Paul Sandoz wrote:
> Alan Hudson wrote:
>
>> Paul Sandoz wrote:
>>
>>> Alan Hudson wrote:
>>>
>>>> A built-in byte encoding algorithm might be useful. My encoders
>>>> are doing algorithm matching on a per element/attribute basis. So
>>>> if I have a string of numbers I test for floats, then ints, then
>>>> shorts... ie to best fit it. Sometimes a sequence of numbers will
>>>> all fit as bytes instead of shorts. I can/have made this a custom
>>>> algo, but it might have applicability across a lot of specs.
>>>
>>>
>>>
>>>
>>> Good point. The ASN.1 group will be meeting in months time to
>>> resolve ballot comments. It is a long shot but it just may be
>>> possible to add as this is a simple localized addition. We shall see...
>>>
>> I assume we could also add it later as an ammendment right? Ie
>> overtime you expect some number of new builtins right?
>>
>
> Yes, we have reserved space for additional built-in encoding algorithms.
>
ok. I'm going to prototype the BYTE one(pretty easy from the SHORT/INT
code). And I'll report back on how it affects X3D compression/parsing
rates.