done and fully implemented). It will be great. Till now were stuck with HTML 4.1 until the entire world moves to IE8 AT LEAST. This is really a technological leap.

I'd prefer a more matter of fact source like this one: "Google Chrome Injects Itself Into Internet Explorer With Chrome Frame" http://www.readwriteweb.com/archives/google_launches_chrome_...

Microsoft will block this like Apple did with Google iPhone apps.

How very cool.

is it April 1 already?

Yo chromeframe, I'm happy for you and I'ma let you finish but Lynx was one of the best browsers of all time.

You know it's TechCrunch when the only credible thing you can find in that article are external. Google Code link + YouTube video. Otherwise, I probably would've assumed some kind of gossip reporting as usual.

Reply on news.ycombinator.com