[9fans] I'm puzzled

[9fans] I'm puzzled

[ron minnich]

I just got this and I don't know why. Never seen it. I thought the key
I set up was never expire.

cpu% auth/secstore -g factotum
secstore password:
auth/secstore: error: account expired

I'm too dumb to figure the way out. Sorry.

So, how do you dig yourself out of this hole? I can just fix the code
to disable that test, surely
there is something better to do here.

This is also another "useless message" example that drives part time
guys like me nuts. Would it really absolutely run counter to all Plan
9 principles, and violate peoples lives, fortunes, and sacred honor,
to PRINT THE FRIGGING DATES? It would at least tell me, if the dates
were totally nuts, that there was corruption going on here. This
message contains (almost) not useful data.

Just wondering.

thanks

ron

Re: [9fans] I'm puzzled

[geoff]

I've just pushed out a new auth/secstore/password.c that has better
error messages and only believes non-zero expiry times.

Re: [9fans] I'm puzzled

[andrey mirtchovski]

how does that deal with the "never" expiring accounts, don't they have
zero as the expiry date?

On 5/28/07, geoff@plan9.bell-labs.com <geoff@plan9.bell-labs.com> wrote:
> I've just pushed out a new auth/secstore/password.c that has better
> error messages and only believes non-zero expiry times.
>
>

Re: [9fans] I'm puzzled

[ron minnich]

On 5/28/07, andrey mirtchovski <mirtchovski@gmail.com> wrote:
> how does that deal with the "never" expiring accounts, don't they have
> zero as the expiry date?
>
> On 5/28/07, geoff@plan9.bell-labs.com <geoff@plan9.bell-labs.com> wrote:
> > I've just pushed out a new auth/secstore/password.c that has better
> > error messages and only believes non-zero expiry times.
> >
> >
>
BTW, the auth/enable did not work at all.

I am running, right now, with a hacked secstored (well, password.c is
hacked with an 'expired, like I care' print to fd 2) that ignores the
expired error. I have no idea what's wrong here. I just know how to
work around it.

thanks

ron

Re: [9fans] I'm puzzled

[Russ Cox]

> This is also another "useless message" example that drives part time
> guys like me nuts. Would it really absolutely run counter to all Plan
> 9 principles, and violate peoples lives, fortunes, and sacred honor,
> to PRINT THE FRIGGING DATES? It would at least tell me, if the dates
> were totally nuts, that there was corruption going on here. This
> message contains (almost) not useful data.

for what it's worth, most of the error messages do
try to be helpful now.  it was harder to be helpful
when the errors were limited to 28 bytes in old 9p,
and the auth server code has not been updated 
since they got bigger.  at least you didn't get EACCES
or something like that.

russ

Re: [9fans] I'm puzzled

[ron minnich]

On 5/29/07, Russ Cox <rsc@swtch.com> wrote:
>at least you didn't get EACCES
> or something like that.

yep, you're right.

Anyways, I'm closer, I found that once I had a working factotum on the
node -- i.e. I got the rsa key into the factotum so that tls would
work -- (this is in essence a combined auth/fs/cpu node) then mail
passwords stopped working.

Closer. Will update wiki when I figure this all out.

thanks

ron