Similar to the Navigation feature now provided in all Livebooks I'd like to add a Raise Issue button so that beta testers and future
students can easily raise an issue.

At this moment in time, there is no way to get to instructions.md (the chapter list), navigate back and forth between chapters or jump to the top of the page (unless you expand sections). I have to hit the back button in my browser to get back to the contents page. Might be a local server limitation as not overly familiar with Livebook and how navigation works but good to let you know about.

We need to replace all calls to Utils.feedback with the new smart cells.

There's the TestedCell which we can write tests directly in the livebook. See the Mad Libs exercise for an example.

There's also the ExerciseCell we can use to create custom smart cells. See madlib.ex and exercise.ex for an example. This has the added benefit of having the solution validated. However, it comes at the cost of creating a new smart cell (which could eventually be cumbersome in the Livebook Smart cell dropdown.)

Existing tests can be stolen and modified from feedback.ex. Some tests could use some work to produce simpler output for students to be able to read better.

Here's a running issues list of things I've found so far. Some of these may have already been fixed, but I plan on going through and fixing these via a PR in the near future, I just wanted to provide a location where it can be easily monitored.

• Typo in calculate loading exercise, change "with" to "will". Previously on line 88 though this may not be the case anymore.
• true OR ___ must evaluate to true. in boolean operators exercise challenge 3 this is being read (to me) as true OR ___ = false in the mermaid.js graphic.
• Your Turn 2 in Learning/Maps, set the variable by default.
• atom_maze does not exist in Exercises/Mazes, this is just due to the variable not being defined properly at the beginning. This should be a simple fix from string_maze to atom_maze in the variable name of the second line.
• W1D4 calculate_force your turn should be mass * acceleration instead of mass * force.
• W1D4 structs - missing example in functions as structs section.
• W1D4 structs - secret_identity should be an atom in "Your Turn" description.
• W1D4 rpg dialogue - consistency issue in :dagger and :daggers in RPG dialogue challenge. Should be consistent throughout document for readability and ease of implementation.

Additionally any of these fixes would make good first issues. I plan to tackle them when I have some spare time and add to the list, but anyone else is free to fix these as well. 😄

The following elixir cell fails to eval correctly:

Utils.table(:example)

When calling a student module for feedback, we currently show the crash. Instead, it would provide better feedback to the student to
attempt calling their module and show the input and the expected value if it crashes.

Describe the issue
Depending on where students run livebook from, or if they are using the Livebook Desktop Application, the File module may behave differently.

We need to verify this by starting livebook from a folder other than the curriculum folder with the File lesson.

Expected behavior
The File lesson should be consistent regardless of the student's current directory. Either we need to make a consistent relative File path (harder) or rewrite the exercise to be agnostic of the current directory (easier).

A couple of ideas to expand the navigation section of each lesson and exercise:

1. Adding a next lesson button to each livemd lesson to allow students to organically move from lesson to lesson in curriculum order without the need to return to the start.livemd file.

2. Adding a "related exercises" button to lessons and "related lessons" button to exercises, allowing easier navigation between lessons and exercises.

Using the File module we can read the start.livemd file which contains the outline of the project, then use this file to automatically generate the Curriculum Outline section of the README.

All links such as [Lesson](reading/lesson.livemd) should be relative.

There are a few incorrect links inside the reading folder that use [Lesson](reading/lesson.livemd) when they should use [Lesson](./lesson.livemd) or [Lesson](lesson.livemd).

The History Section of the IEx reading was not working correctly.

History

You can retrieve previously run commands in the IEx shell by pressing the Up and Down arrow keys.

Each IEx session stores its history but not the history of previous IEx sessions.

If you would like to preserve history between IEx sessions, you can run the iex command
with the --erl flag.

$ iex --erl "-kernel shell_history enabled"

Alternatively, refer to the Shell History documentation
to enable shell_history for your whole system.

It needs to be redone.

Each livebook runtime generates a single board in the Bingo exercise and then uses that same board rather than a randomized board each time.

Overview

When evaluating the comparison operator form in the 'Comparison Operators' chapter, after completing it and reevaluating, there is no output to confirm what you entered was correct or not (true or false).

Recreation steps

1. On the code block, select 'Evaluate' to open the form (that's not immediately obvious by the way).
2. Enter the operators into the form fields and once completed, select 'Reevaluate'.

Behaviour

Confirm there is no output to verify what you entered was correct.

Expected behaviour

There should be output to verify what you entered was correct.

In the 1. Course Tools / Livebook part there's an instruction that indicates that the "Ctrl + Shift + I " keybinding is used to format the code in livebook, however this keybind will prompt the browser element inspector to open

Ideally, students should regularly create pull requests for each day of material to develop familiarity with Git and GitHub.

The start of each day should include instructions on how to make a new branch.

$ git checkout -b week-1-day-1

Then each reading and exercise should include instructions on how to add, commit, and push changes.

Commit Changes

$ git add reading/command_line.livemd
$ git commit -m "complete command line reading"

At the end of the day, students should make a PR, and merge their changes.

1. Make a pull request.
2. Merge the pull request.
3. Pull changes locally:

$ git push -u origin week-1-day-1
$ git checkout main
$ git pull

This may require changing the start.livemd to separate each day into its own .livemd file.

Overview

Similar to the issue reported in #5 , the Integer forms do not appear to evaluate the values input by the learner even after choosing reevaluate.

Similarly, when doing the arithmetic exercises, there is no confirmation of what the answer should be, so it's hard to verify what you did was correct or not (without doing the maths yourself 🙂). I think that seeing the answer you expect to return would help you figure out the right way to do the work. For example, if you had to do 10 divided by 3 and the answer expected was 3.3333333333333335, then you know you need to do 10 / 3 as opposed to div(10, 3) because the latter does not result in a floating point number. Helps cement the learning, unless I'm misinterpreting it?

Recreation steps

1. On the code block, select 'Evaluate' to open the form.
2. Enter a value into the form field and once completed, select 'Reevaluate'.

Behaviour

Confirm there is no output to verify what you entered was correct.

Expected behaviour

There should be output to verify what you entered was correct.

Describe the issue
Code cells using Utils.animate/1 in Livebook never stop computing.

Reproduction Steps (If Applicable)
Steps to reproduce the behavior:

1. Go to 'Arithmetic' section of curriculum in Livebook
2. Scroll down to 'Integers'
3. In the "Your Turn" section, click 'Evaluate' for the code cell (Utils.animate(:biggest_integer)
4. Watch as the numbers start increasing and never stop

Expected behavior
Utils.animate/1 can/does stop at some point.

Possible solutions:

• Utils.stop_animate/1
• allow toggling animate on/off with evaluate and stop evaluate buttons?
• ???

Screenshots
See it here: https://www.loom.com/share/5e528da9f48a44339d26fc91f34a6c63

Livebook has added hidden setup sections on the latest version that we can use to hide the boilerplate setup code for each lesson.

Stack trace:

 1) test boolean_diagram3 (:boolean_diagram3)
     utils/lib/feedback.ex:237
     Assertion with == failed
     code:  assert answer == false
     left:  true
     right: false
     stacktrace:
       utils/lib/feedback.ex:239: (test)

I don't believe there's a way to get the diagram to false either way. When I interpret the top-left portion as true or true that evaluates to true. It's like true or evaluates to true no matter what, only an and would get us into falsey territory.

This screenshot should illustrate this a little better:

Loading bar displays markdown in Basic > Arithmetic > Floats.

Links to Elixir documentation such as https://hexdocs.pm/elixir/1.13.3/Kernel.html should have the version removed to ensure we
do not have to regularly maintain them.

For example, https://hexdocs.pm/elixir/Kernel.html.

Describe the issue
When livebook server start.livemd is called Livebook could not start distributed node.

Note: This is a Livebook known issue (196, 194, 1042) that happens when some erlang files don't show a dialog for its firewall approval on the first Livebook execution after being installed.

As there is no windows version for asdf, my tools setup was done with Chocolately and NVM.

Solution proposal
Add a Windows-specific solution for this behavior on documentation:

• Before calling Livebook, execute on terminal iex --sname test
• On appearing Windows firewall dialog, approve permission for epmd.exe
• Both erl.exe and epmd.exe should appear on firewall-allowed apps

Environment
OS: Windows 11
Erlang/OTP: 24 [erts-12.0]
Elixir 1.13.3 (compiled with Erlang/OTP 22)

Reproduction Steps (If Applicable)
Steps to reproduce the behavior:

1. Oen command line (CMD/PowerShell)
2. From the beta_curriculum project folder execute livebook server start.livemd

Expected behavior
Curriculum Liveboook server should start and load the curriculum.

Actual behavior
Curriculum Liveboook is not started.

Screenshots/Logs

C:\Users\Oct\Elixir\DockYard-Academy\beta_curriculum>livebook server start.livemd

ERROR!!! [Livebook] Could not start distributed node: {{:shutdown, {:failed_to_start_child, :net_kernel, {:EXIT, :nodistribution}}}, {:child, :undefined, :net_sup_dynamic, {:erl_distribution, :start_link, [[:livebook_7isuoce5, :shortnames, 15000], false, :net_sup_dynamic]}, :permanent, false, 1000, :supervisor, [:erl_distribution]}}

Anytime we make calls to the Utils library, we can instead make a hidden cell.

For example, the following in binary_search.livemd.

Utils.graph(:binary_search)

Could be replaced with a HiddenCell smart cell with the following content.

size = 600

widget =
  VegaLite.new(width: size, height: size)
  |> VegaLite.mark(:line)
  |> VegaLite.encode_field(:x, "number of elements", type: :quantitative)
  |> VegaLite.encode_field(:y, "time", type: :quantitative)
  |> VegaLite.transform(groupby: ["color"], extent: [2500, 6500])
  |> VegaLite.encode_field(:color, "type", title: "Big O Notation", type: :nominal)
  |> Kino.VegaLite.new()

init = 1
max = 500

logn = for n <- init..max, do: %{"number of elements": n, time: :math.log2(n), type: "log(n)"}

Kino.VegaLite.push_many(widget, logn)
widget

Smart cell editors are controlled by the :display_editors flag in Utils/lib/flag.ex.

defmodule Flags do
  @flags %{
    # Display Teacher-only editors in smart cells. Students should not see these editors when we launch.
    display_editors: false
  }

  def get(key) do
    @flags[key]
  end
end

Enable the flag to view the hidden cell editor. Ensure you do not commit this change as students should not see the editor.

After walking through the item_search.livemd exercise with a beta tester, we realized it would be more understandable if the exercise were re-themed to books instead of fantasy items.

• re-theme to use books instead of fantasy items
• potentially include step or suggestion to make a matches_filters(item, filters) function.
• improve tests to be more understandable
• improve the clarity of inclusive vs. exclusive search.
• perhaps include runnable input/output examples below the exercise.

TestedCell has been extracted into it's own repository. https://github.com/BrooklinJazz/tested_cell.

We need to add the dependency to the setup section in every livebook.

{:tested_cell, github: "BrooklinJazz/tested_cell"}

And ensure any existing calls to TestedCell continue to work. We may need to recreate them.

Add a mechanism to track student progress and completion of exercises.

Create a new Validated Form smart cell to replace calls to Utils.form.

Ultimately, I'd like to eventually remove the Utils project.

The guide says map will start as a person with a :name, :age, : status and :parents keys but there seems to be mistake later down on the fourth point where it describes as children and also its own keys.

The instructions for installing FontAwesome are incorrect in exercises/home_page.livemd.

<link rel="stylesheet" href="path/to/font-awesome/css/font-awesome.min.css">

We need to either use a different icon library or add the correct FontAwesome instructions.

livebook-dev/livebook#1299

We don't want to accidentally commit calls to TestedCell.Control.show_editors that enable teacher-only SmartCell editors.

We need to write a "Teacher-only editors are hidden" ExUnit test in the Utils project. The test should:

• Be in the utils_test.exs file.
• Read all files ending in .livemd in the /exercises and ./reading folders.
• Ensure the string "TestedCell.Control.show_editors()" is not found in each file.

Some considerations:

• Don't load every file into memory. One file at a time is fine. Streaming files line by line is even better for avoiding memory issues.
• You can use File to list files in the desired directories, read the contents of the file,
• You can use String or Regex to check the contents of a string, and replace values in a string.

Comment on this issue if you are interested, and I'm happy to provide more direction.

By reading the start.livemd file using the File module we can determine the order of reading material and exercises.

Once we have the expected order of reading material we could either

1. write a script that automatically adds an "Up Next" section to each .lvemd file.
2. manually add the "Up Next" section to each .livemd file and create a test to ensure each .livemd file includes a link to the next .livemd file.

For example, here's the first day of the course in start.livemd.

1. Course Tools

• Reading
  • Command Line
  • Git

We would expect reading/command_line.livemd to include the following "Up Next" section:

Up Next

• Git

It may be more challenging, but I'd prefer to write this as a script to make it easier to modify the "Up Next" sections. For example, we may want to include the previous .livemd file.

Kino Lab is about to drop support for using the YouTube video Id directly in favor of using the full URL: jonatanklosko/kino_lab@450c02f

We will need to find all uses of Kino.YouTube and add https://www.youtube.com/watch?v= before the video id.

Several exercises are missing a solution causing tests to fail.

  def rollable_expressions do
  end

  def caesar_cypher do
  end

  def classified do
  end

  def phone_number_parsing do
  end

  def convert_phone do
  end

  def obfuscate_phone do
  end

  def email_validation do
  end

  def cypher do
  end

  def metric_measurements do
  end

  def money do
  end

  def rock_paper_scissors_pattern_matching do
  end

We talk about pattern matching in week 2 with more complex examples, but simple pattern matching could/should be included in the match operator section in week 1.

utils cannot install, File module calls are incorrect.