Hacker News new | comments | show | ask | jobs | submit login
Did Apple Just Quietly End Development Of Safari For Windows? (techcrunch.com)
73 points by Achshar 2007 days ago | hide | past | web | favorite | 62 comments

Genuine question: has anyone here actually ever used Safari for Windows, except for browser testing? I've never met anyone.

So this doesn't really surprise me. Looking up the numbers says Safari is used by 0.46% percent on Windows users -- and it doesn't seem to have ever gotten much higher than that. Why waste resources on it?

Safari for Windows was released more than a year before the first Google Chrome beta. Before that, Webkit wasn't available on Windows at all and it was impossible to test websites on Webkit without a Mac. Google Chrome is now an excellent and widespread browser on Windows so there's no point maintaining Safari for Windows anymore.

I have encountered more than one bug that exists in either Chrome or Safari, but not both. If you only test using Chrome on Windows, you still won't really know how it will behave in Safari.

Curious, are these behavior bugs (expected) or rendering bugs (less so)?

A lot of it is from Safari basically being on its own branch that is occasionally synced with the open source WebKit.

While I'm a Mac user and have used Safari every day for years, I gave it a quick try on my work Windows desktop when it first came out. Since the interface was so much slower than Firefox (probably due to all the custom drawing) I didn't bother to stick with it.

Apple has never really pushed Safari for Windows at all. The theory I've read that always made sense to me was that it was created to allow Windows developers to test pages without an iPhone (or later iPad).

But here we are a few years later and the cost of getting a Mac is lower than it was. Apple already has a ton more developers thanks to the iPhone and iPad. And if you just want to see "will my site look good on an iPhone" you can buy an iPod touch for ~$200.

I'm actually a bit surprised it took them this long. It clearly wasn't an important product for them.

> except for browser testing?

This is where dropping Windows support may harm Safari. Yes we can test against webkit via Chrome now so in theory anything that works on the current PC trinity (FF, Chrome, IE9+) should work just fine in Safari, any bugs that get raised by Safari users that can't be reproduced in Chrome are just going to get labelled "NONREPRO" or "WONTFIX". Chrome and Safari are not completely in-step so there could easily be the odd issue that affects one and not the other.

Admittedly this isn't a big problem for Apple users so the above concern isn't a killer: they can use FF or Chrome if Safari won't cooperate (though good luck convincing some users that switching browser is a valid workaround for any given bug).

I've actually used it for almost a year, as a primary browser. One of the reasons was that it allows (allowed?) switching of font rendering engine to either Apple or Microsoft one. It was part of the experiment whether it's possible to live with a Mac with its extremely blurry though shape-preserving rendering.

Turns out they've bet on ultrahigh resolution displays, where their rendering is perfect.

The other way to go (which I've noticed most Mac people do go with) is comparatively larger fonts. As long as you are above 12px, the Apple font rendering not only looks nicer, but I seem to be able to read it faster.

Of course, I want 12px or smaller unless a display is really high DPI... ;-)

Have you ever considered using GDIPP or Mactype? I used both, GDIPP doesn't work with Chrome so i switched to Mactype a few months back. Everything looks great, but in some situations, certain letters don't show up (like the lowercase "l") which happens very rarely.

I was using it before Chrome arrived. But the font rendering was jarringly out of place in a Windows context.

Only on a netbook. On a couple of Win7 loaded netbooks, safari seemed to use less memory and run noticeably faster than chrome, firefox, and yes IE.

I'd use it as a primary on those systems except for the fact that some of my essential extensions don't work on the windows version.

I install it when Apple says they've made it the fastest browser ever. Then I run the benchmarks and see that it isn't and uninstall it.

My impression was that one of the primary drive of having Safari on Windows back in the day was to get more users exposed to WebKit and to gives developers on windows a way to test there sites. With Google Chrome on mac/windows/linux that seems to be a much more solved problem, not to mention all those old "windows devs" have an iphone now.

I don't know about you, but testing on Chrome and assuming it will also work on Safari is far from "good enough" for me.

On the other hand, testing on Safari/Windows did not ensure it'd work flawlessly in Safari/OSX (still a whole lot better than taking Chrome as a good enough approximation though)

So true. Ran into a few Mac Safari issues that weren't visible in Windows.

I now run a Mac VM to test, which has the added benefit of an iOS emulator.

I've never been a Safari user, on any platform.

Can you tell me if Safari on Windows was a reasonable way to test for Safari compatibility - or was it too far off from the 'real thing' so that you really test 'Safari on Windows or devices that weren't turned on for years' if you run it on a Windows based dev machine?

To give you an idea:

Safari on Windows was pretty often criticized for having ported over the entire OS X font rendering engine (so it looked pretty wildly out of place on Windows), but that was part of what made it a fairly accurate rendition of what it looked like under OS X.

There were still discrepancies even then, sadly.

That font rendering engine was the main reason I used Safari for Windows for so long. I absolutely love it.

I often wonder if Apple knew Google were working on their owner WebKit based browser? Safari was pretty pointless with the release of Chrome IMHO. It was never a very exciting browser.

Your question about the importance of developing Safari for Windows given that Chrome was in the works is interesting for this discussion, but your comments afterward seem out of place since they are about Safari as a whole.

I don't see how the existence of Chrome has any relevance to the creation of Safari as a whole. First, the timing might be off (I'll leave this for others to research.) More importantly, although I don't know exactly what the point of Safari was (presumably Apple wanted to have their own browser to include with the OS just like Microsoft did), I would have to think that if Chrome would satisfy that need then the latest Mozilla browser would have as well.

Apparently, Google, Apple, Nokia, and RIM all submit patches to the Webkit project. Google's contributions have surpassed Apple's in recent years..

Edit: I did a little more digging, and it looks like Chrome did start out as a (not very) secret project for a few years. I'm not sure when they started submitting their own patches to Webkit-core (as required by LGPL?), but they did maintain their own fork of webkit until 2009.

They're not required to submit patches, only release code to people you've distributed binaries to. So before Chrome was released, they didn't have to release anything.

> So before Chrome was released, they didn't have to release anything

They did. While LGPL doesn't mean you need to release all your project's code to users if they request it, it does mean you need to release any modifications of the parts covered by the LGPL.

They didn't have to release the changes in any way that made them easy to reintegrate with the mainline version, but they did have to release them somehow if requested to do so by a Chrome user.

Even the full GPL doesn't require you to automatically release your code back to the original source, it only states that you must provide the source to the people you distribute your product too if they request that you do.

Safari for Windows came out in summer of 2007, Google Chrome came out in September of 2008. Granted Chrome development started in 2006ish which was probably around the time Safari for Windows started. By the time Safari for windows was released those in the WebKit community knew Google was up to something, but didn't know a timetable, any details etc and from discussions they tossed out the first version so when Safari for Windows was released perhaps Google Chrome as we know it today was just starting. Looking at it from a business perspective or even just a 2006 perspective Apple's move makes a lot of sense.

Edit: Some of my memories are a bit fuzzy about what was going on exactly when in relation to everything else so feel free to correct any mistakes in the above.

It would seem they also dropped Snow Leopard support. Not entirely unexpected, but potentially bothersome nonetheless.

Since Snow Leopard was the last version of OSX to run PowerPC binaries, this may actually be quite annoying for all the people who have to be able to run their older apps and have not updated to 10.7 because of that.

Meh. Safari's only utility on OS X is to install Chrome. IE fills that role on the Windows systems.

Aside from all the people (like me) that like Safari. I use Chrome too, but I do genuinely like Safari's UI more than Chrome.

Tell me about it, I still use Camino as my main browser because I find Chrome's UI dreadful and I find Safari's hiding habits annoying when I routinely load 100+ tabs.

The hiding habit used to bug me, but then I started splitting windows of tabs out into logical groups and it bugged me less. It's still the worst aspect of Safari though.

What do you mean by "hiding habits"?

After some number of tabs are opened (depending on window width) the rest aren't accessible as tabs, but from a drop down list of all the tabs.

Different strokes for different folks. I don't like Chrome and am perfectly well served by Safari.

Now that safari has unified the search and url bar, I'm copacetic with either.

or firefox.

It seems as if there are still WebKit Nightlies being built for Windows, it would have been nice if they could tag which ones correspond to certain Safari releases for any Windows devs who want to test their sites in Safari.

I get the impression they already started doing that, when Safari had to work in tandem with WebKit2WebProcess.exe (if I recall the right process), which is pretty good at crashing all my tabs.

Pretty sure Safari from one year ago worked better on Windows than it does now. :/

On the positive side, maybe Apple won't try to force-feed users the browser, when they install and update iTunes, anymore.

Meta-comment: why do tech blogs punch-up headlines with words like "just" and questionmarks?

Because it gets them pageviews.

Because they have no sources or comment from the company, and the entire article is speculation based on poking around the apple website

Wasn't the original point of Safari for Windows that all iOS apps were going to be in‐browser since Apple didn't want to allow the level of access for apps, so it was to allow more people to develop for the iPhone?

And then they went and made the app store and all that went out the window.

I'm surprised they didn't kill it a couple years back.

Apple is a company of extreme focus so this is to be expected. After all, Safari on Windows wasn't going anywhere and it wasn't a substitute for Safari on the Mac.

Hopefully, they'll refocus any Windows Safari people on Webgl.

Man I really liked pointing out that IE was the only major browser that didn't work on every major OS. Apple just lost what little high ground they had left in my book.

I always see Safari as a small GUI around vanilla WebKit -- i.e. Safari basically is WebKit. (Chrome on the other side has pushed it much further and WebKit seems to be only a part of it.)

From the history of WebKit, I also got the feeling that Apple did one of the most important work to modernize it to compete with IE, Firefox, Opera at the time it was released. Afaik, KHTML was not really as far.

Under this view, I don't really see your point. This work by Apple (WebKit/Safari) and by KDE is now part/base of one of the most important browsers, Chrome.

Really though Apple and Microsoft need to both just accept that they lost the browser war and ship with Chrome or Firefox.

How does that even make sense? Apple was never even in a "browser war" to begin with, but even if you want to claim they've "lost" some sort of war, how does that lead to discontinuing their browsers? Besides the fact that Apple created WebKit and ship it as a core component of OS X, how does reducing browser choice benefit anyone?

Apple created "WebKit", but before it was WebKit it was KHTML. Let's not lose sight of history here.

And KHTML was based on the KDE HTML Widget. There's always prior work. But that's not particularly relevant to my point that it makes absolutely zero sense for Apple to discontinue their browser.

Nonless you are still wrong to say that "Apple created WebKit" as the WebKit source code was a fork from the KTHML source code.

Really? Without Apple, "WebKit" would literally not exist. Apple created the project. The code itself was a fork of KHTML, but I'm not talking about the code, I'm talking about the project.

Created is a very strong word to toss around especially in this context. Perhaps you were only referring to the trademark WebKit which Apple did register (but they did offered to have the community manage it at the start if I recall). WebKit literally wouldn't exists without the KHTML developers who were in discussions with Apple and the result of which was the WebKit project. Here on HN where code is first class and many others have incorrectly stated that Apple coded/created/built all of WebKit and they knew nothing of KHTML it is really no surprise that someone would try to correct you when you stated the "Apple created WebKit". Creating something isn't the guy that happen to register the trademark, it is the guys that wrote the code in this context.

Who cares about the trademark? The WebKit project is what I'm talking about. KHTML would not have gone on to become an incredibly popular cross-platform, cross-device, cross-company framework for web rendering if Apple hadn't started the WebKit project. It would have remained nothing more than a web rendering engine for KDE.

Code may be important, but it's not the only thing that matters. You could write the most beautiful, efficient, useful code in the world and it wouldn't matter if nobody ever saw it.

Besides, this is straying further and further from the point that Apple ships and maintains a web rendering engine as a core component of the system. It's so firmly embedded that it actually powers most text rendering on iOS. In light of this, saying Apple should discontinue their browser which is built upon this engine is preposterous.

If by "created" you mean "enhanced and renamed".

By that standard UC Berkeley "created" BSD and Tim Berners-Lee "created" HTML.

Safari on OS X is excellent. It was complete garbage on windows, but safari 4+ on OS X is an ideal browser IMO.

Looks like I will have to start using IE then

I used Safari for Windows when it first came out, it was pretty good actually, but the later versions were a disaster, I don't know how but it ran like hell.

Somehow Apple managed to make Safari 5 slower than previous versions on Windows.

I can't exaggerate how unstable it was, it crashed all times and for no reason.

While I imagine next to no one uses Safari as their actual browser on Windows, it was nice that it was there so most cross browser testing could be done on one box.

Does anyone really care?

breaking news!


Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | DMCA | Apply to YC | Contact