Bruce Conrad

Need a better parser/compile error message for mis-placed bindings/assignments

1

KRL doesn't allow name/value bindings except in a prelude or in a function definition. When these are tried in other places, the error messages are unclear/misleading and sometimes don't fail...

Need a String operator to do URI component encoding

1

This could be easily implemented with JavaScript's `encodeURIComponent()` function.

before a pico is deleted, its subscriptions should be deleted

3

otherwise it becomes impossible to remove the subscription from the other pico, even though it still remains

But running it through `krl-compiler --verify` didn't report an error. Trying it without the `--verify` flag... No error reported, but the generated Javascript code includes ``` var url3 = $state.setting["url"];...

Phil says, "a pico engine-supported link to IPFS would be excellent."

We ought to consolidate code in Wrangler for computing a URL given an absoluteURL and a RID

1

To install a ruleset, you have two options: 1. Provide the absolute (full) URL of where the source code is located, or 2. Provide an absolute URL of an already-installed...

1. install a ruleset useable as a module (call it `ruleset_one`) 2. install a ruleset which has `use module ruleset_one` in its `meta` block (call it `ruleset_two`) 3. modify `ruleset_one`...

From Phil Windley: "We need to update `wrangler:picoQuery()` and `event:send()` to better support location transparency. As currently implemented, the programmer is responsible for knowing (and often hard-coding) the pico's location...

The application that triggered issue #572 uses a [ruleset](https://github.com/Picolab/fully-sharded-database/blob/main/krl/io.picolabs.pds.krl) named `io.picolabs.pds` (which is how it avoids the use of entity variables). This is another, indirect way in which a pico...

Uncertainty in UX when changing out a ruleset used as a module

1

Changes made to an application repo in [this commit](https://github.com/Picolab/fully-sharded-database/commit/01d2073c5265724d58d44445e210eac3fd668660) and the previous one (and some that follow) have revealed some odd behavior, and multiple issues. Notice that we changed from...