headers
Joe Touch | 2 Jul 2012 20:16
Picon
Favicon

Re: Should security requirements be MUST?


On 7/2/2012 10:52 AM, Marsh Ray wrote:
> On 07/02/2012 12:34 PM, Joe Touch wrote:
>>>
>>> Seriously, you think Netscape would have shipped a security system that
>>> required you to install a kernel module?
>>
>> They could have shipped with a security system that used an existing
>> security kernel module, just like they shipped with a communication
>> system that used an existing transport kernel module.
>
> I'm sorry dude but that's just ridiculous.
>
> If fact, Netscape shipped a time when kernel TCP/IP modules didn't exist
> on many clients.

Clearly everything Netscape first shipped with hasn't changed in 17 years.

Oh, and why it's still available.

Things change, and there have been many opportunities to use kernel 
modules over the years. "First out" isn't always the winner (or are you 
all using SGML and Fetch?)

Joe
Permalink | Reply |
headers
Theodore Ts'o | 2 Jul 2012 20:24
Picon
Picon
Favicon
Gravatar

Re: Should security requirements be MUST?

On Mon, Jul 02, 2012 at 11:16:21AM -0700, Joe Touch wrote:
> >If fact, Netscape shipped a time when kernel TCP/IP modules didn't exist
> >on many clients.
> 
> Clearly everything Netscape first shipped with hasn't changed in 17 years.
> 
> Oh, and why it's still available.
> 
> Things change, and there have been many opportunities to use kernel
> modules over the years. "First out" isn't always the winner (or are
> you all using SGML and Fetch?)

Sure, but what it means is that SSL had the first mover advantage.  If
IPSEC was going to displace SSL it would have needed to be
significantly better in terms features, functionality, ease-of-use,
ease-of-deployability, etc.  Unfortunately, the lack of standard API,
the unevenness which OS providers adopted IPSEC, etc., meant that in
the eyes of application developers IPSEC was significantly worse than
SSL, not better.

Yes, today IPSEC support is in most OS's.  And in theory someone could
create a simple, easy to use userspace library which applications
could use, and this API could expose enough information on a per-TCP
connection basis so that a server could make authorization decisions.
(The discussion of whether x.509 certificate names is a wash, given
SSL is also using certificates, so it doesn't matter for the purposes
of this discussion.)  But even supposing all of this work got done,
what's the killer feature that would cause people to want to migrate
to using IPSCEC instead of SSL?
(Continue reading)

Permalink | Reply |

Gmane