[Biojava-l] ontology exception, addSequence & BioSQLSequenceD B

[Matthew Pocock]

I think this is safe. The core_ontology terms would be used if you 
commited any interpretation logic (e.g. rules saying what feature types 
are allowed or not allowed). However, we aren't doing that sort of thing 
yet.

The workaround for the last bug is apparently to mark the name field on 
term as BINARY - this ensures that it does a case-sensetive indexing. 
Postgresql does case-sensetive indexing automatically, but MySQL doesn't.

Best,

Matthew

Len Trigg wrote:

> Simon Foote wrote:
>
>> The new code does indeed break existing BioSQL databases.  As far as 
>> I can tell it relates to the changes in the core_ontology.
>> To test this, I found that within my existing databases, the 
>> core_ontology terms are never used, so I removed all the 
>> core_ontology terms from the database.  I then re-ran one of my DB 
>> programs, so the core ontology would be re-built, but it fails with 
>> this error. 
>
>
> So, is this the recommended upgrade method? (I'll have to check 
> whether or not the core_ontology terms are used in my db's, but I 
> don't think they are)
>
>
> Cheers,
> Len.
>
> _______________________________________________
> Biojava-l mailing list  -  Biojava-l at biojava.org
> http://biojava.org/mailman/listinfo/biojava-l
>