[HECnet] Configuring py-decnet --> patches

[Johnny Billquist]

On 2020-08-26 22:50, Thomas DeBellis wrote:
> Only some of them don't show up?  Boy, that sure would drive me nutz!

Yup. I haven't been able to spot a pattern, but only turned up when it 
got above a certain number of entries. But then it seemed to disappear 
again, and now I've come to a point where it reappeared again.

And it's about 8 nodes that don't show up in about 800. So about 1%.

> I have a batch job which fires up weekly and pulls down FIX.T20 from 
> MIM.  If a connection can't be made, then it whines and tries again 
> after a while.  The file is downloaded and changes are documented with a 
> file comparison (FILCOM).  It is then converted into an internal format 
> by SETNOD and loaded into the monitor

Well, at least the files can be trusted.

> A planned future enhancement is to compute the set difference; that is, 
> remove nodes from the monitor database which are no longer defined.  I 
> had stumbled over what appears to be undocumented way to do this while 
> implementing other functionality; but it needs application logic to do 
> the set difference.

That is one thing that is on my mind somewhere in the future as well. 
It's not handled that nicely today, and require some manual fiddling 
when it happens.

   Johnny

-- 
Johnny Billquist                  || "I'm on a bus
                                   ||  on a psychedelic trip
email: bqt at softjar.se             ||  Reading murder books
pdp is alive!                     ||  tryin' to stay hip" - B. Idol