You'll need to log into the website or join GitHub to contribute.
The easiest way to suggest improvements is to open an Issue. Here are a few tips to follow:
-
Refer to entities in the site using the urls to their detail page where possible. This conveys the id for the entities that a developer will need to process your request. Examples:
Referring to the type of entity and its id directly is OK too. Ids are easy to mix up, so it wouldn't hurt to write out your propositions in human language, too.
-
Include citations in peer-reviewed sources to support assertions when possible. If none can be found, a short proof may suffice. If an assertion is unclear, reviewers may decide to defer adoption of the submission until a reference can be found.
-
Be mindful of scope when creating an issue. A large multipart issue that is not well-focused will be more difficult to take longer to resolve.
You may want to submit separate issues to organize your requests better. Examples of good focus would be- for a single example, suggest one (or many) resolution to the example's properties
- for a single property, suggest one (or many) examples and their statuses for the property
- suggest a new theorem ("logic") that could be used to make deductions
-
(When suggesting an example) mention what the example contributes that is new to the database. Usually an example is added to make a particular distinction between properties. If there are already other examples
very similar to the proposed one, it might not make the cut. Sometimes an example is interesting enough in its own right to add, without a particular use. -
(When suggesting a new property) support its relevance with peer-reviewed citations. If you can make the case that it interests algebraists, then you'll have good chances at inclusion of the material.
After discussion of your suggestion, any contributor can make a Pull Request (see below) to implement your improvement.
To add a pull request implementing any changes to the repository, fork this project to your own account, make whatever changes you'd like, then open up all pull request for review.
Our reviewers will
- check correctness
- evaluate quality of supporting citations
- make suggestions for improvements or corrections
- ensure the sumbission's usage of terminology is consistent with the conventions being used
- (when necessary) run post-processing to maximize the deductions
Once the PR is approved, a developer can go about integrating it into the master branch.
This is the least flexible method, but if you really won't use git, then you can log into the website and send a message via the suggestion form.
Not all suggestions can be approved. There are an infinitude of conditions and examples out there, but we need to balance quantity and quality to try to create the best experience. Deliberations in a GitHub issue should suffice for deciding unusual cases, and ultimately it's up to the discretion of the maintainers.
Ideally, contributions should reference a peer-reviewed publication.
Sometimes an online resource like the stacks project, mathoverflow or math.stackexchange post
will also suffice.
Submitting bibtex for references is ideal.
The google scholar chrome extension
can produce bibtex entries on demand, although occasionally it messes them up or has incorrect information.
These are our current mathematical reviewers.
- Ryan Schwiebert, @rschwiebert
- Jose Brox
Please feel free to contact us with any questions. If you are interested in becoming a reviewer, please make contact.
Data in this repository - including any merged user contributions - is licensed under CC-BY-4.0. Please see LICENSE.md for details.