Re: WG ACTION: 2 weeks to discuss [LL69] Move constants to beginning of doc

On 5/4/04 6:24 PM, "Erik Guttman" <[email protected]> wrote:

> [Stuart]
> 
>> 9.  Constants
>> 
>>    The following timing constants are used in this protocol.
>> 
>>    PROBE_MIN              1 second
>>    PROBE_MAX              2 seconds
>>    PROBE_N                2
>>    NUM_PROBES             3
>>    MAX_COLLISIONS        10
>>    RATE_LIMIT_INTERVAL   60 seconds
> 
> Can we list this *before* they are used in the text?
> 
> Abstractions are great for experts, but not for people learning. Think
> about children. They learn specific examples first, and then from a
> collection of specific examples generalize to abstractions. Beginning
> with the abstract is not the way to inform the reader.
> 
> [Erik]
> 
> Listing constants early in the RFC is not common practice.  Constants
> are not abstractions.
> 
> [Stuart]
> 
> Using a symbolic identifier in place of a concrete value is not
> abstraction?
> 
> Since when?

I support this. It's far better practice to define terms prior to use.

john

-- 
"Keecking butt fahr gudness"
Baldur's Gate