Bug? in OVadd ??

Russ Allbery rra at stanford.edu
Mon Apr 11 22:34:40 UTC 2005


Todd Olson <tco2 at cornell.edu> writes:
> At 11:04 -0700 2005-04-11, Russ Allbery wrote:

>> This change is going to mask other, more serious problems.  I'm not sure
>> that it's a good idea in general.

> Hum ... So I need to recompile without the change and redeploy
> (I'm beyond the Xref stage anyways)

> Thanks for the heads up.

Don't do that yet; let me think about this some more.  I'm still missing
some piece of the overall picture and need to understand it.

> ... but ... a question or two ...

>     a) Heath's comments suggests that the other OVadd methods worked the way
>        that his suggested fixed worked for tradindexed
>        ... so are they broken too?

No, they should be fine.  Making tradindexed consistent with everything
else is reasonable.  That means that everything else wasn't working in the
way that I was expecting, so I need to go dig into the code and understand
the assumptions better.

>     b) It seems to me that the code in question is not particular to Xref
>        slaving (but that my be my very limited understanding of the code).
>        So I am concerned that if I get a message from a peer, posted to a
>        group I have, cross posted to a group that I don't have, that I will
>        start seeing these errors, even in the no slave case.  Can you set
>        my mind at ease on this?

No, INN only puts groups that it carries into the Xref header and overview
information is only stored in groups that it carries.  Xref slaving is a
special case.

-- 
Russ Allbery (rra at stanford.edu)             <http://www.eyrie.org/~eagle/>

    Please send questions to the list rather than mailing me directly.
     <http://www.eyrie.org/~eagle/faqs/questions.html> explains why.


More information about the inn-workers mailing list