Arjay McCandless hunts bugs hiding between perfectly working functions
His debugging tutorial tops a five-upload week, alongside Git-server stress tests and a new eight-core programming challenge.
Why this story? The channel's weekly summary: what went up this week, what is drawing views now, and where the channel stands among similar channels.
10/10 prompt for debugging.
Next story: Arjay McCandless wouldn't ship his fastest server experiments
Arjay McCandless has a specific request for your coding agent: "Find all the implicit state machines in my app." The useful part of his most-watched upload this week is the explanation that follows, not just the prompt to copy.
In "10/10 prompt for debugging.", he uses an online order to show how software moves between states: pending, paid, shipped, delivered. When the rules governing those changes are scattered across functions and files, an order can slip into a state nobody intended, such as shipped but never charged. Each individual function might work; the trouble lives between them.
The tutorial has reached nearly 96,000 views. In the comments, developers describe recognizing a pattern they had used without knowing its name, while others discuss applying the idea at work. One viewer's practical obstacle to trying it immediately was the hour: well past bedtime.
His longer experiments continued, too. "How Many Developers Can ONE Git Server Handle?" puts bare Git, Forgejo and GitLab CE on the same hardware, then adds simulated developers working with repositories including Django and the Linux kernel. It's a concrete way to ask what the conveniences of a full Git platform cost.
The newest upload, "Which Programming Language Can Handle the Most Users on 8 Cores?", currently has the strongest viewing momentum. Arjay follows requests into Postgres and examines community submissions in languages including Java, Swift, C and Rust. That expands the hardware canvas of the small-server experiments readers may remember.
The channel gained about 1,000 subscribers over the week, though daily views are running somewhat below their usual level. Among nearby technology channels, Tool Finder had a slightly larger weekly viewing total, while Digital Spaceport had a quieter week. For Arjay, the fresh eight-core challenge is now drawing views fastest, with "Deploying a Vector Database" next among his active uploads.
Published on Trari
Where next?
What viewers say
You've just articulate something I unconsciously learned to do over a decade of being a software engineer. That was great
@jonathanmarler5808See it on YouTubeI would've tried it right now, if it wasn't 3:30am, but thanks for explanation
@SimpNickSee it on YouTubei always just called this "operation flow" or "decision tree", but i see now why things always felt close but never exact, and that is because i was addressing it by the wrong name. this is really the information i needed to fill the gap i kept trying to avoid!
@herbertmilhommeSee it on YouTube
4 more comments
I will use this at work, we have lots of these
@gianlucav.4780See it on YouTubeThis is why my application is both serverless and stateless. No bugs!
@slothmaster912See it on YouTubeHard core channel with deep meaty tech insights. Love it.
@trevthea5781See it on YouTubeHonestly interesting pattern to think about. Mega common as well implicitly.
I think an existing repo will have far too many. Maybe going forward it would be wise to incorporate this.@IhsanMujdeciSee it on YouTube
Public comments from the video's page on YouTube
Want to know when somebody answers you?
The videos drawing views on Arjay McCandless right now
- Channel
- Arjay McCandless
- Subscribers
- 137,000 ▲ 1,000 this week
- Channel views, last 7 days
- 477,583
- Videos published this week
- 5
- Video published
- November 10, 2026
- Video views so far
- 95,855
- Views on its first day
- 55,063
The numbers
Here is how Arjay McCandless's week looks in views and subscribers.
💡 Test your knowledge
Which database underpins Arjay's eight-core backend challenge?
The challenge description specifies an eight-core server with Postgres.






