Fork me on GitHub

I had the same problem upgrading to 569. I gave up.


Hi @U0773UB6D Do you recall if this was your first upgrade on a split stack? If so, we have identified this issue as a bug and are working to address in a future release. In the interim, you can get around the issue by running the upgrade with “reuse existing storage” set to false and it should succeed. Note, you will still have your existing storage and it will be used, this option just moves the CF down an alternate path.


I don't think this is the first upgrade. I'll try that workaround.


A while ago a watched a presentation by Stu about a Typed Java DSL for accessing Datomic? Does this exist anywhere? I can't find any information on it.


Question about keywords in Datomic. In it says: > Keywords are interned for efficiency. What does this mean? I know keyword literals are interned in Clojure. Does Datomic intern keywords it encounters? Say keywords are dynamically generated (by parsing incoming JSON requests from a client for example) and are stored as keywords in Datomic. Does Datomic have an optimized way of storing/querying them?


After experimenting, it seems clojure also interns dynamically created keywords, so does datomic do anything special in addition?


Interesting case I encountered today when renaming an attribute. It seems that in order to rename an attribute without downtime you have to: 1. update your code to specifically pull your old attribute (`[:old/attribute]`), [*] will return the new attribute name as soon as you change the schema in step 2 2. update the schema {:db/id :old/attribute :db/ident :new/attribute} 3. update your code to use the new attribute Does it make sense? Do I miss something?


Correct. I think the lesson should be “don’t use star” if you know the specific attribute you want. Star is for repl exploration not produciton code


It seems that way. I wanted to double check here before I accept the lesson 🙂