Finally. This is the Microsoft I've been waiting for! No more rolling over and no defeatist talk of adopting WebKit or whatever. Microsoft is going to use its muscle and position to make a truly competitive browser.
We need more competition, and we need the default browser in Windows to be just as good as Chrome and Firefox.
Microsoft, I hope you pull every -ms- vendor flaggin', ring-0 kernel hookin', micro-optimizatin', site-specific D3D driver tweakin' trick in the book as you reach for parity. This is going to be so fun to watch.
I'd like to see some competition as well, but in my perspective, they're taking the wrong approach.
This entire project seems to be about improving compatibility with older websites, by adding more algorithms and layers to the rendering process. I don't believe everyone is using Firefox and Chrome because they have better compatibility than IE.
I think they should focus on cutting the fat, and making the lightest weight, fastest and standards compliant browser possible, with some popular third-party add-ons available, such as ad block. Or, leave IE11 as the default now for compatibility, and spin off a new browser called IE Lightning.
If people start to switch over, then more sites will be developed to work in IE Lightning.
In short, create a faster browser with a smaller memory footprint and ad block, get users, watch compatibility fix itself, then make this ship as the default browser in 5 years.
Actually, our approach is effectively yielding the results you're looking for.
Removal of old IE legacy cruft is slimming Spartan's disk and memory footprint when compared with IE. Advances in the Chakra engine are pushing performance ahead. We're rearchitecting our DOM, which is yielding perf and security wins too. We're planning an extension platform also, with top add-ons like ad block being a clear target.
There's a bit of a catch 22 with "get users" and "watch compatibility fix itself" as broken compatibility is often cited as a top reason for users to switch browsers. It's hard to grow users without investing in compatibility.
So what we're doing is defining our "blend" of investments. Right now we have a heavy amount of interop investments in our blend as we think that's important for users. Over time (months, not years) the major interop gaps will disappear and I expect we'll see a shift in that blend to increase investments in other areas.
I strongly hope that there will be a drive among the major browser makers to do what it takes to make the future Web as powerful a platform as possible, even at the expense of delegating support for some old websites to legacy-support plug-ins. I personally would not have a problem with a dual-path approach to the Web: one set of browsers that specify a "current standards" Web platform with maximum capabilities, strict rules, and no legacy baggage, and an optional set of browser plug-ins, to which older, non-conforming sites are delegated.
This would allow the Web platform on small-capacity devices such as watches or headsets to grow in capability like an iOS device with its frequent OS upgrades and deprecation of older code, but would make the whole Web back to the earliest websites still accessible to somewhat larger and more capable devices (desktops, laptops) that could afford to include plug-ins for any old crazy code from the past.
I really don't want the Web to be the "we'll keep hanging on to the past" platform, while the native APIs are the "we'll keep bringing you the future" platforms.
I think you got downvoted for asking a foolish question. If there were linux builds in the plans, they would have announced it. If open sourcing the whole thing was planned... well, you get the idea.
I don't see what is foolish about the question (and was taught [and believe] that there are no foolish questions).
MS has opened up their flagship development platform (.NET) and even has it building on Linux. MS will also rent you Linux virtual machines.
The post also talks about interoperability and running across many kinds of devices, including "try out RemoteIE which will stream our latest browser from the Azure cloud to Windows, iOS or Android devices."
First of all, the team who open-sourced .NET is a different one from that that is building IE. DevDiv has been very open-source-friendly for a few years by now.
Then there is the fact that .NET is far less tied to the platform than IE is. .NET is nothing more than a VM, runtime and GC. That's fairly easy to get cross-platform, compared to a GUI application that makes use of platform libraries for hw-accelerated drawing and text rendering.
So when your browser doesn't work on modern web sites that make up 90% of what people visit, people will site "broken compatibility" and then move to Chrome or Firefox. You'll then reiterate your focus on the few outdated sites that no one visits stating that "broken compatibility" is the cause.
I think you may be mistaking broken compatibility with backwards compatibility. I don't use IE because it doesn't work with modern web sites.
"If people start to switch over, then more sites will be developed to work in IE Lightning."
Based on this sentence, I get the feeling you don't appreciate the value of web standards? If Microsoft wants to be taken seriously ever again (they had 10+ years to make that effort with IE before they lost the browser war) all they need to do is render pages correctly, not in their own special way. That also means not inventing special-to-Microsoft codes.
Microsoft has a long record of screwing the greater community for their own self-interest. For example, sabotaging Java.
I should clarify, I think they should be strictly following web standards, and by doing so, they help to force those sites not following standards to update their code, or die. I don't want my browser using extra resources, and having extra features and menus to improve compatibility with someone's sloppy, hacked together website from the 90s. I'd rather that site appear broken because it's not following standards, and by doing so, it's encouraged to update, or be replaced by someone else.
I believe this would allow for a smaller browser footprint, and we'd see better coded websites in the long run, instead of carrying them through.
For example, we get a new IE browser, that follows standards to the T. If your site doesn't work in IE, it's not because "it's IE, use a real browser", it's because you didn't code your site properly. In the video they talk about how they would previously look at the top 10k websites and ensure compatibility, and now they're looking at millions and billions. They shouldn't look at anything. If Reddit doesn't work in the new version of IE, because they're not following standards, then so be it. Don't look at methods of relaxing those standards and bending the rules to get Reddit working. Leave it broken, and let the users apply pressure to Reddit because the site isn't working in IE. As I said above, they can't laugh off IE anymore, because it's not some dumb, outdated browser. Instead, it's following standards, so Reddit is broken, not IE. It would be the gold standard... in standards. You check to make sure your site works in IE, because if it does, then you coded it right. If your site displays fine in Firefox and Chrome, you don't really know if it's coded properly, or if they just held your hand to make it display correctly.
I thoroughly disagree. Net Neutrality, today, is case in point. Until Chrome came along, Firefox was going it alone against IE. Until Google Fiber came along, we all had less than 10Mb downloads for $80/month.
There's much, much more. Pick better battles. Google is not your enemy.
A company can achieve good things, and still be an enemy. There are other examples of unsavoury things Google has done as well.
That said, I think applying morality to a company is silly; despite what the law may say, I'll never consider a corporation a person. It's a collection of people, and people can do good and bad things.
That's suggestive that Google is our friend, which it is not. It's a for-profit entity, thus just as much as our convenient friend as any other and in general an enemy whose sole purpose is to separate you from your money. Don't get it twisted any other way.
lol people like you are going to be happy when in a decade, everyone is going to be wearing google glasses, browsing through google chrome, having a google fiber internet connection, and 'renting' self-driving google cars. Cameras everywhere, control of the source internet, control of the browser, and I bet through some magic prominent universities decide that Go is the perfect language to teach beginners, over Python/C++/others- and start issuing free chromebooks to students.
Just look at how they've been fucking up google voice to drive users to Hangouts to get more control. Look at how sometimes, Hangouts randomly doesn't give you notifications of new messages, so that you're kept on your toes and keep checking your mobile frequently. Look at how google maps absolutely _degrades_ and the 'famed' UI starts to look retarded when you turn off 'Location sharing'. Haha, google is definitely going to have the last laugh. Edward Snowden must be crying somewhere right now.
There's going to be a day, when your 3 months old search for 'google services open source alternatives' gets you flagged as disqualified to being hired in Google/(maybe) other companies. And it's not going to stop. Google is going to keep getting bigger and bigger, keep growing tentacles, and keep getting more and more in control, and omniscient. I wonder how the great hackers of ages past allowed the whole technology industry to come to this, to let the situation become so bad. There should have been dozens of rms', but everyone kept quiet and rolled over. Now, it's near impossible to be free of google's touch when browsing the internet.
EDIT : Hmm, downvoted without a comment. Looks like the 'crawlers' picked this up. I'm scared now.
>This entire project seems to be about improving compatibility with older websites, by adding more algorithms and layers to the rendering process.
Uh... isn't that exactly the opposite of what they're actually doing here? Did you read the article? They're creating a new rendering engine that has all the backwards compatibility crap stripped out, so that they don't accidentally break compatibility while making changes to the layout engine they're using for legacy sites. (I.E. They're removing layers from the rendering process, not adding them.)
> Or, leave IE11 as the default now for compatibility, and spin off a new browser called IE Lightning.
Have you used IE in the past few years? IE is (by far) already the fastest browser on my two Windows 10 machines right now. That was also the case when they were running Windows 8 and 8.1 in the past. The latest versions of IE make Chrome feel like Netscape Communicator 4, and they're noticeably faster than Firefox too.
If it had a nice LastPass extension, Ghostery, Buffer, and a few others, I'd probably be using IE as my default browser right now.
Sadly, judging from FF and Chrome both, extensions lead to unavoidable slowdowns. Which makes them a "can't live with them, can't live without them" proposition. Every time I create a new profile, I am shocked by how fast it is. Then the avalanche of ads and tracking pixels forces me to install ABE or uBlock, and the downward spiral begins.
Definitely, that's probably one of the biggest dangers of opening IE up to more extensions. On Windows at least, the divide between Chrome/FF and IE is more fundamental though. Last month, I spent a week really seriously trying to evaluate switching to Firefox or IE, using equivalent extensions (essentially just LastPass, IIRC). Chrome was still awfully clunky compared to Win10's IE TP, especially around touch scroll/zoom/swipe inputs.
Speed is not the mark of a good browser. If you are using speed as your mark, then you have never coded for the web and have never gone through the trials and tribulations of IE. If speed was the target, then Lynx would be the best of browsers.
I agree that speed isn't everything, but switching to IE is often a relief these days when Chrome has managed to spin up my Surface 3 Pro's fan yet again just browsing average sites.
> then you have never coded for the web and have never gone through the trials and tribulations of IE
I've been developing for the web since before any version of IE existed. You?
Speed is NEVER the mark of a good browser. It can be a point on a list but, if it fails to properly implement standards or work in all devices, that's a failed browser. It's not the whole picture.
But you say you agree with my point.
If your Surface 3 can't handle the computing needs of a browser, that's not a resounding recommendation for owning a Surface.
As far as developing for the web, I've own a web dev company for 10 years and, among our list of clients, we manage two web sites you have visited before, one of which you probably visit every week or so. I was also, once, invited to work at Mozilla but declined. Does that answer your question?
So, you're telling me that Chrome running poorly on an i7 processor (while Photoshop, Visual Studio, WebStorm, Firefox, and IE all manage to run fine) is the Surface's fault? It's awfully hard to take any of your points seriously with these lines of reasoning.
Seriously though, everyone I knew who jumped from Firefox to Chrome ~5 years ago did it 100% for performance. We were switching to Chrome for speed even despite it lacking a lot of key features (e.g. dev tools, extensions, third-party cooking blocking, Flash blocking). Speed might not be the only important mark of a good browser, but it is a crucial ingredient.
html5test might not be the best comparison either. They test things that are not part of the official html5 spec. The scoring seems sometimes arbitrary.
On my machine, Chrome Beta channel scores 51% in 52ms and IE TP scores 44% in 41ms. If no browser is even making a passing grade on that test yet, it seems more aspirational than anything.
We have a already have lot of competition. In fact, it's never been better: Chrome, Firefox, WebKit, Chromium, Opera. You can even get the source code for Firefox and Chrome and contribute. I'm glad Microsoft is making a better go at it, but honestly, Microsoft could have built a more compliant browser any time they wanted too.
Chrome and friends aren't webkit anymore, but only by way of forking. Otherwise Chrome, Chromium and Opera 15+ are the same engine. The family tree kind of looks like this:
-- Internet Explorer
-- Firefox
-- Safari
\
-- Chrome, Chromium, Opera 15+
-- Opera 12 (abandoned)
I'm a little unclear on how differentiated Webkit and Blink are in practice, but sadly we have only 3.5 major engines competing at this point.
-- Internet Explorer 6-8
-- Internet Explorer 9-10
-- Internet Explorer 11
-- Internet Explorer new?
-- Firefox
-- Konqueror (abandoned?)
\
-- Safari
\
--Chromium, Opera 15+
\
-- Chrome
\
--IOS safari
-- Opera 12 (abandoned)
and that's why i use and only advertise firefox. It is the only browser that respect webdevelopers.
Konqueror is not abandoned, but Rekonq is generally the premier KDE browser nowadays. Any KDE install still includes Konqueror, even if it is not the default browser, because its backends power a lot of various parts of a KDE desktop.
What about chrome doesn't respect web developers? Chrome's web debugging in browser is very good, and until recently firefox didn't even compete in that regard.
Are you kidding? Firebug was the gold-standard for a very long time, and for a very good reason. If you're talking Firefox's built-in tools, then sure, but talking about Firefox while ignoring extensions is laughable.
"but sadly we have only 3.5 major engines competing at this point." I actually, don't think rendering engine competition is a great thing. I think competing browser products is a great thing. Think about node.js and io.js or rails and merb... it's not necessarily a good thing. Everyone kind of assumes that more rendering engines is good thing but I think it's just important that we have at least one healthy open source option. We have 3
blink
webkit
gecko
The closed source solutions are the ones we need to hope and pray die away trident. They're evil because there is no hope for any innovation as we saw in the last decade with IE6. MS can do anything they like to re-market their browser but as long as it's closed source engine it has no place in the web IMO. I hope it reaches 1% market share and is forced to go the way of Opera.
We got the Gecko family of browsers, the Trident family of browsers, and the WebKit family of browsers. Sure, there are varying JS engines, and Blink and Servo are on the way, but WebKit is the 900lbs gorilla. It is a monoculture in a lot of ways. I use and promote Firefox because I want to check the other vendors.
Microsoft makes, hm, how can I be positive here ... patchy software. There are always lots of rough patches reflecting pointy-haired boss decision making process, and there's cruft that builds up. But, very often you'll see these weird, beautiful bits just sticking out all over the place in Microsoft code. The weirdo geniuses at MSFT are allowed to do their thang and leave their mark. And the seminal releases of their software are always the best. Like, there are parts of IE6 that make me go, "how the HELL did Microsoft pull this tech off in 2000?" IE6 was this crumbly tower of awesome. If IE6 were a game, I won that game.
Anyhow, if development on Spartan goes RIGHT, then this browser will be the 2015 version of IE6. It will be the completest vision of kickassitude. There will be weird, crazy bits, and it'll take Microsoft and the rest of us a decade to untangle the proprietary parts. I can feel my blood pressure rising already!! The rage and the triumph and the glory!
IE6 was pretty damned awesome at release... IE5.x was a bit of a mess in that transition, but 6 was pretty damned awesome (proprietary bits included)... That said, it got stale pretty fast (by the time Phoenix was renamed Firefox) and slid even faster as others worked to make a better browser experience. The sad thing is, we're just now shedding IE8, and IE9 is still holding on... I'll be happy once I can get to an IE10+ only position...
For me, one thing I've been wanting to do is a compute heavy challenge-response for API requests.. meaning web workers, and that's only IE10+ ... will probably do a proof of concept with node (server) with scrype (client/server) against a pool of computed hashes... which could be very interesting in terms of reducing a DDOS against an API.
Haven't they? Since IE7, every new version is more and more compliant. IE11 gives a good experience and seems easy to work with (at least with the web dev I do). It just seems that they got sick of the baby steps approach and needed to start from scratch. They figured out yet another way to handle legacy code and want a new engine, and by new I mean, a Trident-based engine without some/most of the fluff IE has for legacy sites. Especially the active-x fluff that has no place in non-enterprise browsers.
I really don't see this as some big move. They've been doing this stuff for years. I think the consumer oriented Spartan browser is a clever move, but MS was getting there anyway. Seems like they're being clever and using Spartan and web-standards as a way to sell Win10. We're not getting these toys on 7 or 8. Want the new fancy browser? Get Win10.
No. Despite the experience of those who do a little bit of web dev, those of us who treat the web as our operating system still curse IE on a daily basis.
I hope that all those who currently talk about how wonderful IE11 is take note that Microsoft is, essentially, saying they're abandoning it for all these reasons.
I only curse IE9/IE10 when they don't support a moden feature, but I don't mind using display:table instead of flexbox (to name an example) as long as it gets the work done in IE9/IE10.
Although I don't bother testing IE10- to be honest, I just blindly trust the document modes of IE11, I'd rather spend time testing different mobile devices than testing different IE versions.
PS. I might be biased in the sense that I have been doing web development since like 15 years ago, and I'm already accustomed to the broken box model of IE. With Firefox I literally have to ask myself every time "what the f%@k did I do wrong???", but with IE literally is "oh you silly IE, I just have to set X and Y extra attributes and it's done".
If you are having problems in the other browsers, and not IE, then you're doing it wrong. Never, ever use IE as a reference for how things should work. THAT is the reason you are having trouble with the other browsers.
Your assumption is wrong, IE is usually the last browser I test. In my experience the rendering of Webkit and Gecko have more issues than Trident.
I don't remember every edge-case I have faced since ever, but for example, the following selector used to fail with Webkit browsers, and the response from Apple were "WONTFIX for performance reasons":
Also, in my experience, the sites that look broken in IE11 (ex. Facebook) is because the website author (or the compatibility lists from MS) set an old/faulty document mode via X-UA-Compatible.
I'm not trying to claim that IE is a better browser than Firefox/Chrome, because this is subjective (and personally I prefer Opera), but I'm trying to say that you must not develop a website in Firefox/Chrome assuming they never have faults.
IE is not as broken as people claim, and there are edge cases where IE is right and the other browsers wrong.
Opera, Chrome and Chromium are based off the same engine which is a fork of WebKit. If anything competition has gone down the last couple of years with WebKit based browsers becoming the standard despite its bugs & non-standard behavior. It reminds me of IE6 in its heyday.
No web developer worth his salt writes code for any one browser. You write code to the standard and test in browsers. Anyone doing differently should be fired on the spot. I'm not aware of any professional web developer who does that.
To second what Navarr said, I don't really see IE issues these days unless it's on a site that needs to support versions prior to IE9. As often as not, issues in modern IE are due to developers writing code that targets only a single vendor-prefixed version of an API instead of supporting standards, which makes IE look deficient (instead of that sloppy developer) even when it has good support for the APIs that a site uses.
To name just one thing that IE still doesn't support: setting the download attribute on a link. Neither does Safari.
This makes it very clumsy/impossible to write real web apps that need to save to the local file system. The scenario is generating a document client-side using JavaScript. Now you want to let the user save it. In Firefox, Chrome, Opera, and no doubt others, this isn't a problem. Just set the download attribute on the link to the file name you want to use.
> Microsoft, I hope you pull every -ms- vendor flaggin', ring-0 kernel hookin', micro-optimizatin', site-specific D3D driver tweakin' trick in the book as you reach for parity. This is going to be so fun to watch.
Well that's great, but if I'm on Linux and OSX this is just a pain in the ass for devs to support YAWB
Microsoft is going to use its muscle and position to make a truly competitive browser.
Isn't this the line with each and every release of IE? Why is this different? Each prior version of IE was proclaimed some revolution of some sort.
EDIT: Downvotes? Really? I see the front page is starting to get curiously Microsoft heavy again.
We've heard the "this time it's different" argument with IE7, IE8, IE9, IE10, and now IE11. Each time some shill would declare "finally! This changes everything! Microsoft has the resources and know-how to beat everyone" And the world moved on.
Do you or I really care abouts boasts from the marketdroids? Cut The Rope doth not a great browser make!
You and I know the truth. There are seminal, world-beating versions of software that Microsoft has released in the past.
And then, after dominance is attained, they touch up the pig's hair and makeup a bit in the subsequent releases. IE7-11 has been hair and makeup and some botox.
I don't know how Microsoft works internally, I only know history. And, reading their statements, Spartan might be a world-beater at least in quality, features, or developer interest. When Microsoft feels threatened, that's when the sparks fly.
I do not EVER underestimate Microsoft. How quickly did they turn that billion dollar Mojang acquisition around, 2 days?
Why the fuck did anyone ever think vendor prefixes are a good idea? If you want to prefix, prefix with versions.
I should be able to type `-beta-mask-image: url()` and any browser that implements that version of the spec can pick it up. There's no good reason to have to type the same exact thing 3 times to target each browser engine.
Because the vendors sometimes proposed incompatible variants of the same feature before it was standardized. And they would implement it behind their individual prefixes so that you could experiment with both the IE and FF variants, for example.
If you don't implement the same spec, bump the version. This is not a complicated idea.
Instead of -webkit-mask-image and -moz-mask-image, have -beta-mask-image and -gamma-mask-image. When IE comes along and wants to play with WebKit's implementation, they implement -beta-mask-image and all the sites that use it get IE support for free.
It was most often not the case that (a) vendor prefixed APIs matched a particular "version" of a spec nor (b) were interoperable across browsers at any point (e.g. -moz-foo didn't necessarily have the same behavior as -webkit-foo).
The issue isn't vendor versus version prefixes. The issue is sites using early versions of an API, relying on them, and then never removing/upgrading that code.
-webkit-border-radius for example doesn't necessarily match a particular version of a spec. Unprefixed, standard border-radius has been in all the browsers for several years [1]. Yet over 60% of page views (according to Chrome data [2]) still requires this property. So other browsers are forced to support the webkit version of this property for compatibility, which may not have a particular version of a spec for those browsers to reference when implementing support.
If you're interested, I gave a talk with other browsers at Mozilla HQ last week about some new ideas for how we can further improve the strategy for experimental APIs going forward. Check it out and give me feedback if you'd like (@jacobrossi on twitter).
No, that just shifts the burden onto every website instead of onto the few browsers. If they render differently someone needs to fix their browser rather than everyone has to build sites that work both ways.
Yes and no - our lives would be much easier if we could just assume that every user runs a WebKit fork. From testing to development, it would be enable devs to push apps to the limit without testing across multiple platforms. And I'm sure adding those vendor-specific extensions to a WebKit fork is also do-able. Since WebKit is open-source, I'd prefer to see companies put in more effort in enhancing its performance instead of developing new proprietary rendering engines (unless they are 10x better of course).
> Yes and no - our lives would be much easier if we could just assume that every user runs a WebKit fork.
Just as our life "was easier" when we could just assumer every user ran MSIE.
It was easier for a short while, then the rot set in. I'm not interested in having Google control the web in the same way, that's a road we've taken, we know where it leads and it's a terrible place.
Yes! Because trident wasn't different to WebKit and they definitely worked hard to make it superior!
Wait, no, this is Microsoft we're talking about, and they'll use their monopoly to force everyone to have the "new IE", and they'll break the web, again.
Honestly, I don't know why you think they'd do anything different. They've managed one release of IE that was actually innovative and good for its time - IE4.
I have little faith that they won't just produce some xslt spewing horror show that only works with drm monitors and complains if the backend isn't .net.
I'm glad to see that Microsoft is finally ripping out all the weird legacy modes -- IE11 essentially ships with 6 different rendering engines! [1]
* IE5 (quirks mode)
* IE7 (compatibility view)
* IE8, IE9, and IE10 (available from the x-ua-compatible meta tag)
* IE11's actual rendering engine
As an aside, the hilariously long user-agent string is perhaps the best evidence that string-based browser detection is something that web devs should avoid at all costs. It manages to include "Safari", "Chrome", "AppleWebKit", "KHTML", "Mozilla", and "Gecko".
>As an aside, the hilariously long user-agent string is perhaps the best evidence that string-based browser detection is something that web devs should avoid at all costs. It manages to include "Safari", "Chrome", "AppleWebKit", "KHTML", "Mozilla", and "Gecko".
I'm sure the reasoning behind those modes is because they had big (paying) customers who needed them there to help their "never ever getting an update" mission critical system webpages continue to work properly.
if you're not detecting on the client (via javascript), then you have to rely on headers. well, maybe via IE conditional comments + cookie setting on transparent pixel request
In an ideal world, perhaps. Sadly, in the real world, a browser seemingly supporting a feature doesn't mean it gets it right, and not all features can be detected by testing.
I really really hope Microsoft adopts the strategy from competing browsers:
- Make one version of Spartan and keep it AUTO-UPDATED, instead of shipping a new version every couple of years that is, more or less, completely decoupled from the previous version.
- Decouple the browser from the operating system. I don't really care if they make it cross-platform (though it would be nice), but if the browser stops updating unless you upgrade to the next version of Windows you WILL end up with legacy sites again and again that cater to a specific version, and then we can start over and just replace every rant against IE 6-7-8 with Spartan 1-2-3.
> Make one version of Spartan and keep it AUTO-UPDATED, instead of shipping a new version every couple of years that is, more or less, completely decoupled from the previous version.
This is what IE 11 currently does, so I would imagine they will continue doing so with Project Spartan.
We have actually been updating Internet Explorer for some years now. We don't touch machines that have opted out of updates, but for everybody else we've been moving them forward regularly.
Spartan is neat because it's a separate app, built on a different architecture that will not only continue to allow us to update in an evergreen fashion, but it will allow us to do so at a much faster rate.
I like the fact that you guys are sampling more and more of the web rather than sticking to the top 9K sites. This will make fixing issues in one browser less of an issue (hopefully I'm crossing my fingers).
It would be really nice to know what technical reason there was with freezing IE development on Windows XP before the OS itself reached EOL.
It has been standard practice for Microsoft for many years to ship the DLLs required for a feature in the OS as redistributables, features such as updates to OLE, new runtimes for Visual C++ or Visual Basic, new APIs, new images processing libraries and audio and video codes, and many other things. Software like Office would even ship with new UI components that could be reused.
Building a version of IE that did not require OS sandboxing features was also not an impossible task and would have maintained the backwards compatibility that Microsoft holds almost sacred better than a version of IE that was is still frozen in time, and is the last version that will ever be shipped for Windows XP.
This has broken the trust and relationship with web developers, even those that used IE as the gold standard and built their sites to match every spec and feature promulgated by Microsoft in the past. (VML, behaviors, filters for instance).
I work in an industry where compatibility with IE8 is expected, and spend the same amount of time that many others do maintaining a veneer of compatibility with that browser, that rendering engine, that HTML parser. This requires heavier libraries than I would prefer to use (such as jQuery) because it maintains an abstraction over the incompatible parts of the DOM implementation.
Another example of this is the incompatibility with SNI which has driven a requirement for unique IP addresses per SSL site and driven those than do not want to pay for such as limited resource (assuming IPv4 addresses) to share SSL certificates, use wildcard certificates that open issues of their own, or forgo the security provided by SSL entirely if they still require compatibility with Internet Explorer versions and OS versions that don't implement SNI natively.
This also applies to current more secure versions of NTLM which are and were supported by third-party browsers on the same OSs that Microsoft did not offer support, by using a non OS provided SSL library which implemented the Microsoft driven standard.
It's true that the IE team has done a better job of communicating with IE Blog and other channels, and has done a better job of getting out in front of upcoming standards, delivering impressive performance on complicated implementations (WebGL) and otherwise kicking some ass.
But they and Microsoft still have to rebuild the trust and support of the developers they lost not just as users, but the developers who have come to dread working with technologies they once loved due to the heartbreak that IE6, 7, and 8 inflicted upon us.
I'll end where I began, with a plea for an explanation for not shipping a newer version of Internet Explorer on Windows XP before the end of life.
And best of luck.
Edit: I include 6 and 7 in reference to current support requirements, not the advances they represented when first released.
But it will still be non-free, and probably won't sync history/bookmarks/open tabs with mobile, except perhaps windows phone. And it only works on windows.
I don't see what would entice users to switch from Firefox or chromium or opera (or midori, konqueror, etc) to a new Microsoft browser.
Its probably still worth telling users not to use internet explorer, just in case they ever end up buying a Mac or Linux PC, and out of principle to support free software.
Its good to see competition, but I theorize that this will hurt Mozilla and benefit Google and Microsoft. The quality of chrome has been stagnating recently due to its large market share, while Firefox has been steadily advancing (because Mozilla's primary agenda is to move the web forward). Users have started a slow migration back to Firefox, which I love. A good internet explorer, paired with effective marketing, could threaten chrome and force google to get their act together, ending the Firefox rennaissance.
But realistically I don't see Microsoft marketing this effectively, and internet explorers reputation as a piece of crap will haunt it for years to come.
Personally for me? I am deeply concerned that internet explorer becomes not terrible.
People would be stupid enough to then use it.
Then microsoft would have more clout.
They would immediately use that clout to cause enormous industry problems, push bad standards, fail to comply with open ones, push windows specific extensions, ect ect ect.
I am 100% concerned that internet explorer might not suck. If that happened, the internet would be in a much worse place.
This interoperability-focused approach brought the obvious question of adopting an existing open-source rendering engine such as WebKit. While there were some advantages, upon further investigation it was not the right path forward for two important reasons. First, the Web is built on the principle of multiple independent, yet interoperable implementations of Web standards and we felt it was important to counter movement towards a monoculture on the Web. Second, given the engineering effort required, we found that we could deliver an interoperability focused engine to customers significantly faster if we started from our own engine (especially if unshackled from legacy compatibility concerns), rather than building up a new browser around an open-source engine. We will continue to look at open source and shared source models where it makes sense and look forward to sharing more details in coming posts.
This is a powerful paragraph from the article. I'm all for competition, and if they actually produce the next great browser, that'll just mean better browsers overall (and probably even better tooling for us devs). I'm a bit skeptical whether they can pull it off, but at least I have a good feeling that they're out to prove us all wrong.
(general reply to some of the negative comments in this thread)
When someone is taking the right steps, I think they should be praised. Whatever bad feelings about Microsoft people have, they seem to be a newly revitalized company of late, are implementing cultural changes the way they interact with other ecosystems, and are shipping more open source, and seem to be playing well with others.
Haha very true. It reminds me of the time Apple announced that Internet Explorer was going to be the browser for classic Mac OS, with Bill Gates calling in, to shouts of NO!!! NO!!!!! NO!!!!
It makes me chuckle. It always boggles the mind that the web is mostly just rendering of documents and flinging backwards and forwards of requests, yet the passion it invokes with browser wars and web developers. A browser is essentially a HTML Help File viewer yet you get followings and wars! It doesn't make sense.
Even less sense is made when they reinvent all application features in a web browser to make web pages behave like desktop applications. Kind of useful, but an incessant broken/fix/reimplement cycle
Impressive to see the difference between browsers there, though. IE is unusable with ~420 MiB of memory for that tab. Chrome is unusable as long as the page loads; after that only slowed down slightly, with ~540 MiB of memory for that tab. Firefox is snappy, fast and takes only 180 MiB of memory for that tab.
> the new engine began as a fork of MSHTML.dll but has
> since diverged very quickly
Curiously, this means that Spartan will still be able to trace its lineage back to 1992's NCSA Mosiac (in contrast to Mozilla's Servo, which is a greenfield project). I expect that it will be fascinating to compare the two projects as each matures.
To be pedantic: even if there is code from MSHTML.dll in Spartan, there's no code from NCSA Mosaic in MSHTML.dll. Internet Explorer originally licensed Spyglass Mosaic, which licensed the name (and the code) from NCSA, but did not actually use any of the code. Though I suspect that there's not actually any Spyglass Mosaic code left in IE these days.
(Edited, I said that there's no MSHTML.dll code in Spartan. That I do believe exists.)
Up to IE6, the about dialog mentioned Mosaic licensed code. There are UI related glitches (mouse pointer and scroll related code) in IE8 that were never fixed since Mosaic. There is still a lot of UI code and API related stuff from IE3 in IE11.
I feel like the "A new Web rendering engine is born" section covers quite honestly why they have chosen to fork IE. It seems, to me, to be pretty frank and doesn't contain much "doublespeak."
PS - This was originally a reply to a comment which got removed.
The amount of that article which was focused on legacy was very depressing. When it started to go about how it would break some Little League site that was IE specific that got me very unexcited. Microsoft today needs to learn the lesson of Apple in the very late 90s and just say "no".
They can't. All of the success MS has is based on enterprise - Apple are lucky in not having that restriction. At this point, if MS decided to say F everyone it would come back to bite them very hard.
I have a lot of respect to these IE devs for not just quitting and doing something easier, to be honest.
Actually it seems strange now, but back in the 90s Apple (and NeXT too) were very focused on the enterprise. They were just "lucky" to have done badly at it. In fact they actually even made servers: http://en.wikipedia.org/wiki/Apple_Network_Server
Apple didn't have a huge market to lose. Or to sustain through several decades of compatibility.
Also, Microsoft makes the vast bulk of its income (90-95%) from partners. It can't just decide to change everything, it has to move thousands of partners along.
Is it just me or is there some kind of an air of sadness about the video?
I like that they have real developers and not pr people talking about the product, and I don't like fake positivity. I guess it's impossible to please people like me, if you're Microsoft.
I'm just amazed that they had to figure out that the "long tail" matters. It seems so obvious. It is almost like the owner of a hardware store coming to the conclusion that his customers want more than just a swiss army tool.
@IEDevChat on twitter is now (and for the next two hours) answering questions regarding Internet Explorer and Project Spartan. Just ask your question(s) with the hashtag #AskIE.
I understand the language. I don't understand what they mean by "patterns". Common edge cases where js/css/html is broken in specific ways?
If you write html/css in a relatively compliant way, things will not be so broken that a browser vendor needs to account for them in a "long tail". If your js is broken, well tough luck.
Are they just talking about how to implicitly handle commonly malformed HTML, as all browsers do?
If people in the past have written bad js which worked in IE, Microsoft aren't now going to break that code. They don't consider saying "tough luck" to businesses with code which works for them today acceptable.
my problem is they're selling this as a "clean break" with no baggage. but then in the same breath explain how they've found a way to still accommodate that baggage and calling it a "long tail".
IE is still gonna live next door (to be used for enterprise/compat) probably for a while. I don't understand what they are saying about solving incompatibility issues of IE-specific code in Spartan - which is presumably NOT IE.
Basically, they took Trident, refactored it and stripped out all the compatibility shims, and made it into EdgeHTML, which they will be working on making similar to WebKit/Blink/Gecko in terms of how it renders web pages. For pages that still need IE, you can switch to Trident, which will still be around.
When MS decides to release an open source rendering engine - we can start taking them seriously... until then they're never going to be worth taking seriously.
For add-ons and DevTools plugins: it would be cool to have AtomIO style architecture. Easy to install, easy to develop for, easy to publish to npm, and written in pure JavaScript. IMHO it would quickly propel Spartan to the forefront of the developer community and would really help move the web forward.[1]
Now if they could just force this new web browser onto every single Windows XP and later installation out there, so we could get ride of the horrendous amount of IE version < 11 still in use, I would see this in a positive light. The only way I am using IE at the moment if for corporate website that are engulfed in MSFT technologies.
If Microsoft tried that, it would probably end up back in court. Microsoft actually has very little power over its large customers, or those customers wouldn't still be using XP, launched in 2001.
If Microsoft could wave a magic wand then everybody would be on IE11, but it can't. In fact, it can't even make them patch security holes that were fixed years ago.
we felt it was important to counter movement towards a monoculture on the Web
Am I alone in wishing there was an actual monoculture on the web? Where you could spend time making your app/site more functional instead of more compatible?
As devs we don't necessarily want a single browser to be dominant even though we might jokingly pray for such a thing during testing phases. We just want them to be standards compliant, right?
Don't get me wrong: there is a certain freeing joy when you work on a project where you know the target browser is the only one you need to think about (like say if you build kiosk apps or apps on a strict intranet).
But at the same time, if one player is too dominant or we only have a single voice about "this is what a browser should be" we get the crap we have lately with Apple's voice being too strong on what is or is not accepted into the HTML standard; or when IE had the 90%+ share it used to have.
Well, let's flip the question around -- why are there three major open-source rendering engines in the world right now already? Why did Google fork Blink from WebKit?
The answer seems to be, major browser vendors have a hard time cooperating at the engine level. Apple and Google had different opinions over WebKit, which eventually led to it being forked. Microsoft has no real reason to think that it could partner with Apple or Google any more successfully than they partnered with each other in developing a rendering engine.
I'm a developer with Azure. Most of the developers still have individual offices (a few folks are also 'doubled up' in individual offices because of space constraints). Only some teams have moved to an open-space office layout.
This. Disbanding the IE team after "winning" the first browser war was a terrible mistake. IE fell behind technically, but they caught up again in recent years. However their reputation among web developers and even end users hasn't really recovered.
As an aside, a great way to anger an IE engineer is to remind them that the entire reason their team exists is because of Mozilla and Firefox. I truly believe that had Firefox not come out, the web browser winter would have continued until the rise of smart phones.
I don't quite follow your reasoning here. It's got a really bad reputation with devs and people "in the know," but for the average user turning on the computer, it's perfectly fine, and that has to count for something.
That's the part I find interesting - despite being good enough for all the major sites on the internet, they kept bumping up against cases where it didn't work. But I don't understand why they needed a new browser to fix that; the article implies that all it required was a change in focus.
Exactly, this is the thing people keep forgetting. IE has a market for people who don't change the default browser from their computers, and it's a good thing they are getting a better, modern browser (not that IE11 is bad, mind you).
If you think of the Browser-as-Operating-System, then building a legacy-free IE makes a lot of sense. That kind of OS development is right in Microsoft's wheelhouse and they need to do it to stay competitive if they want to continue to be a client-side application host.
If this Spartan team does a good job, then they'll brain drain the rest of Microsoft the way XBox did. Building a web browser from the ground-up in 2015 would be a pretty sweet project to work on. I'm nerd-salivating.
That's true. I so wanted to like the Windows Phone. I bought a Lumia 1520 and it was a beautiful piece of tech. And I like the default interface of the home screen (Metro works on a phone, not on desktop imho).
But .. then I tried to use the store to get apps and it was a desert. Or the things I could find, worked poorly.
That's the thing, it might well be on "mum's (next) tablet" (probably not phone).
There's a raft of really cheap (7/8/10 inch) Windows 8.1 tablets that are price competitive with their Android equivalents. At the moment they are running full Windows with an Intel Atom processor. From what I've read, these will all be capable of being upgraded to Windows 10 when release.
Spartan is going to be released for both Windows Phone 10 (or whatever they are calling it) in addition to Windows 10, from what I've read.
Please, let's not pretend remote.modern.ie OR the VM's are ANYTHING close to running the browser locally. They are, at best, a band-aid. Yes, pjmlp, safari is OS X-only now but just because I'm a Mac user does not mean I support that decision, I don't. Also you know how many times in the past 10 years I've heard someone say "Well it works in Chrome/FF/IE but not in Safari?" Never. That's not to say it never happens but I've spent many agonising hours trying to make IE behave when it works PERFECTLY in Chrome/FF/Safari. I've had a handful of "It works in FF but not Chrome" and "It works in Chrome but not FF" but I've got a landfill full of "It doesn't work in IE".
Things I need before I'm going to pretend IE isn't a steaming pile of shit, hell I'd settle for just 1-2:
* Evergreen browser (I'm sick of dealing with 7/8/9, I don't look forward to playing the same song and dance with 10/11/12/etc)
* Cross-platform (Testing on Linux or a Mac is painful)
* Open source (IE is the ONLY major browser that is not OS, and yes I know Chromium != Chrome but I'd be fine with IE doing the same sort of thing)
From the instructions it appears it works with a current RDP client (likely the version required for connecting to Windows 8 or 2012R2)
Interesting, as their partner for Modern.IE, BrowserStack, uses a Javascript based remote desktop. (It may also support Flash on certain browsers, but I'm not sure about that part.)
> This new rendering engine was designed with Project Spartan in mind, but will also be available in Internet Explorer on Windows 10 for enterprises and other customers who require legacy extensibility support.
This is a good point. While I have my doubts that it will be an entirely new browser, by simply renaming it they might be able to offer a version for non-Windows OS-es without making earlier statements seem ridiculous. And they should.
The actual court case you're referring to has to do with the bundling of IE with Windows and how it can't be pulled out. Remember that IE actually used to be cross-platform until version 5 with IE for Mac although it used a different rendering engine (Tasman vs Trident).
I know this is this is the narrative they are trying to promote but I remove code from my projects all the time but don't feel it makes it a different project.
Spartan isn't a subset of IE either. We've added support for nearly 50 standards, and more. The delta between IE and Spartan is massive, and growing quickly.
Interesting, if it's a complete fork does that mean IE is not going to be maintained? I thought IE was going to continue on for the time being. If there isn't a common codebase does that mean you're going to be duplicating work? Why?
Internet Explorer will stick around for legacy applications, but Project Spartan uses EdgeHTML (a fork of MSHTML) and is meant to be used with the web moving forward.
“Microsoft stated that the merging of Microsoft Windows and Internet Explorer was the result of innovation and competition, that the two were now the same product and were inextricably linked together and that consumers were now getting all the benefits of IE for free. Those who opposed Microsoft's position countered that the browser was still a distinct and separate product which did not need to be tied to the operating system, since a separate version of Internet Explorer was available for Mac OS.”
This isn't legacy IE but a new engine. It's like saying Chrome can't run on Linux because MS said IE can't be separated from Windows. Nonsense, but I get your point ("MS lied back then").
Hahaha. Microsoft and Internet are fundamentally incompatible. All you need to do is go to their homepage to see they don't get the Internet. Never have, never will.
The main thing I would want from a Microsoft browser is actually keeping up to date with the latest in security protocols and crypto algorithms. IE has always been several years behind everyone else in this area, which just meant developers had to keep their websites less secure just to be able to serve IE users as well. That needs to end.
I don't necessarily want Microsoft to be a leader in this area (not sure I'd trust Microsoft with any new security protocols anyway), but at the very least it should be a fast-follower. Looking forward to Microsoft adopting ChaCha20-Poly1305 or at least helping to speed up the standardization at IETF and then adopt that standard if their excuse right now is that they can't adopt "non-finalized standards".
Here's an idea for an architecture that doesn't suffer from compatibility problems ever:
- Every web site refers to the bytecode for its required render engine
- This engine is loaded in an intelligent/cached way, and run in a sandboxed environment whenever the website is visited
- Since the website picks its own render engine (or provides its own), the developers of the website know for sure that it will render the website correctly
- Besides the render engine, also the scripting language could be referenced/provided in the same way.
An architecture like this could boost the proliferation of open-source render engines and in-browser languages.
This approach doesn't need to depend on Java. It would preferably use a minimal, non-garbage-collected VM, or something along the lines of NativeClient (or simply asm.js initially).
"It can't be indexed" happens to hold for a lot of websites that are basically single-page web-apps and which build their contents from within javascript. Fortunately, Google and others are using AI techniques to index pages. No need for special markup, or special structure.
Text-selection can be built into the render-engine of choice. Granted, this is not a guarantee that it will work for all websites, but even today you can turn off text-selection for a website, so you don't have that guarantee now either. Also, AI or OCR could help here. As an added benefit, it would allow one to even select text in an image.
Also note that this doesn't work so well in JS, because JS is not a multi-threaded language and it typically suffers from garbage-collection pauses. This is not good enough for UI work, unfortunately. And also not good enough as the target of compilers.
But, granted, to get this started, JS (or the asm.js subset) could be used initially (and be replaced later).
Errr. Sorry MSFT. The web pretty much does work. So long as you don't use IE.
Seriously, I haven't touched IE in probably four years. OK, here and there when there's no choice for some reason. I could probably count those instances -over four years-- with two hands.
We don't even test on IE any more. It's shit and I have no problem ignoring it. MS needs to understand that standards compliance is important. Don't waste my time and money dealing with your crap just because you were too incompetent or couldn't bother to make your browser standards compliant.
And it is slooooooow.
This coming from someone who's been with MS since day one. They are going to have to walk on water for me (and I am sure others) to give the new browser more than 30 seconds of a chance. That's how much trust they've lost.
Oh, yes, what's up with this business of breaking fonts with a routine OS update. Do any of you folks at MS actually use the operating system you create and maintain? Nobody tested this KB3013455 update? Nobody saw that it is destroying fonts on various systems? And you are going to create a new browser?
Sorry. A bit tired of bullshit. Don't tell me. Show me.
Finally. This is the Microsoft I've been waiting for! No more rolling over and no defeatist talk of adopting WebKit or whatever. Microsoft is going to use its muscle and position to make a truly competitive browser.
We need more competition, and we need the default browser in Windows to be just as good as Chrome and Firefox.
Microsoft, I hope you pull every -ms- vendor flaggin', ring-0 kernel hookin', micro-optimizatin', site-specific D3D driver tweakin' trick in the book as you reach for parity. This is going to be so fun to watch.