Home page logo

bugtraq logo Bugtraq mailing list archives

RE: RIP: ActiveX controls in Internet Explorer?
From: "Drew Copley" <dcopley () eeye com>
Date: Tue, 2 Sep 2003 11:17:47 -0700

Hash: SHA1

-----Original Message-----
From: sb () xiongmao otago ac nz 
[mailto:sb () xiongmao otago ac nz] On Behalf Of Simon Brady
Sent: Sunday, August 31, 2003 6:43 PM
To: bugtraq () securityfocus com
Subject: Re: RIP: ActiveX controls in Internet Explorer?

On Sat, 30 Aug 2003, Alun Jones wrote:

The descriptions I've heard of this suggest that this 
patent could be 
applied equally to prevent (or grab payment from implementors of) 
Javascript, Java, Flash, etc.

I'm with you on the security issues with ActiveX (and 
Javascript) - I 
disable ActiveX on the principle that it has no security 
consideration, and Javascript on the basis that it's been 
implemented in a vulnerable manner.  But this is a considerably 
further-reaching patent than merely killing off ActiveX.  Before we 
sing "ding dong the witch is dead", let's have some concern for the 
peaceful Wiccans that might be next on the chopping block.

Java and Flash aren't exactly free of security issues either. 
In fact, I 
would go further and argue that the whole notion of a controlled 
client-side runtime environment for remote code has been an 
disaster for the web (and this is solely from a security 
perspective - see 
http://members.optusnet.com.au/~night.owl/morons.html for a 
take on the usability crisis they've caused).

I'm not just referring to current implementations with their 
defect rates. All client-side runtimes, no matter how well-written,  
inherently reduce security. That's their function: to give outsiders 
access to your machine they otherwise wouldn't have.

Even more insidiously, their prevalence numbs users into a 
mode of thought that it's quite normal and healthy to let 
this happen. How can the security community promote safe 
browsing when users are being forever brainwashed into 
ignoring or disabling security features for the sake of 
pointless but pretty downloadable applets? How can we 
encourage content developers to reduce attack surface when 
fashion demands yet more gratuitous bells and whistles?

Web applications belong on the server. The more widely this 
patent gets applied the better off the browsing public will be.

Client side applications have problems. Server side applications have problems. Anything you use has problems. Anything 
could have a security hole in it. Further, when you are talking about "all client side runtimes" -- you may not have 
meant this, but you are talking about the browser itself.

Client side applications may often involved "pointless but pretty" websites, but they are also used in just about all 
ecommerce sites. The pros and cons of these sites people could forever debate, but personally I love being able to buy 
stuff in the middle of the night or from across the planet. Who doesn't? And, where else can you shop for a new car and 
not have to deal with annoying sales people? Where else can you surf for buyer opinions, the best rates, and do your 
shopping -- but on the internet? [I am not even mentioning doing online bills, loans, or countless other such tasks 
which are made easier by these things.]

I will not lie, there is a lot of "attack surface" in a browser. There will only be more and more as the years 
progress, if anything. The browser is our most central and basic connection to the internet. As such it is destined to 
become more, not less.

Client side attacks will continue to be popular, but bugs being found in clients does not mean these clients have 
become less secure. In fact, if you look at applications' bug histories you will note that the bugs generally become 
harder and harder to find, and the time between bug fixes becomes further and further apart. This rule holds generally 
true with all applications.

But, why will client side attacks continue to be popular? There are some very simple reasons. One, everybody has a 
client. Not everybody has a server. Two, client side attacks bypass just about every type firewall people may have. 
Three, client side attacks remain remote attacks, despite the fact that everybody has a client. Four, yes, there is a 
wider attack surface than that which can be found on servers (generally). 

Four good reasons why these attacks will remain popular.

At the end of the day it all comes down to one's acceptance of new technology. Are we demanding that we stay with 
yesterday's technology, or are we able to go forward with new technology? Must we insist websites use 1994 type 
technology? Is this kind of insistence not really based on nostalgia? Is there no danger to nostalgia?

Apologies for all of the posts this week.

Simon Brady                             mailto:simon.brady () otago ac nz
ITS Technical Services
University of Otago, Dunedin, New Zealand

    I don't speak for my employer, and they don't speak for me.

Version: PGP 8.0


  By Date           By Thread  

Current thread:
[ Nmap | Sec Tools | Mailing Lists | Site News | About/Contact | Advertising | Privacy ]