Clojurians Log v2

love it. may i know any eta you have on [soon] items?

Hey, the goal was to get this project started and we are actively working on the next parts of the tutorial with no defined time table. We will definitely incorporate feedback. Thank you!

nice, but one question while reading, doesn't the :db/retract example retract just the single :item/text attribute? leaving the entity as an orphan. Haven't used datomic in a while, and I would have expected to see :db/retractEntity to remove all attrs and refs to it

That is correct, db/retractEntity is to remove all the datoms of the entity. In that example is to show how retract works in general, db/retractEntity will be use later.

also

(d/transact conn [{:db/id [:list/name "life"] :list/items [{:db/id "new-item" :item/text "todo 1" :item/status :item.status/todo}]}])

but it makes sense if you are starting from basics, to introduce the api stuff

Yes, it's nice to see that the tutorial is provoking to think on alternatives ways to achieve the same thing. What we decided to put there is what we think it's a good path to navigate the learning curve of Datomic. What you suggested also works well, the difference here is that instead of resolving the lookup-ref to get the eid in the transactor, you do it in the peer. In some cases this can have some performance benefits for the tempid, yes you can use a string, although it's good practice to have a programatic way to generate those instead of hardcoded. i.e. if you transact many item in the the same transaction, that "string" tempid needs to be different for each one

I like the lookup refs approach because I don't need to pass a db around when generating the tx data, but good to think about the peer vs transactor

Love Part 1 - really helpful - can't wait for part 2 :)