Re: Proposed ISSUE: Flatten algorithm identifiers

Yeah, I agree that this is largely an issue with WebIDL - eg: just
because WebIDL isn't expressive enough doesn't mean we should hold
things back.

The next ED will attempt to resolve this - although expect a few bugs
with language to creep in.

On Mon, Mar 18, 2013 at 4:21 PM, Mark Watson <watsonm@netflix.com> wrote:
>
> On Mar 18, 2013, at 4:14 PM, Richard Barnes wrote:
>
>> (Brought this up earlier, but I don't think it made it into the tracker...)
>>
>> The current spec requires that algorithm parameters be encapsulated in a "params" field within an algorithm identifier.  For example:
>> OLD: { name: "AES-GCM", params: { iv: [0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0], tlen: 128 } }
>> NEW: { name: "AES-GCM", iv: [0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0], tlen: 128 }
>>
>> The only reason this extra complexity would be necessary is if we expected an algorithm to need a parameter called "name".  This does not seem very likely; I see no problem making "name" a reserved word for the purpose of algorithm parameters.
>>
>> This requires the useless interface "dictionary AlgorithmParameters {};" from which individual algorithms specialize.  Proposed revisions to WebIDL:
>> -- Remove the AlgorithmParameters definition
>> -- Remove the "AlgorithmParameters params;" from the Algorithm interface
>> -- Change all descendants of AlgorithmParameters to instead inherit from Algorithm
>>
>> Whatever the opposite of "syntactic sugar" is, this is it.  "Syntactic vinegar"?
>
> "Syntactic arsenic".
>
> i.e. I agree.
>
> ůMark
>
>>
>> --Richard
>>
>
>

Received on Monday, 18 March 2013 23:25:23 UTC