Hi Tony, 

> On Apr 28, 2025, at 7:29 PM, Tony Przygienda <[email protected]> wrote:
> 
> 
> 
> On Mon, Apr 28, 2025 at 7:48 PM Acee Lindem <[email protected]> wrote:
> Speaking as WG member:
> 
> I support adoption. 
> 
> thanks Acee
>  
> 
> I do think draft terminology could be improved. 
> 
> Despite the risk of confusion with AI-based landscaping, I've resigned myself 
> to you using "pruner". 
> However, for a node that is not preforming any flooding reduction, this 
> should be a "non-pruner" rather than
> a "zero-pruner" or, worse yet, "zero".
> 
> "non-prunner" I think is OK a term we could go with if people think that 
> makes it more readable than "zero prunner". 
> 
> We can even move away from "pruner" for this class of algorithms if someone 
> has a significantly better term that crosses their minds. 
>   
> Additionally, I think the term "connected set" for a group of routers running 
> the same flooding reduction algorithm 
> is a poor choice. You say these are "colloquially often" called "flooding 
> topologies" which is exactly what they 
> should be called. If we're going to come up a clever term, I'd suggest "Flood 
> Uniform Clique" since the routers
> in one such grouping can reduce their LSP advertisement without regard to 
> what is being done in another such
> grouping. 
> 
> Those are not "clever" terms. Those are terms used for many, many years in 
> graph theory and describing "precisely" what the "thing" has to be to work. 
> Also, clique is a term used in the context of graphs and means something 
> rather different than a "connected component". 

The alternate term wasn't really serious. 

> 
> AFAIS nothing prevents people from calling the stuff "flooding topology" of 
> course as long we have a framework in place making it clear that means a 
> "edge dominating set allowing cycles/loops" rather than having later to 
> discuss in future in context of subtle algorithms what a "flooding topology" 
> really is. Hence the document spells that stuff out. 
> 
> 
> Finally, I don't see the need to optimize the output to the non-ASCII test 
> format when this is not
> the default you get when the draft is displayed in the datatracker - 
> https://datatracker.ietf.org/doc/draft-ietf-lsr-distoptflood/
> 
> I fail to read your point somehow here. You mean an ASCII figure showing this 
> is necessary. Yes, work item to do once draft adopted of course.

I think the embedded markup is more of a problem than the figure since readers 
are more used to complex figures being omitted. 

Thanks,
Acee



> 
> -- tony 
>  
> 
> Thanks,
> Acee
> 
> > On Apr 27, 2025, at 10:39 AM, Acee Lindem <[email protected]> wrote:
> > 
> > LSR WG, 
> > 
> > This starts the Working Group adoption call for 
> > draft-prz-lsr-interop-flood-reduction-architecture-01. Please send your 
> > support or objection to this list before Monday, May 12th, 2025. 
> > 
> > Thanks,
> > Acee
> > 
> 

_______________________________________________
Lsr mailing list -- [email protected]
To unsubscribe send an email to [email protected]

Reply via email to