Hi Florian, Thank you for clarification.
Vyacheslav Pascarel -----Original Message----- From: Florian Müller [mailto:f...@apache.org] Sent: Friday, October 21, 2016 4:09 AM To: dev@chemistry.apache.org Cc: Vyacheslav Pascarel <vpasc...@opentext.com> Subject: Re: "null" values for multi-value properties and CMIS specs Hi Vyacheslav, If I recall correctly, most vendors involved in the specification had trouble distinguishing null strings from empty strings. That's why null values are not supported. From a protocol and OpenCMIS low-level API point of view you can send null values, but that would be outside the CMIS specification. - Florian > Hi Everyone, > > A have a general question regarding CMIS specs. CMIS 1.1 says: > > ------------------------- > A property, either single-valued or multi-valued, MAY be in a "not > set" state. CMIS does not support "null" property value. If a > multi-valued property is not in a "not set" state, its property value > MUST be a non-empty list of individual values. Each individual value > in the list MUST NOT be in a "not set" state and MUST conform to the > property's property-type. > A multi-valued property is either set or not set in its entirety. An > individual value of a multi-valued property MUST NOT be in an > individual "value not set" state and hold a position in the list of > values. An empty list of values MUST NOT be allowed. > ------------------------- > > Does anybody know what is the reason behind not allowing individual > values to be in "value not set" state for multi-valued property? Is > there any way to work around that limitation (beside "reserved" values > or extensions)? > > > Regards, > > Vyacheslav Pascarel
