Music Blocks/2026-03-08: Difference between revisions
Created page with "Reports: Hari -- Discussion about git backend Syed -- Working on v4 Ashutosh -- Installed Sugar on laptop; worked on performance-related issues; cleaned up spam PRs Jenish -- CICD pipelines Kartik -- How to submit a proposal? Send to sugar-devel Parth -- (left before called) Vanishka -- Working on Broken Calculator Activity; working on reading PRs Vyag -- Looking at reflection Naitik -- Worked on reviews Nikhil -- Git backend project Shreya -- Had a meeting on..." |
m Suryaanshchawla moved page Music Blocks/2026-02-08 to Music Blocks/2026-03-08: Misspelled title |
||
| (One intermediate revision by the same user not shown) | |||
(No difference)
| |||
Latest revision as of 11:20, 19 March 2026
Reports:
Hari -- Discussion about git backend
Syed -- Working on v4
Ashutosh -- Installed Sugar on laptop; worked on performance-related issues; cleaned up spam PRs
Jenish -- CICD pipelines
Kartik -- How to submit a proposal? Send to sugar-devel
Parth -- (left before called)
Vanishka -- Working on Broken Calculator Activity; working on reading PRs
Vyag -- Looking at reflection
Naitik -- Worked on reviews
Nikhil -- Git backend project
Shreya -- Had a meeting on lessons
Aman -- AI Optimization Project
Om Suneri -- Needs time to discuss after meeting
Sonal -- CICD
Krishna -- worked on security and stability
Discussions
CICD Pipelines and security
Assign a code owner files to restrict who can merge into PRs <-- right now, only a select few are able to merge, so this may not be an issue
Use semi-automation to help
GTK projects
Met with Krish; the scope may be too big
AI Guidelines
We have guidelines on Sugar AI
Devin TODO: https://github.com/sugarlabs/musicblocks/issues/5865 and 5866
Constructionism
Aesthetic -- helping a user by creating an environment that they feel comfortable trying things out and asking what they like, which they prefer, do these things compliment each other; if not, perhaps create something that does (for example in the Phrase Maker).
AI Guardrails
Devin's suggestions:
- No image creation
- No profanity -- perhaps add a time penalty if user uses profanity
- Other corner cases -- AI bots getting too friendly**; off-topic; fabricating lies from well-known science/history things without warning
- I've seen kids get the bot to say it was their grandparents or parents, after asked by students. We should probably avoid that.
Ashutosh -- how improved MB