Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Apple Releases Safari 3.2 with Improved Security, Anti-Phishing
- Thread starter PeteB
- Start date
- Sort by reaction score
same here. i have uninstalled all my 3rd part add-ons but safari crashes when i open a new tab.
You're right. We should only use Apple-approved software from now on. iLife, iWork, and that's it. Forget about any third-party apps at all.
God you're a bit rude ain't ya. So you don't install ad blockers in Safari or other plugins that make Safari better and add features that Safari is lacking?
Third party apps are fundamentally different from things like Saft. Safari "addons" are modifying Safari without its knowledge in ways that make assumptions about how it works. In many cases it would literally be impossible to make any improvements to certain areas without breaking compatibility with certain addons.
So take your pick: addons keep working between versions, or Safari improves.
The Firefox developers are in somewhat the same situation, which is why Firefox 3 broke most extensions and the authors had to scramble to fix them. Indications so far are that Firefox 3.1 is going to break a bunch of them again (see, for example, http://quotes.burntelectrons.org/4095).
<edit>
Adding a semi-hypothetical example (hypothetical because I don't know that any addon currently does this, semi- because I actually did make this change):
In trunk WebKit, I made some of the node traversal functions into inline functions for increased performance, and moved some infrequently used data to an optional storage spot. The upside is that this was a nice perf boost on a number of DOM benchmarks, and a few megs less ram used on complex sites. *However*, if an addon were making calls directly into the WebCore versions of the functions, it would now crash since they're inlined and no longer exist in the vtable for that class.
</edit>
As far as I know that's correct. I didn't pay attention to exactly which commits were merged, but I'm pretty sure SF/SFX wasn't among them.
No, but the Webkit nightlies do and, because they also hook into the underlying technologies in OS X that were installed along with Safari 3.2, the new anti-phishing features works too.
This just in from sunspider:
TEST COMPARISON FROM Webkit(38386) TO Safari 3.2 DETAILS
==========================================================
** TOTAL **: 3.91x as fast 2832.2ms +/- 0.4% 724.0ms +/- 1.6% significant
==========================================================
3d: 2.66x as fast 346.6ms +/- 1.3% 130.2ms +/- 4.9% significant
cube: 2.32x as fast 117.4ms +/- 2.2% 50.6ms +/- 13.0% significant
morph: 2.45x as fast 117.4ms +/- 2.2% 48.0ms +/- 3.7% significant
raytrace: 3.54x as fast 111.8ms +/- 0.5% 31.6ms +/- 2.2% significant
access: 6.74x as fast 445.0ms +/- 0.6% 66.0ms +/- 1.9% significant
binary-trees: 6.43x as fast 60.4ms +/- 2.3% 9.4ms +/- 11.8% significant
fannkuch: 15.0x as fast 209.4ms +/- 1.3% 14.0ms +/- 0.0% significant
nbody: 3.42x as fast 124.6ms +/- 1.1% 36.4ms +/- 1.9% significant
nsieve: 8.16x as fast 50.6ms +/- 1.3% 6.2ms +/- 9.0% significant
bitops: 10.8x as fast 384.2ms +/- 1.8% 35.6ms +/- 4.0% significant
3bit-bits-in-byte: 18.6x as fast 55.8ms +/- 1.0% 3.0ms +/- 0.0% significant
bits-in-byte: 12.7x as fast 86.4ms +/- 5.5% 6.8ms +/- 8.2% significant
bitwise-and: 17.8x as fast 149.2ms +/- 2.2% 8.4ms +/- 8.1% significant
nsieve-bits: 5.33x as fast 92.8ms +/- 0.6% 17.4ms +/- 3.9% significant
controlflow: 18.5x as fast 74.0ms +/- 0.0% 4.0ms +/- 0.0% significant
recursive: 18.5x as fast 74.0ms +/- 0.0% 4.0ms +/- 0.0% significant
crypto: 4.83x as fast 201.8ms +/- 0.5% 41.8ms +/- 1.3% significant
aes: 5.20x as fast 66.6ms +/- 1.0% 12.8ms +/- 4.3% significant
md5: 4.79x as fast 67.0ms +/- 0.0% 14.0ms +/- 0.0% significant
sha1: 4.55x as fast 68.2ms +/- 0.8% 15.0ms +/- 0.0% significant
date: 3.52x as fast 253.6ms +/- 0.3% 72.0ms +/- 1.2% significant
format-tofte: 3.87x as fast 113.8ms +/- 0.9% 29.4ms +/- 3.8% significant
format-xparb: 3.28x as fast 139.8ms +/- 0.4% 42.6ms +/- 1.6% significant
math: 4.13x as fast 392.8ms +/- 1.8% 95.2ms +/- 2.1% significant
cordic: 4.45x as fast 153.8ms +/- 2.8% 34.6ms +/- 4.1% significant
partial-sums: 3.69x as fast 169.0ms +/- 1.6% 45.8ms +/- 1.2% significant
spectral-norm: 4.73x as fast 70.0ms +/- 1.8% 14.8ms +/- 3.8% significant
regexp: 4.40x as fast 172.4ms +/- 0.4% 39.2ms +/- 3.5% significant
dna: 4.40x as fast 172.4ms +/- 0.4% 39.2ms +/- 3.5% significant
string: 2.34x as fast 561.8ms +/- 0.8% 240.0ms +/- 1.6% significant
base64: 4.39x as fast 86.0ms +/- 1.0% 19.6ms +/- 5.7% significant
fasta: 4.63x as fast 150.0ms +/- 1.7% 32.4ms +/- 4.4% significant
tagcloud: 1.30x as fast 113.0ms +/- 0.8% 87.0ms +/- 1.0% significant
unpack-code: 2.15x as fast 112.0ms +/- 1.1% 52.0ms +/- 1.7% significant
validate-input: 2.06x as fast 100.8ms +/- 2.5% 49.0ms +/- 2.5% significant
My plug-ins are still a little unstable, even after a re-install and cache-clearing session. If it bugs me too much, I'll seek alternatives like inquisitor and Safari Adblock. But, for now...
As long as you still have the security lock icon (and a https url), you are safe.
The "green text" is a higher up security certificate. Basically, the company in question (such as eBay) handed over more $$$ so a web browser can display who actually owns the web site where the user can see it.
So for example, no one will be able to have the green text say "eBay, Inc." without it actually being ebay.com.
It is still relatively new, so many companies are still upgrading their certificates. Bank of America did theirs a month or so ago for example.
From that website Safari opened all the links I tried, Opera however bought up the fraudulent warning everytime.
However after closing Safari and then reopening it the links were shown as fraudulent in Safari
For those who said it didn't work have you tried restarting Safari.
It is a step in the right direction for Safari, was it the last major browser to get this Opera, IE7, FF and Safari ?
Delay changing to ssl-tab
I get a noticeable delay when i change to a tab that contains a secure site. Anyone else getting this? It's annoying
I get a noticeable delay when i change to a tab that contains a secure site. Anyone else getting this? It's annoying
Still waiting for the REAL big update FF 3.1 beta 2
So with FF 3.1 beta 2 Firefox will finally have private browsing and yes it has it's own 'snappier' JS engine. It's nice see the world standardizing around 2 layout engines, Gecko and WebKit, that are obsessed with supporting standards.
So with FF 3.1 beta 2 Firefox will finally have private browsing and yes it has it's own 'snappier' JS engine. It's nice see the world standardizing around 2 layout engines, Gecko and WebKit, that are obsessed with supporting standards.
Is it just me (probably is) but this seems to improve speed visiting MobileMe.
Trying both from my macbook and XP station with safari 3.2 installed and it seems somewhat quicker than firefox and previous versions of Safari
It also seems alot more stable and faster when uploading to idisk too . . .
Trying both from my macbook and XP station with safari 3.2 installed and it seems somewhat quicker than firefox and previous versions of Safari
It also seems alot more stable and faster when uploading to idisk too . . .
While we wait for my 3.2 to finish downloading, here is how mine looks with 3.1.2.
Attachments
oh no! are you serious? i just installed google video chat yesterday, tried it out with a friend and was SOOO impressed! i hope yours is an isolated case, because i would be extremely disappointed if apple would sink so low...
Apple provides ways of extending it; these 'extensions' which are causing problems are not using the provided interfaces - they are using undocumented parts of safari that were never designed to be used outside that of Apple.
Stop being a hyperbole queen and stay on planet earth. The issue is using third party components that rely on undocumented features. Features which Apple have NEVER disclosed to developers and thus never promised to maintain compatibility with updates/upgrades.
How on earth did you make it this far.
Learn the difference between extensions which use API"s designed and supported by Apple versus hackware which uses undisclosed and thus unsupported hacks using internal API"s.
Dear god, please - send the computer back. Stick with a type writer because that is where your capability is right now.
KítscheñÇinqµe
macrumors regular
AFAIK, "anti phishing" checks are a (small) WOB if you're already immune to phishing. however, your granma or cosmetologist might need some assistance...
There's a major difference between Firefox and Safari: Firefox is open source and developers don't have to rely on assumptions to make their code work.
Over the last weeks, I became more and more disappointed with Safari. Too many websites don't really work with it. Try selling something on eBay with Safari, for example, or homebanking with certain banks. Several corporate pages that I had to use professionally also did not work with Safari.
I had to learn that except for a few die-hard-Internet Explorer-only pages, almost everything works fine and faster with Firefox.
That's not really making me happy, because for some irrational reasons I actually prefer using Safari. It's probably that "underdog-thing" that appeals to me. Unfortunately, Safari does not live up to Apple's marketing promises and fails too often in real life.