This is the source for content for docs.timescale.com, starting with release 2.0. All documentation for previous versions is in the deprecated repository called docs.timescale.com-content.
The docs website site uses this repo as a submodule and converts the files directly into pages using a bash script and markdown parser.
All files are written in standard markdown.
We welcome and appreciate any help the community can provide to make TimescaleDB's documentation better!
You can help either by opening an issue with any suggestions or bug reports, or by forking this repository, making your own contribution, and submitting a pull request.
Before we accept any contributions, Timescale contributors need to sign the Contributor License Agreement (CLA). By signing a CLA, we can ensure that the community is free and confident in its ability to use your contributions.
There is a version of the docs for each supported version of the database, stored in
a separate git branch. Our docs site parses those branches to allow users to choose
what version of the docs they want to see. When submitting pull requests, you should determine
what versions of the docs your changes will apply to and attach a label to the pull request
that denotes the earliest version that your changes should apply to (2.0
, 2.1
, etc.)
The admin for the docs will use that as a guide when updating version branches.
Once a PR has been started for any branch, a GitHub action will attach a unique URL to preview your changes and allow others to more effectively review your updates. The URL will be added as a comment to your PR.
Any new commits to that PR will rebuild and be visible at the same URL.
Each major section that is incorporated into docs.timescale.com has a navigation
hierarchy governed by the appropriate page-index.js
file and underlying
directory structure. Depending on the content you are adding or editing, you may
have to modify the page-index.js
file that exists for that specific sub-project.
For instance, if you are adding a new function to the 'API Reference' section,
then you need to modify the page-index.js
file inside of api-reference
folder,
specifically at api-reference/page-index/page-index.js
.
Most content is then held within the parent folder as defined by the hierarchy. For
instance, the Overview section is a parent to a number of other pages and
folders. Therefore, there is a folder in the repo for overview
with more Markdown
files and folder inside as the hierarchy dictates.
As a general rule, every folder must have an index.md
file inside which will
have the content and layout for the "landing page" of that parent folder.
The format of this file is generally self explanatory, but the following rules should be understood to ensure you get the results you expect.
-
href: The minimum detail for any node in the tree is the
href
element. This is used as both the name of the Markdown file (eg.example-file.md
) to utilize for this URL and as the text in the tree (Camel Case and all hyphens replaced by spaces). -
title: If the navigation tree text should be anything other than the
href
as described above, add atitle
tag provide the exact text you would like to appear in the tree (and browserTitle
area). -
pageComponents: Any page that has children pages can list the children titles in a "card" or "list" style at the bottom of any other content for the parent page. To direct how a parent page should format links to the child object, use a value of either
['featured-cards']
or['content-list']
-
children: An array of additional
title
/href
objects, assumed to be inside of an on-disk folder with the same name as the parent. -
type: In some special cases, a tree element may have a special
type
associated with it. This is rarely needed in day-to-day documentation updates, but when specific functionality is required, it may be necessary to inquire about other page types. Currentlypage
anddirectory
the the two magjor supported types in documentation. -
excerpt: For a few select nodes in a tree, the
excerpt
may be used to create navigation "cards" on high-level landing pages. It is only intended to be used for the highest level pages at the current time.
Example
href: "overview",
pageComponents: ['featured-cards'],
children: [
{
title: "What is time-series data?",
href: "what-is-time-series-data"
},
{
href: "core-concepts",
children : [
{
title: "Hypertables & Chunks",
href: "hypertables-and-chunks"
},
{
href: "scaling"
},
...
]
}
]
In this example:
-
overview
is a parent and has no unique navigation title. If you look at the tree under the TimescaleDB section in a browser, you will see the output is Overview (Camel Case) -
Overview will have a visual indicator in the browser that it has children, in this case at least What is time-series data? and Core Concepts (among others)
-
The content for the Overview page will be displayed first, and then all
title
elements for the children will be displayed as "cards" below any content because of the pageComponents setting -
Two of the examples above display a navigation and page title text that is different from the name of their source Markdown files. For instance, the content for Hypertables & Chunks is found in the Markdown file
hypertables-and-chunks.md
.
None of the internal page links within these files will work on GitHub inside of the raw Markdown files that are being reviewed. Instead, the review link discussed above should be utilized for verifying all internal links.
All external links should work.
By default, any H2 or H3 heading ('##' and '###' respectively in Markdown) will have an anchor tag generated automatically.
If you want to link to a specific part of the page elsewhere in your document, you
need to use special anchor Markdown next to your anchor text; eg. [](anchor_name)
.
Your anchor name must be unique in order for the highlight scrolling to work properly.
Command formatting
When showing commands being entered from a command line, do not include a character for the prompt. Do this:
some_command
instead of this:
$ some_command
or this:
> some_command
Otherwise the code highlighter may be disrupted.
Syntax highlighting
When using a code block, add the appropriate language identifier after the initial three backticks to provide the appropriate highlighting on the rendered documentation site.
Programming language samples aside, most code blocks will usually be one of:
bash
, sql
, json
.
To maintain consistency, please follow these general rules.
- Maintain text editor width for paragraphs at 80 characters. We ask you to do this to assist in reviewing documentation changes. When text is very wide, it is difficult to visually see where text has changed within a paragraph and keeping a narrow width on text assists in making PRs easier to review. Most editors such as Visual Studio Code have settings to do this visually.
- Most links should be reference-style links where the link address is at the
bottom of the page. The two exceptions are:
- Links within Tip/Warning Callouts. These must be inline links for now
- Links to anchors on the same page as the link itself.
- All functions, commands and standalone function arguments (ex.
SELECT
,time_bucket
) should be set as inline code within backticks ("`command`"). - Functions should not be written with parentheses unless the function is being written with arguments within the parentheses.
- "PostgreSQL" is the way to write the elephant database name, rather than "Postgres". "TimescaleDB" refers to the database, "Timescale" refers to the company.
- Use single quotes when referring to the object of a user interface action. For example: Click 'Get started' to proceed with the tutorial.
To create a callout around a paragraph of text, wrap it with the following custom React component tag. Reminder, any links within the callout text MUST have inline link styles.
The "type" can currently support a value of "tip" or "warning"
<highlight type="tip">
Callout text goes here...
Example link style would [look like this](https://link_to_something.com/)
</highlight>
There are some custom modifications to the markdown parser to allow for special formatting within the docs.
- Adding
sss
to the start of every list item in an ordered list will result in a switch to "steps" formatting which is used to denote instructional steps, as for a tutorial. - Adding
:FOOTER_LINK:
to the start of a paragraph(line) will format it as a "footer link". - Adding
x.y.z
anywhere in the text will be replaced by the version number of the branch. Ex.look at file foo-x.y.z
>>look at file foo-0.4.2
. - Adding
:pg_version:
to text displayed in an installation section (i.e. any page with a filename beginninginstallation-
) will display the PostgreSQL version number. This is primarily to be used for displayed filenames in install instructions that need to be modular based on the version. - Wrapping a link with
<tag type="download">[file link here](https://link-to-file.com)</tag>
will create a tag "bubble" wrapper around the link and append a 'download link' icon to the end of the link inline. - Designating functions
- Adding
<tag type="community">Community</tag>
next to a header (for example, in the api section) adds decorator text "Community". - Adding
<tag type="function">TSDB Function</tag>
next to a header (for example, in the api section) adds decorator text "TSDB Function".
- Adding
Make sure to include the space after the formatting command!
There is a specific format for the API section which consists of:
- Function name with empty parentheses (if function takes arguments). Ex.
add_dimension()
- A brief, specific description of the function
- Any warnings necessary
- Required Arguments
- A table with columns for "Name", "Type", and "Description"
- Optional Arguments
- A table with columns for "Name", "Type", and "Description"
- Any specific instructions about the arguments, including valid types
- Sample Usage
- One or two literal examples of the function being used to demonstrate argument syntax.
See the API file to get an idea.