Re: [hybi] Framing take IV [why fragment]

[Scott Ferguson <[e-mail]>]

Ian Hickson wrote:
> On Wed, 4 Aug 2010, Patrick McManus wrote:
>   
>> But I also believe fragments with fixed lengths are simpler than a BER 
>> because properly aligned lengths are always correct data - there are no 
>> error cases as all the bits are defined to be some quantity regardless 
>> of their arrangement. BER has error cases and determining the expressed 
>> quantity requires multiple indirections and a significant algorithm to 
>> use it.
>>     
>
> As far as I can tell there are no error cases as defined in WebSockets.
>
> Variable-width lengths also doesn't preclude chunking at a future point 
> should it prove to be needed (which we would be able to tell because we 
> would see authors making their application-layer protocols implement 
> chunking on top of WebSockets).
>   

No.

Chunking is a direct consequence of using a buffered stream, like fwrite 
and fread. It is not an application-level concept at all. If you do not 
add chunking, an implementation cannot implement an efficient basic 
stream for it.

If you do not understand why server-side HTTP use streams as the basic 
interface: CGI, PHP, Java, etc, you don't understand HTTP.  If you do 
not understand how streams work or understand why servers would choose 
that API, then you don't understand the basic concepts of chunking.

If you do not understand how fwrite works, or think fwrite or PHP's echo 
or CGI or Java's OutputStream are enormously-complicated concepts (or 
even new technology), then you might consider refraining on commenting 
on the usefulness of chunking until you learn these basic concepts.

-- Scott