[Info-vax] login information capture

[pcoviello at gmail.com]

it's fine as long as they fill it in I'm sure I wont have a problem! :-)

thank you!

On Wednesday, February 19, 2014 11:59:58 AM UTC-5, Stephen Hoffman wrote:
> On 2014-02-19 16:39:20 +0000, pcoviello at gmail.com said:
> 
> 
> 
> > it looks like they could just hit return and not put anything in with 
> 
> > restricted on :-(
> 
> 
> 
> The SYSUAF RESTRICT setting on the username involved has nothing to do 
> 
> with how much text might be input by the user.  If any.  That's 
> 
> completely independent.
> 
> 
> 
> Even if you decide to test the text content or text length or other 
> 
> attributes, there's no reasonable way to differentiate whether the 
> 
> reason string is reasonable and semantically valid, or just random 
> 
> text.  Well, short of performing grammar analysis or such -- simple 
> 
> Bayesian won't get you there -- and that sort of processing is a little 
> 
> much for a typical DCL procedure.  Some reasons will be cryptic.  Some  
> 
> humorous.  Even grammar analysis would have trouble with "fx rx efi ipb 
> 
> vms boot err", which could be perfectly reasonable for some support 
> 
> vendor connecting into the server, for that matter.
> 
> 
> 
> Included mostly for humor value, there's 
> 
> <http://www.commitlogsfromlastnight.com> -- same general problem arises 
> 
> there.
> 
> 
> 
> The READ would need an /ERROR or /END_OF_FILE to specifically process 
> 
> ^Z input, as well.
> 
> 
> 
> If you really want to know who is accessing and why, then you're likely 
> 
> headed for some form of two-factor authentication, and that's not what 
> 
> you had originally asked for.
> 
> 
> 
> 
> 
> 
> 
> -- 
> 
> Pure Personal Opinion | HoffmanLabs LLC