Why can't I update a schema that has data ?

I am using Tamino 3.1.1 on Windows 2000 Professional.
I have a very simple schema and it contains data in it.
When I try to insert a new field to the Schema and update the Schema I recieve the following error code. INOXDE7956
The picture of the error is in the attached gif file.

Note : The minOccur of the new field I am trying to add is 0.
A friend of mine tried the same thing on W?ndows NT and Tamino 3.1.1 he said he did not recieve any errors.
Is this error special to W?ndows 2000 ?


I have the same problem but my component’s type is choice. I am using w2k prof. too. Is it really possible that this caused the error? I just hardly believe it!!!

My error message code is: INOXDE7955 (it says that you can add components), the same is true with the seq. type but you must set the minoccurs to zero, and the order is cannot be arbitrary.

So, did you solve your problem, if yes, how?


probably your TSD3 schema has set the validation

probably your TSD3 schema has set the validation

(sorry I obviously hit the wrong button before)
probably your TSD3 schema contains a
tsd:contentopen</tsd:content> for the respective doctype. That means, that the TSD3 validation allows for arbitrary child nodes.
On the other hand, the basic restriction during an update-define is to grant that all existing instances stored in a doctype do validate against the new schema as well. However, adding additional nodes in the schema may be in contradiction with additional nodes being allowed by the old schema due to open-contant validation.
That is probably the reason, why your new schema is being rejected.
Probably, a future version of Tamino will provide a schema evolution with explicit validation.

Hello Ulrich,
Thanks your reply, it clears some points in my mind. Please give me an advice or guideline:

What should I do if I want to store up to 150000 documents?
I have many different search queries, with some pre-defined nodes in the search filter. Later, the user can modify the search filter (adding searchable nodes or removing searchable nodes), and I must correct the schema to improve the search (adding index, or mapping), and in the future the schema will be changed (some new nodes will appear).
A, I define the schema according to my best knowledge using open-content, but, later I can’t update my schema (unload/undefine/define/load).
B, I define the schema using closed-content, I can update my schema, but I must correct the schema before anybody wants to insert it to the database, and all nodes must appear in the schema.
C, I define the schema using open-content, and I set the document type to ANY storing all documents under this node.

So, what should I do? What is your experience? I think that I make something wrong.

I have tried the A and C with a 31111 database about 25000 documents. I use taminoclient.query, plus a count so the query runs twice (unfortunately, the getTotalCount not reliable).
I experienced that the A (maybe the B too) generated a very good response time, less than 2 sec together.
But the C choice, it generated a very bad response time, the query runs 18-20 secs, but the count takes 222 secs.

thanks in advance,