Re: flow id's and .tr files

[Tarik Alj <aljtarik@ozias.inrs-telecom.uquebec.ca>]

> If there's a Good Reason for how all this is done other than leftover
> code from iph->flowid()/p->class_ conversion, I'd really like to
> know what it is.
> 
> As far as I can tell, if anything should break it should be class_,
> going on Kevin's comment (in up to v1.59), and fid_ ought to work
> these days.
> 
> L.
> 
> look, can we stick to binding variables of the same name? Once? Please?
> 

I tend to agree with Lloyd, it is confusing... (that and PacketSize_ vs 
packet_size_ too) Shouldn't there be a deprecation message when using class_ 
instead of fid_? 

(btw, i have been using "$udp set fid_ $x"; and it works fine with my b5. From 
the look at agent.cc from the cvs, there is no reason it should not.)

Most people use old names because they either old found examples scripts from 
tcl/ex tcl/test, or else?

It's not the first time somebody comes up with a deprecated name, such as 
Agent/CBR instead of Application/Traffic/CBR for example. Usually it does not 
prevent one from running his/her simulation correctly but can give unexpected 
results, still.

Tarik.