{"user_signed_in":false,"allow_comments":true,"is_author":false,"request":{"endpoint":"https://exercism.org/api/v2/tracks/go/exercises/react/community_solutions/soniakeys/comments","options":{"initial_data":{"items":[{"uuid":"3ae4f314-08d3-11ec-81c5-853579c25f94","author":{"handle":"petertseng","avatar_url":"https://assets.exercism.org/avatars/92961/0","flair":null,"reputation":"18.9k"},"content_markdown":"An interesting observation to make about DAG vs slices: Let's say there's a reactor, first we create an input cell, then one compute cell connected to the input cell. Then we create 100 cells doing various things, not connected to these first two cells at all. Then we update the first input cell. The DAG solution would know to only update the one compute cell, whereas with the slice based solution it would iterate through all the 100 cells calling their compute functions only to find that nothing changed for them. Yeah I see the advantage is that you can do the value updates and callbacks in one pass since no cell can possibly depend on a cell that comes before it in the slice. At what number of cells does one approach overtake the other? Yeah don't know, I don't think I have the time to try to test that. Just speculating I guess.","content_html":"

An interesting observation to make about DAG vs slices:

Let's say there's a reactor, first we create an input cell, then one compute cell connected to the input cell. Then we create 100 cells doing various things, not connected to these first two cells at all. Then we update the first input cell. The DAG solution would know to only update the one compute cell, whereas with the slice based solution it would iterate through all the 100 cells calling their compute functions only to find that nothing changed for them.

Yeah I see the advantage is that you can do the value updates and callbacks in one pass since no cell can possibly depend on a cell that comes before it in the slice.

At what number of cells does one approach overtake the other? Yeah don't know, I don't think I have the time to try to test that. Just speculating I guess.

","updated_at":"2018-07-13T10:01:29Z","links":{}},{"uuid":"3ae4f3fa-08d3-11ec-81c5-853579c25f94","author":{"handle":"soniakeys","avatar_url":"https://assets.exercism.org/avatars/13984/0","flair":null,"reputation":"2,166"},"content_markdown":"Oh, and RemoveCallback here has a bug. Yes it passes TestVersion 2 but I think it might crash an upcoming TestVersion 3.","content_html":"

Oh, and RemoveCallback here has a bug. Yes it passes TestVersion 2 but I think it might crash an upcoming TestVersion 3.

","updated_at":"2018-07-13T10:01:29Z","links":{}},{"uuid":"3ae4f472-08d3-11ec-81c5-853579c25f94","author":{"handle":"soniakeys","avatar_url":"https://assets.exercism.org/avatars/13984/0","flair":null,"reputation":"2,166"},"content_markdown":"Iteration 1 data structure was a pointer-linked DAG. Slices are usually better than pointer-linked things so this iteration uses slices. Recalc processes the ordering from the changed cell on, just recomputing everything.","content_html":"

Iteration 1 data structure was a pointer-linked DAG. Slices are usually better than pointer-linked things so this iteration uses slices. Recalc processes the ordering from the changed cell on, just recomputing everything.

","updated_at":"2018-07-13T10:01:29Z","links":{}},{"uuid":"3cf89d4a-08d3-11ec-81c5-853579c25f94","author":{"handle":"mariusbalcytis","avatar_url":"https://assets.exercism.org/avatars/329252/0","flair":null,"reputation":"29"},"content_markdown":"Really cool example how an integer can implement all those interfaces. I also like how you just iterate through all cells – if some change in the middle, dependent ones will get updated later in the same loop. The only downside I can see here is that if there would be really large of cells, but only some would be related, there would be many unneeded computations. Great solution!","content_html":"

Really cool example how an integer can implement all those interfaces.

I also like how you just iterate through all cells – if some change in the middle, dependent ones will get updated later in the same loop. The only downside I can see here is that if there would be really large of cells, but only some would be related, there would be many unneeded computations.

Great solution!

","updated_at":"2019-08-23T18:46:26Z","links":{}}]}}},"links":{"create":"https://exercism.org/api/v2/tracks/go/exercises/react/community_solutions/soniakeys/comments","enable":"https://exercism.org/api/v2/tracks/go/exercises/react/community_solutions/soniakeys/comments/enable","disable":"https://exercism.org/api/v2/tracks/go/exercises/react/community_solutions/soniakeys/comments/disable"}}

Language Tracks

#48in24 Challenge

Your Journey

Exercism Perks

Community Videos

Brief Introduction Series

Interviews & Stories

Discord

Forum

Getting started

Mentoring

Docs

Contributors

Donate

About Exercism

Our Impact

Insiders

SWAG

soniakeys's solution 2,166 to React in Go

About the Go Exercise

Other solutions to React in Go

soniakeys's solution
2,166

to React in Go

Other solutions to React
in

Go