Code Driven Manifesto
Want to back the stream? The «داعم» membership on YouTube is EGP 60 a month: badges, emoji, and I answer your questions first on stream. Become a supporter
The gist
The stream centers on an unfinished hands-on review of Burnish and its Code Driven Manifesto workflow: generating, editing, and exporting a manifesto; turning it into internal instructions for the storefront, articles, visibility, reputation, and discovery; and considering how its status and Autopilot integration should appear. Testing surfaced issues with generation, progress, navigation, state updates, scoring, auto-publishing, and background jobs. Some next-batch logic appeared to work in one attempt, but critical bugs and questions about cancellation, timeouts, concurrency, and persistence remained under investigation. Luqman emphasized backend wiring, user experience, and debugging before interface polish, arguing that premium-positioned software needs commensurate reliability while acknowledging that no launch can eliminate every bug. Other discussions covered an exploratory Telegram community workspace, audience questions about learning programming with AI and the advantage of experience, uncertain opinions about coding-tool quality, pricing, and context capacity, and acquisition and retention measurement without invented figures, plus shorter conversations about YouTube music copyright, Algerian music, personal goals, and screen-use glasses.
85 moments from the stream
- 1:11:19
Learning from the Market Before Prolonged Study
An experience-based view recommends testing a field’s practical value before committing to prolonged study, recalling learning digital marketing through direct market experience and later studying it for six months to a year.
- 1:13:44
Completing Application Phases Before Debugging
The guidance is to carry the remaining phases through to completion instead of pausing to tweak details along the way, then review the finished application and begin debugging.
- 1:19:25
Fixing a UI Link, Spacing, and Reinstalling the Application
A UI review identifies an action that appears to have a broken link for “start over with the questionnaire,” calls for a Shopify secondary CTA and a refine button, requests appropriate spacing between nearby elements, and plans to delete and reinstall the application.
- 1:26:21
Plan to Finish the Storefront and Wire the Manifesto
The plan is to finish the remaining wiring, complete and stabilize the merchant product-management storefront, then return to the Manifesto and connect it downstream because it also produces a system prompt.
- 1:27:32
Manifesto Data Flow Across the Application
The speaker explains that the system generates a merchant manifesto and extracts structured data from it through a system prompt designed to feed the wider application, including topics and articles, visibility, repetition and discovery prompts, and product fields in the storefront. He also outlines the intended work sequence and notes that backend wiring and bug fixing take more time than final polish.
- 1:31:24
Context Capacity and Conversation Compaction in GPT Models
While expressing uncertainty about the pricing of a future GPT release, the speaker discusses why a larger context capacity matters to him. He says limited context could cause a conversation to be compacted after only a small number of prompts, leaving insufficient context available.
- 1:35:49
Limiting the Primary Market to One Country
During a pre-push and merge review, the speaker identifies that the target-audience question should allow only one primary market. He proposes replacing the one-item-per-line input with a selector that displays countries, or major countries, for a single-country choice.
- 1:39:36
Quality Versus Price in AI Coding Tools
The speaker comments on something he had heard about Codex prices potentially falling to compete with Claude, but argues that the more important competition should be on quality rather than price because many people in the field prioritize quality. The possible price reduction remains reported speculation and opinion, not a confirmed fact.
- 1:39:59
AI: Capability Versus Price and Different Usage Markets
The speaker argues that some users prioritize capability over price, while qualifying that his perspective is limited to coding tools and that mainstream chat-product usage may present a different picture. No verified figures are given.
- 1:41:05
Experimenting with a Collaborative Virtual Workspace in Telegram
An experimental Telegram-based virtual workspace where participants could have desks showing current work, demos, timelines, help requests, and possibly public GitHub activity. The concept is still exploratory; the repository might be shared for collaborative development if there is interest, with no current intent to monetize it.
- 1:43:58
Deferring Supabase Setup Until the Current Task Is Finished
The speaker says the application will need Supabase but prefers to finish the current work before setting it up, illustrating a simple implementation-prioritization decision.
- 1:47:23
A Navigation Logic Issue in the Questionnaire Restart Flow
A possible navigation-logic issue is identified: after choosing refine and then restarting the questionnaire, a “back to options” button appears. The excerpt ends before the impact or solution is explained, so the full nature of the bug remains uncertain.
- 1:53:22
Resetting the App to Test the First-Time User Experience
The app was prepared for uninstalling and a full data reset—including manifesto and product data—to examine a genuine first-time-user experience after a reported successful merge and deployment.
- 1:55:18
Considering System-Prompt Storage and Downstream Effects
An architectural question was raised about where the system prompt is stored and how it might later be created to affect the intended downstream behavior, explicitly as brainstorming rather than execution.
- 1:58:15
Responding to Boredom by Defining a Goal
A general opinion suggested that boredom can sometimes be connected to lacking a vision or goal, with advice to reflect on goals in life, work, and the future and choose something purposeful to pursue.
- 2:09:01
Unexpected AI-Readiness Score Change After a Rescan
The AI-readiness score was observed to rise immediately after using the rescan button even though no products had been optimized, prompting questions about the discrepancy and how the initial score was calculated. This page provides no answer or diagnosis.
- 2:17:23
Fixing Bugs in Parallel During a Storefront Review
The discussion considers assigning a specific bug to a separate session so the storefront review can continue uncovering other issues. It also asks for a concise, plain-language description of the fix while following the repository contract, with a preference for resolving bugs individually.
- 2:21:10
Testing Autopilot and Next-Batch Logic
The review focuses on autopilot as a major, bug-prone feature, checking whether recent optimizations appear and whether products move into a ready state after a delay. The displayed result suggests the next-batch logic worked in this attempt, after which testing turns to the approval threshold for unlocking the next level.
- 2:26:12
Debugging Across Different Store Sizes
The discussion explains that debugging requires focus and anticipation of varying conditions, including large differences in product counts between stores, so the application can behave correctly across different store sizes.
- 2:30:32
Testing Autopilot Level Progression
A hands-on test checks the autopilot logic by publishing a manually optimized product page, observing how it appears and is counted, then reaching level two and reviewing the requirement for progression toward full autonomy. The captions also indicate remaining errors and some uncertainty during the test.
Got a question about something in this stream?
Ask in the group