# A university course index that is one README and a hundred dead ends waiting > This repository is a hand-curated index of Chinese university course resources, and its entire content is a Markdown table of links organised by school and school type. That makes it genuinely useful for finding a course strategy in one click, and genuinely fragile the moment any of the linked repositories moves. **nwuzmedoutlook/university** — 120+ course resources from Chinese universities, hand-compiled; contributions and corrections welcome. - Repository: https://github.com/nwuzmedoutlook/university - Website: https://workhard.eu.org - Stars: 1,230 · Forks: 90 - Language: Unknown - License: Apache-2.0 - Published: 2026-10-12 · Updated: 2026-10-12 · Language: en - Canonical page: https://hysenlabs.com/projects/nwuzmedoutlook-university ## What the repository contains: three files The top level of this repository holds three entries: a README, a terms file and a CNAME. There is no code, no scripts, no data file and no documentation beyond the README. That fact shapes everything else about the project. It is a hand-maintained index rather than a tool, and its quality depends entirely on the discipline of whoever edits the table. The repository description says as much, describing it as 120 or more domestic university course resources, hand-collected, with contributions and corrections welcome. The CNAME is the one piece of configuration, and it points at a custom domain, which means the README is served as a GitHub Pages site under a domain the maintainers chose rather than under a github.io address. That is a small decision with a large consequence for this kind of project: if the maintainer stops renewing the domain or the Pages setup, the whole index becomes harder to find, and the repository itself remains the only copy. The README also links out to a separate organisation site for joining the group, a career and postgraduate study experience site, and a third-party blog page offering free courseware. Those are community links rather than documentation, and they are the same kind of external dependency as everything else here. So the honest characterisation is this: a table of pointers maintained by hand, published as a web page, whose entire value is the curation and whose entire risk is link rot. ## How the table is organised, and what the tiers mean The README is a single Markdown table with four columns: a sequence number, the school name, the school type, and the resource links. The links cell uses line breaks to stack several entries, each a short Chinese label followed by a URL. The school type column is the interesting design choice, because it encodes a classification the reader would otherwise have to know. Early entries are marked Top, and later ones use the Chinese Double First Class and 985/211 designations, split into A-class and B-class with the 211 tier noted. So the table sorts institutions by national tier, not alphabetically, which means a reader looking for a strong computer science programme has an implicit filter. The entries themselves are uneven in scope, and that unevenness is informative. Some rows are broad course sharing plans that cover a whole faculty, such as a computer science course sharing plan per university. Others are narrow and personal: a single course, a single exam, a single set of notes. Named examples across the visible rows include a machine learning course from an economics and management school, a computer architecture course, a signal processing course, a computer vision course with six lab sessions, a programming aptitude test solution set in plain C, a final exam collection for a database course, and quantum optics and advanced linear algebra final exams. Survival handbooks appear too, in the literal sense: a student survival manual for an information security programme and an updated 2008 edition of a university survival handbook. Those are not course materials at all, and their presence tells you what students actually look for at exam time, which is often the practical answer rather than the syllabus one. ## What the links are, including the awkward ones Following the link shapes tells you how this ecosystem stores course notes, and it is not a single tidy repository. Some links are dedicated per-school repositories with their own documentation sites, which is the most durable category in the table. Others are Read the Docs builds. Others are course pages on university domains, such as a listing of courses taught by a retired professor, which is a personal page that survives only as long as the department keeps it. A striking number are mirrors. The same computer science course sharing plan appears on GitHub, on its own documentation domain, and on Gitee, a separate git host widely used in China, sometimes under a different account. One entry points at a specific tree path inside a fork with a name like revert-61-master, which is a branch created to undo a commit rather than a repository anybody intends to maintain. The curation reality shows in the naming too. One entry links to an exam repository with a profane name, which is a reminder that student-run collections are named by students at odd hours rather than by a documentation team. And at least one cell contains a link to a user profile with a repositories tab rather than to a repository. That is the weakest link form in the entire table, because a profile URL resolves as long as the account exists and tells you nothing about whether the content is still there. For a user, the practical consequence is that some fraction of this table is a dead end, and nothing in the repository tells you which fraction. ## Contributing, and the two ways to do it The README invites contributions in two ways, which tells you something about the project's scale. The first is a direct pull request against the README file, with an explicit link to the edit page for it. For a table like this, that is the right mechanism: the change is one line, the diff is readable, and review is a matter of checking that the link resolves. The second is a form, hosted on an external site, for submitting additional resource links from other universities. Two contribution paths for one job usually means the pull request path is for small edits and the form is for people who would rather not touch git at all. The README also asks for a star, in the tradition of Chinese open source community projects, and carries a motto about taking and using resources in moderation and sharing knowledge. There is a channel link for discussion as well. What is missing is any standard for what a good entry looks like. There is no stated preference between a dedicated repository and a personal page, no guidance on whether a resource should be mirrored or linked, and no convention for recording a date so a reader can tell a current course from a 2019 one. Several entries do include a year in the label, which points to the convention existing informally rather than by rule. For a project whose failure mode is rot rather than incorrectness, that missing standard is the interesting gap. A schema, even three fields of institution, resource type and date, would make the table filterable and reviewable. ## Reading the table: three entries in detail Reading a few rows closely shows both the value and the limits better than describing the format. The first entry, Tsinghua University, is marked Top and points at a computer science course strategy repository, a second strategy repository, its Read the Docs build, and a separate postgraduate entrance exam strategy. Three of the four links describe the same subject area from different angles, which tells you the entry is a summary of what exists for one department rather than one resource. The second entry, Peking University, is longer still and mixes levels: a general course resource repository with its own documentation site, a computer science final project collection, an introduction to computer science course, a programming internship course in its C++ edition, a retired professor's personal course listing on a university domain, and a computer architecture course. A personal page for a retired professor is the most fragile link form in the table, because it depends on a department keeping one person's page online. The third visible row, Zhejiang University, shows the widest spread in a single cell: a course sharing plan, a mirror, a gitee copy of the mirror on a branch path, a documentation build, an architecture course, software engineering review notes, a knowledge base for computer science and geospatial science, an electronics course, a signal processing course, an optics course, plain C aptitude test solutions, a multimedia security course, a control engineering course, an electrical engineering course, a mixed class course collection, a computer vision course with lab sessions, and two more general collections. One cell with sixteen entries in it is where scanning stops working. Counting is a reasonable proxy for coverage: the first row carries four links, the second six, the third more than fifteen. A reader looking for a signal processing course has to read the whole cell to find it, which is exactly the cost that a topic index would remove. The pattern across rows is consistent: a broad sharing plan at the top, then faculty or institute specific collections, then individual courses and exams. A reader who wants coverage should take the first link; a reader who wants one course should read down the cell. ## Terms, and what they do not cover The repository declares Apache 2.0, and that is where the legal detail stops being simple: it covers the file in front of you and nothing behind the links. It does not reach into the linked repositories, each of which carries its own terms or none at all, and its own judgement about what a student may publish from a course. That matters here more than in a project that links to documentation, because the linked material is exactly the sort of thing that sits in an ambiguous space: course slides authored by teaching staff, exam papers that may or may not be intended for circulation, and student notes whose authorship and permission are unclear. The index does not adjudicate any of that, and its own terms cannot. A reader who is systematic about this should treat each link as its own object with its own provenance, which is more work than the table invites but the only defensible position. The practical counterpoint is that most of this material is openly published by the people who made it, and an index that links rather than mirrors keeps the original attribution intact. Mirroring everything into one repository, which some indexes do, would raise exactly the questions the current design avoids. Nothing was committed to this repository after 2026-09-28. ## Where this index is the wrong tool, and what to use instead Three limits are structural rather than fixable by tidying. The first is that discovery is manual. There is no search, no filter, no topic index and no machine-readable format; the only way to find a machine learning course is to read down a table cell. A list of this size, organised by institution, is close to the limit at which scanning stops working and an index stops being worth maintaining by hand. The second is that quality is unassessable. There is no signal about whether a linked course strategy is current, complete, or accurate, and the only verification available is opening it. That is not a criticism of the maintainers, who are curating for themselves and their peers, but it does mean the index optimises for finding a starting point rather than for finding the right thing. The third is that it competes with the resources it points at. The dedicated per-school projects, several of which maintain their own documentation sites and contribute through pull requests, are the durable home for course material. Once a school has one, a central index adds a second, thinner layer that must be kept in step. Given those limits, the sensible use is as a discovery layer with a short half-life. Use it to find that a course sharing plan exists for a school you care about, then move to that project's own documentation and work from there, and expect to come back to the index only when you are looking for the next school rather than the next resource. ## Conclusion Adopt this index as a discovery starting point rather than as a resource. A table organised by university and by tier turns a vague search into a shortlist, and the per-school projects it points at are where the actual notes, exam archives and course strategies live. Do not treat the links as durable, because the repository contains no copy, no verification and no mirror of its own, so a dead link is indistinguishable from a bad one and there is no way to check the index as a whole without walking it. Verify three things. Open the target before you plan a semester around it, since entries include personal profile pages and forks on two different git hosts, both of which move or vanish without notice. Check the date in the resource title where one exists, since several entries name specific academic years such as 2019 to 2020 courses, and course structure changes. And treat the terms carefully, since the list's own licence says nothing about the linked repositories. If you want one course tracked properly, go straight to the per-school project such as the computer science course sharing plans it links, and start there rather than from the index. ## FAQ ### What does this university course resource repository contain? The top level holds a README, a License file and a CNAME. The README is a Markdown table listing universities with a school type classification and links to course resources, and the CNAME points the Pages site at a custom domain. ### What kinds of resources are linked in the index? A mix: faculty-wide course sharing plans, individual course notes, final exam archives, survival handbooks for students, and programming aptitude test solutions. Entries are often mirrored across GitHub, documentation sites and Gitee. ### How do I add a university or a resource link? Through a pull request against the README file, with an edit link provided in the README itself, or through a submission form hosted on an external site. The README asks contributors to add resources for other universities. ### Can I rely on the links staying working? Not indefinitely. The index contains no archive or verification, and some entries point at personal profile pages, forks with branch-specific paths, or university department pages, all of which can disappear without the index changing. ### What licence applies to the linked material? The index itself is under Apache 2.0, which covers the table and its curation. Each linked repository carries its own terms or none, and some of the notes, such as course slides and exam papers, sit in an ambiguous position that the index does not resolve. ## Sources - [Issues](https://github.com/nwuzmedoutlook/university/issues) - [License: Apache-2.0](https://github.com/nwuzmedoutlook/university/blob/main/LICENSE) - [nwuzmedoutlook/university on GitHub](https://github.com/nwuzmedoutlook/university) - [Project website](https://workhard.eu.org) - [README](https://github.com/nwuzmedoutlook/university/blob/main/README.md) --- Hysen Labs editorial analysis, written from the project's own repository and release notes. Cite the canonical page: https://hysenlabs.com/projects/nwuzmedoutlook-university