Well, whether they are less secure is subjective.

WPA-PSK does not require a username or password. It just requires you to enter 
the key and get access. That’s handled by the WPA2 process.

Any captive portal requiring a username and password will be managed by either 
the hotspot that displays the captive portal (and as such one would hope the 
security certificate for said portal will be one signed by a commercial CA that 
is trusted by the device OS, because the captive portal pages are 
browser-based), or it will be sent (hopefully securely) to a third party. Under 
*NO* circumstances should *any* student/staff member *ever* enter their 
university/organisation usernames and passwords into a captive portal system 
*unless* they are categorically sure that it is run by their 
university/organisation. The risk to the university/organisation and its 
systems is much too great. In the day and age of ransomware attacks, all it 
takes is *one* username with access to a system, which can be used to break 
into others, to inflict untold damage to the university/organisation and its 
students/employees/both.

However, since universities and other organisations (including large 
enterprises) tend to be paranoid (which is a good thing), they would very much 
not like to have usernames and passwords used on *their* systems handed over to 
anyone. This is why WPA-Enterprise works the way it does. It provides a level 
of trust in the sense that devices will be either manually configured or set up 
through a mobile device management (MDM) solution (like Apple Configurator or 
Microsoft’s InTune) which in turn will ensure that only the right certificate 
is trusted (usually that involves a CA certificate and a set of CN and/or 
subjectAltName entries), or, in the most basic case, a certificate is pinned. 
The latter becomes a problem when a certificate is renewed as the new 
certificate will no longer match the pinned fingerprint, and the trust breaks 
(and no authentication takes place). In the user’s eyes that’s extremely 
suboptimal, although from the security officer of the user’s organisation’s 
perspective, it’s exactly what’s expected.

By ensuring the CA certificate is either one of your own (and you’ve put it on 
the device with your MDM (or something like geteduroam, the eduroam CAT website 
or, on Android, the now virtually obsolete eduroam CAT app), you ensure that 
any server claiming to be your home server better be signed by your CA, or, 
just like a browser would (in strict mode), be rejected for being a liar and a 
fraud. You can use a commercial certificate, sure, but by doing that you trust 
that the commercial CA who signed that certificate doesn’t make a mess (like 
several have in the last 12-18 months).

I hope that explains things a little more from the security/tech aspect.

With Kind Regards

Stefan Paetow
Federated Roaming Technical Specialist

t: +44 (0)1235 822 125
gpg: 0x3FCE5142
xmpp: [email protected]
skype: stefan.paetow.janet
In line with government advice, at Jisc we’re now working from home and our 
offices are currently closed. Read our statement on 
coronavirus<https://www.jisc.ac.uk/about/corporate/coronavirus-statement>.

jisc.ac.uk

Jisc is a registered charity (number 1149740) and a company limited by 
guarantee which is registered in England under Company No. 5747339, VAT No. GB 
197 0632 86. Jisc’s registered office is: One Castlepark, Tower Hill, Bristol, 
BS2 0JA. T 0203 697 5800.


From: radiator <[email protected]> on behalf of "Ullfig, 
Roberto Alfredo" <[email protected]>
Date: Thursday, 9 September 2021 at 17:41
To: Heikki Vatiainen <[email protected]>, "[email protected]" 
<[email protected]>
Subject: Re: [RADIATOR] Certificate Not Trusted - InCommon?

You're right - there is no username. Just an SSID and credentials. So why do 
these services not require trusting a certificate? Are those services less 
secure? Is the shared password going over in clear text?

---
Roberto Ullfig - [email protected]
Systems Administrator
Enterprise Applications & Services | Technology Solutions
University of Illinois - Chicago
________________________________
From: radiator <[email protected]> on behalf of Heikki 
Vatiainen <[email protected]>
Sent: Thursday, September 9, 2021 10:59 AM
To: [email protected] <[email protected]>
Subject: Re: [RADIATOR] Certificate Not Trusted - InCommon?

On 9.9.2021 18.11, Ullfig, Roberto Alfredo wrote:

> No, I'm referring to WiFi offered at airports, coffee shops, bars, or at
> someone's home etc... You are given a username and password and the
> phone shows the SSID, you just enter the username and password and are
> connected. There is never a window asking you to trust a certificate.

Could be WPA-Personal, also known as WPA-PSK, where a shared secret (Pre
Shared Key) is needed, or Wi-Fi Protected Setup? But I think these don't
require a username. One option is a built-in client that launches a
customised browser to help with captive portal login.

The problem with the above is that Wi-Fi client software in your device
does not know how to authenticate the network. The network demands that
credentials are used but the Wi-Fi client software has nothing to demand
from the network.

Thanks,
Heikki

--
Heikki Vatiainen
OSC, makers of Radiator
Visit radiatorsoftware.com for Radiator AAA server software
_______________________________________________
radiator mailing list
[email protected]
https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Flists.open.com.au%2Fmailman%2Flistinfo%2Fradiator&amp;data=04%7C01%7Crullfig%40uic.edu%7C223082975583478ac9e208d973aaf2f4%7Ce202cd477a564baa99e3e3b71a7c77dd%7C0%7C0%7C637668000789719424%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C2000&amp;sdata=kwMBZeGt00F7dz4i7dD%2FH8Op%2FwfX6VZmzKxDEOqZe28%3D&amp;reserved=0
_______________________________________________
radiator mailing list
[email protected]
https://lists.open.com.au/mailman/listinfo/radiator

Reply via email to