[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: NLDM LDM future... 32 types, server side restriction.



Hello Peter,

My apologies for the delay in responding.  Way too many meetings!


Peter Silva wrote:
> 
> Thank you Anne!
> 
> What about bridge type services, one paper mentions "everyone switching
> over at once"... ouch..

Yes, not only a painful thought, but probably infeasible.

> Could there be some kind of glue which would take in a NLDM feed, and
> translate it into LDM 5 or 6.  You would clearly lose some things in the
> process.  One of my clients is using LDM5 (We are feeding unidata with
> ldm6.) and they have commercially acquired visualization systems.  I am
> not sure how easy it is for them to rip out LDM, and replace it
> outright, especially since they (and we) have NOAAport downlinks, and we
> are unlikely to transition that in the near future.   The glue would be
> around for a while.  I imagine dump & reingest is the simple answer,
> just wondering if there is more to it than that.
> 

I modified some LDM queue scanning code to extract products from the LDM
queue, wrap them into a format acceptable to NNTP and post them.  That's
how I'm getting the data with which to test.  It would be easy to write
something to go the opposite direction.  So, glue of that type is
certainly a possibility.  Should this come to pass, I imagine growing
the NNTP network while shrinking the LDM based networks, coupled with
some glue-type pieces like these.

> I found two of your papers on the net, one was the paper which I think
> was for the AMS last week, the other was something un-dated on your web
> site.  I wasn't clear on which conclusions were more recent,  Is using
> INN just a demonstration, with a re-implementation planned for a new
> LDM, or is the intent to add integration around standard NNTP servers
> (of which inn is but one example.)   The web page seems to say the
> former, the AMS paper the latter.   Is that sort of thing decided?  Is
> it likely that LDM v 7 will be some form of NLDM ?
> 

I see I should date those papers...

I seriously doubt we would reimplement INN.  It already does a very good
job, and has the benefit of the attention of a number very knowledgable
maintainers plus interested parties in the open source community.  

My view is that the data relay should work with any NNTP server.  As
part of testing I'm considering trying dnews as well.  However, the
issue of general NNTP compatibility will be visited if/when we decide to
go this route as it's not totally my own decision.


> 
> Your answers were great!  Excitement makes any job easier.
> 
>  Thanks again!
> 

Happy to be of assistance.

Anne

-- 
***************************************************
Anne Wilson                     UCAR Unidata Program            
address@hidden                  P.O. Box 3000
                                  Boulder, CO  80307
----------------------------------------------------
Unidata WWW server       http://www.unidata.ucar.edu/
****************************************************