Try services, network, inetd, configure, http, configure.
--
You received this message because you are subscribed to the Google Groups "Alt-F" group.
To post to this group, send email to al...@googlegroups.com.
To unsubscribe from this group, send email to alt-f+un...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/alt-f?hl=en.
But remember that the root password is transmited in clear form through the wire. Use the stunnel package for secure http.
On Jan 29, 2012 1:09 AM, "Joao Cardoso" <joao.fs...@gmail.com> wrote:Try services, network, inetd, configure, http, configure.
> > On Jan 28, 2012 7:52 PM, "Keon91" <kpsc...@gmail.com> wrote: > > Hi, > > I have forwarded the...
That's the browser job. If it does not complain (apart from the self-signed
server certificate) when the url is 'https:/...', then all is alright. It
should display a lock icon or similar
> How can I be sure that my root password isn't transported clear trough
> the wire?
>
> On Jan 30, 4:45 pm, Keon91 <kpscha...@gmail.com> wrote:
> > Isn't it a good idea to make the stunnel a standard enabled service in
> > Alt-F and a part of the http under the service inetd?
Yes, if there wasn't only 8MB of flash memory available for the kernel and the
root filesystem. The DNS-325, e.g., has 128MB!
Regarding the login page: yes, it is possible to only display the status page
after a successful login, "security by obfuscation". But a "man in the middle"
attacker could get some server info from the http headers.
The auth screnshot you posted is tipically used by the browser when the server
asks for http authentication (look at the swat login, services->network->smb-
>configure->advanced page).
The "basic" http auth is not better than the current forms-based
authentication, and I'm not sure if the busybox http server fully supports
digest-based (md5) auth. If it does, and there are signs in the code that it
does, then Alt-F could use it -- passwords would not be transmitted in the
clear.
But again, a man in the midle attacker could forge the server header, asking
the browser to use "basic" authentication.
I'm not a security expert, and my Alt-F concerns regarding security only
covers the trivial casual user cases.
I think that "what a man can do other can undo" (apart from breaking an egg,
of course :-)
But you can fill an issue report, asking for http digest-based auth, so I will
not forget to see if busybox http server fits the job.