On Tue, 7 Jul 2020 19:25:46 +0800, Timothy Sipples <[email protected]> wrote:

>Have those "powers that be" offered a list of acceptable alternatives? 
No, of course not.


>Unless they insist, I don't think NTLM over HTTP is a good protocol idea 
I don't think so either.

>nowadays for a variety of reasons, so can we skip that one?
Probably not. See above… :-(


>The IBM HTTP Server for z/OS supports TLS client certificate 
>authentication with RACF. That's not basic authentication, so it 
>ostensibly qualifies. It's also widely accepted. Have you considered that 
>option?
Not yet, because it opens a different can of worms: that of having to manage 
the client certificates. I am not sure I want to do that… But I agree: it would 
be a good alternative.


>Or you could adopt a token-based approach. The classic way is forms-based 
>authentication, i.e. some application-based mechanism.
Hmmmm…. That would mean I need to code to interrogate RACF, with all the 
problems of running authorised (z/OS lingo) code. I don't want to go there 
unless I really, really, *really* have to.

> Another, widely 
>accepted choice is OAuth 2.0. However, OAuth 2.0 would require either a 
>custom, additional module or an authenticating proxy arrangement of some 
>kind.
I don't know that one. But does look as a lot of hassle for something where 
basic authentication over HTTPS is all I really need.

Thanks for the suggestions. At the very least, they give me some ammunition…

Cheers,

Jantje.

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN

Reply via email to