Presently transactions are a block of code. Everything has to occur within it. This made using ArangoDB difficult for me when creating a self referencing data model that spans collections.
I've had better success with PostgreSQL for this particular problem. The key factor is that I can create a transaction, insert, get keys, and build up the model on the client side. The transaction spans many operations. Given what I know about Arango's internals such as the MVCC, it should be possible to create a transaction resource much like a query resource. I should be able to create a transaction, get a token and then make a series of calls (reads, posts) with the transaction token as part of the message. Then I can call PATCH to complete the transaction. Behind the scenes Arango uses the MVCC magic to ensure everything works. What do you think? I know you've got your hands full right now for the next release. How far off am I? Thanks, JPD -- You received this message because you are subscribed to the Google Groups "ArangoDB" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/d/optout.
