Hi,

No more comments before v4. There are quite many small/easy things to fix 
already for v4.

Include update to this piece of text ...
"Current version of interface allow Compression ONLY,and
both Compression + hash ONLY sessions."

... You could describe the operation modes with a bit more context: when / how 
to use compression + hash vs. the normal just compress/decompress. No need to 
speculate on current version vs. future version.

About operation() vs compress() + decompress(). There's no need to minimize the 
number of function prototypes in the spec. Usually, it's easier to understand 
code which uses compress() and decompress() vs operation(x). Also specification 
is easier to write and maintain when you can refer to actual function calls vs 
function call + mode. Implementation may have more optimization opportunities 
with separate functions (e.g. if your HW implements   only decompress, and 
compress needs to be pure SW etc.).

-Petri


> -----Original Message-----
> From: shally verma [mailto:[email protected]]
> Sent: Tuesday, June 06, 2017 8:27 AM
> To: Verma, Shally <[email protected]>
> Cc: Savolainen, Petri (Nokia - FI/Espoo) <[email protected]>;
> [email protected]; Challa, Mahipal <[email protected]>;
> Narayana, Prasad Athreya <[email protected]>
> Subject: Re: [lng-odp] [RFC, API-NEXT v3 1/1] comp: compression interface
> 
> Do we have any more comments on same?
> 
> Petri
> 
> Could you get time to review feedback?
> 
> Can we consider v3 as accepted?
> 
> Thanks
> Shally
> 


Reply via email to