Steve McCanne

We should improve the DX of the zed command error message when encountering an unknown command, e.g., ``` $ zed lo "zed": no such sub-command "lo": options are: auth branch...

Since the Zed data model is not based on tables and schemas but rather sequence of heterogenoues, deeply typed values, we should update all the references to row/column in the...

Thanks to [Timothy Green](https://twitter.com/TRiG_Ireland) for catching this bug in the ZSON spec per a recent tweet https://twitter.com/TRiG_Ireland/status/1538325276253380608 ``` It says that all JSON documents are valid ZSON documents, and then...

record literals with quiet errors should drop them

3

This should result in an empty record but doesn't: ``` echo 'error("quiet")' | zq 'yield {foo:this}' - ```

The new zed lake implementation has temporarily removed the data object cache. The old code still was removed and can be found in the package `lake/immcache` in the commit history....

index & slice operators on bytes/strings

2

expr.Index() should be able to index a `bytes` type and return a `uint8`. It should also be possible to index a string type and return a rune. (The index operator...

streamline PEG grammar

1

There are many PEG rules in the grammar that can be rewritten to reduce backtracking overhead. This perf issue became more apparent with the introduction of "over expressions" where sequential...

Reorganize and update the lake format specification so it is consistent with the current implementation and easy to update as we continue to develop the zed lake, add crud, incorporate...

The text below was removed from the Zed lake specification as these are really design notes for pending items. We should create issues from these notes and update the docs...

schema argument for CSV reader

1

The CSV reader should take a schema argument to handle cases where there is no CSV header in the file but simply a sequence of comma-separated values. The schema would...