Hi,

Does ICE implement it’s own data type representation which is different from MYSQL’s? Because i found that ICE converts data to MYSQL’s representation in MYSQL’s tuple interface fill_row().

Does any encoding method depend on data representation or there’s some considerations other than that?

regards.

Hello Kurt,

Internally, we have two major generic data types: numeric and alpha-numeric.

kurt - 25 May 2011 10:18 PM

Does any encoding method depend on data representation or there’s some considerations other than that?

I’m not sure how to reply to this part of your question…

Could you please clarify?

Looking forward to your further questions,

Dominik

Dominik Slezak - 03 June 2011 01:13 PM

Hello Kurt,

Internally, we have two major generic data types: numeric and alpha-numeric.

kurt - 25 May 2011 10:18 PM

Does any encoding method depend on data representation or there’s some considerations other than that?

I’m not sure how to reply to this part of your question…

Could you please clarify?

Looking forward to your further questions,

Dominik

Thanks for your help.

I want to know why ICE implements it’s own data types instead of MYSQL’s?

Does ICE use any compression method which depends on datatype implementation?

Hello Kurt,

kurt - 07 June 2011 09:59 PM

Does ICE use any compression method which depends on datatype implementation?

Yes, ICE uses its own compression algorithms for data packs related to numeric and alpha-numeric columns. Indeed, it was far more reasonable to design it for a few generic data types. Actually, the same situation can be observed at the level of Knowledge Grid. We have specific Knowledge Nodes for data packs related to numeric and alpha-numeric columns. In summary, query optimization and execution is resolved at the generic level of both data and metadata.

Best greetings,

Dominik