Dustin Getz

hf/*$* when stage is broken

1

Introduced in https://github.com/hyperfiddle/hyperfiddle/commit/2750d74c24e352d655800f9b8d399686ae952d2c

Building here a list of cases to consider: * Manually staged tx fails security * Manually staged tx is malformed (e.g. :db/unique1 is not an attr) * ...

race condition in `with-entity-change-route!

3

If you beat the server, you can clobber the last entry

Fixed in norby demo branch

s/cat inst? renderer is missing css class .-db-type-instant

2

  The ref renderers changed as well, a couple bugs here.

``` (defmethod hf/tx (keyword `school-requested-block) ...) [:hf/iframe `sub-rename :tx-fn ":zero"] [:hf/iframe `school-requested-block :tx-fn (str `school-requested-block)] ```

Cleanup Datomic facade so system fiddles can d/q without hardcoding a Datomic product line

1

https://github.com/hyperfiddle/hyperfiddle/blob/3dcd1dd786ec3a6dd0b487c0ea5744e53e96070e/src/hyperfiddle/foundation.cljc#L139 Userland fiddles e.g. hello-world are happy to hardcode the Datomic dep, but hf provided fiddles must be agnostic. The current Datomic facade sucks for this, needs untangling.

select options sorting crashes for ref options

2

This commit is the crash https://github.com/hyperfiddle/hyperfiddle/commit/3a222c7f61ba382c9febc70c83f1182393535af9 Error is: `Cannot compare {:region/id :ca.district/napa-south} to {:region/id :ca.district/twin-no-livescan}`

Search field ui gray database color should stay in dynamic scope

3

Here, ::sub is tagged with sd/ref? aspect and thus gets a ref renderer for free, but should stay gray 

Widgets migrate to (f e a rets adds)

4

Here is what the picker interface could look like ```clojure (defmethod hf/render #{:rosie.hyperfiddle.sub/locations :swing/display-sub} [ctx props] #?(:cljs [picklist (assoc ctx ; This way is too much programming, not declarative enough...