LessonThis is a lesson restated in new words. Any figure in it is an example from the source, not this desk's record.
A single workshop is the right shape when the goal is one technique. A deep dive is the right shape when a public repository already exists and you can name the behavior you want explained or extended. The lecture’s own public code, including trading-system projects and a Python package installed with pip install qfin, is the menu it offered for that second choice.
Why it matters
Open time gets wasted in two opposite ways. A deep dive with no question becomes a tour of files. A workshop with no technique becomes a status meeting. The lecture, recorded in the middle of an office move, is unusually explicit about the fork: keep doing one-off sessions on a fresh topic each time, or pick a project and walk through it in detail, including extensions. Crypto was one area they were willing to enter if that was the interest. The fork is a study-design choice, not a comment on which topic is fashionable.
The concrete inventory matters because a deep dive needs an artifact. The lecture points at a public GitHub account, open-source projects, work related to algorithmic trading systems, and qfin as a packaged library you install with pip. It does not, in this check-in, document qfin’s functions or show a trading system’s results. Those would be the content of a later deep dive. This page is about how to decide that a deep dive should happen.
The idea in plain words
A one-off workshop has a single payload. You arrive needing a method, you leave having seen that method once, and the next session may be unrelated. That matches a calendar of isolated topics. It is a good fit when you cannot yet point at a repository, or when the thing you lack is a definition rather than a change to running code.
A project deep dive has a spine. One repository, one entry point, and a sequence of sessions that stay on that code. The lecture’s offer was specific about the kind of follow-up that fits: explain a project that is already public, or extend it. “Extend” means a behavior the code does not have yet, proposed before the session, not discovered by scrolling. Trading-system repositories are one pile. qfin is another, and the only install instruction given here is pip install qfin. Crypto was named as a possible neighborhood for the work, not as a completed system.
The check-in itself demonstrates the boundary. It does not teach a library call. It asks which format should come next. That question is answerable only if the next session’s input is clear: either a technique with no repo, or a repo with a technique attached.
How to use it
Write one sentence before you ask for a session. If the sentence names a technique and does not name a file, schedule a workshop. Examples that fit the lecture’s past pattern are a single numerical method, a language feature, or a definition. The session should end when that technique has been shown. It should not drift into a second, unrelated tool.
What usually goes wrong
Asking for a survey of every public project produces a directory listing. The lecture wanted a choice, not a catalog. Depth on one repo beats a paragraph on each.
Starting a deep dive from the package name alone fails in a different way. qfin is identified here only as an installable library sitting next to trading-system projects. Without a function or a bug, there is nothing to extend and nothing to explain past the install command.
A workshop that secretly needs a codebase will stall, because the one-off format has no place to keep state between sessions. If you notice the question depends on last week’s file, you are already in a deep dive and should say so.
Questions people ask
What is qfin in this lecture?
A packaged Python library on the public account, installed with pip install qfin. The check-in does not describe its functions. Those belong to a later session if you choose that package for a deep dive.
When is a one-off workshop the wrong format?
When you already know which repository and which behavior you care about. A single fresh topic each time will not accumulate into an extension of that repository.
Did the lecture pick crypto as the next project?
It named crypto as one possible interest, alongside the choice between one-off sessions and a shared deep dive. It did not start the deep dive or report a crypto result.