GitHub AE can display several common image formats, including PNG, JPG, GIF, PSD, and SVG. In addition to simply displaying them, there are several ways to compare differences between versions of those image formats.
- GitHub does not support comparing the differences between PSD files.
- If you are using the Firefox browser, SVGs on GitHub may not render.
You can directly browse and view images in your repository on GitHub AE:
SVGs don't currently support inline scripting or animation.
2-up is the default mode; it gives you a quick glimpse of both images. In addition, if the image has changed size between versions, the actual dimension change is displayed. This should make it very apparent when things are resized, such as when assets are upgraded to higher resolutions.
Swipe lets you view portions of your image side by side. Not sure if colors shifted between different versions? Drag the swipe slider over the area in question and compare the pixels for yourself.
Onion Skin really comes in handy when elements move around by small, hard to notice amounts. Did an icon shift two pixels to the left? Drag the opacity slider back a bit and notice if things move around.
GitHub AE can host and render 3D files with the .stl extension.
When looking directly at an STL file on GitHub AE you can:
- Click and drag to spin the model.
- Right click and drag to translate the view.
- Scroll to zoom in and out.
- Click the different view modes to change the view.
When looking at a commit or set of changes which includes an STL file, you'll be able to see a before and after diff of the file.
By default, you'll get a view where everything unchanged is in wireframe. Additions are colored in green, and removed parts are colored in red.
You can also select the Revision Slider option, which lets you use a slider at the top of the file to transition between the current and previous revisions.
If you see this icon in the corner of the viewer, then the WebGL technology is not available on your browser:
If your model is invalid, GitHub may not be able to display the file. In addition, files that are larger than 10 MB are too big for GitHub to display.
For example, if your model's URL is
github.com/skalnik/secret-bear-clip/blob/master/stl/clip.stl, your embed code would be:
By default, the embedded renderer is 420 pixels wide by 620 pixels high, but you can customize the output by passing height and width variables as parameters at the end of the URL, such as
ref can be a branch or the hash to an individual commit (like
GitHub supports rendering tabular data in the form of .csv (comma-separated) and .tsv (tab-separated) files.
When viewed, any .csv or .tsv file committed to a repository on GitHub AE automatically renders as an interactive table, complete with headers and row numbering. By default, we'll always assume the first row is your header row.
You can link to a particular row by clicking the row number, or select multiple rows by holding down the shift key. Just copy the URL and send it to a friend.
If you want to find a certain value in your dataset, you can start typing in the search bar directly above the file. The rows will filter automatically:
Occasionally, you may discover that your CSV or TSV file isn't rendering. In those instances, an error box appears at the bottom of your raw text, suggesting what the error may be.
Common errors include:
- Mismatched column counts. You must have the same number of separators in each row, even if the cell is blank
- Exceeding the file size. Our rendering only works for files up to 512KB. Anything bigger than that slows down the browser.
GitHub supports rendering of PDF documents.
Currently, links within PDFs are ignored.
Commits and pull requests that include prose documents have the ability to represent those documents with source and rendered views.
The source view shows the raw text that has been typed, while the rendered
view shows how that text would look once it's rendered on GitHub AE. For example,
this might be the difference between showing
**bold** in Markdown, and bold in the rendered view.
Prose rendering is supported for rendered documents supported by github/markup:
You can click to see the changes made to the document as part of a commit.
We provide a tooltip describing changes to attributes that, unlike words, would not otherwise be visible in the rendered document. For example, if a link URL changes from one website to another, we'd show a tooltip like this:
Commit comments can only be added to files within the source view, on a line-by-line basis.
As with other rendered prose documents, hovering over a header in your document creates a link icon. You can link readers of your rendered prose diff to specific sections.
Some pull requests involve a large number of changes with large, complex documents. When the changes take too long to analyze, GitHub AE can't always produce a rendered view of the changes. If this happens, you'll see an error message when you click the rendered button.
You can still use the source view to analyze and comment on changes.
We don't directly support rendered views of commits to HTML documents. Some formats, such as Markdown, let you embed arbitrary HTML in a document. When these documents are shown on GitHub AE, some of that embedded HTML can be shown in a preview, while some (like an embedded YouTube video) cannot.
In general, rendered views of changes to a document containing embedded HTML will show changes to the elements that are supported in GitHub AE's view of the document. Changes to documents containing embedded HTML should always be reviewed in both the rendered and source views for completeness.
GitHub AE supports rendering geoJSON and topoJSON map files within GitHub AE repositories. Simply commit the file as you would normally using a
.topojson extension. Files with a
.json extension are also supported, but only if
type is set to
topology. Then, navigate to the path of the geoJSON file on GitHub.com.
When you click the paper icon on the right, you'll also see the changes made to that file as part of a commit.
Maps on GitHub AE use Leaflet.js and support all the geometry types outlined in the geoJSON spec (Point, LineString, Polygon, MultiPoint, MultiLineString, MultiPolygon, and GeometryCollection). TopoJSON files should be type "Topology" and adhere to the topoJSON spec.
You can customize the way features are displayed, such as specifying a particular color or adding a descriptive icon, by passing additional metadata within the geoJSON object's properties. The options are:
marker-color- valid RGB hex color
marker-symbol- an icon ID from the Maki project or a single alphanumeric character (a-z or 0-9).
stroke- color of a polygon edge or line (RGB)
stroke-opacity- opacity of a polygon edge or line (0.0 - 1.0)
stroke-width- width of a polygon edge or line
fill- the color of the interior of a polygon (GRB)
fill-opacity- the opacity of the interior of a polygon (0.0-1.0)
See version 1.1.0 of the open simplestyle spec for more information.
For example, if your map's URL is github.com/benbalter/dc-wifi-social/blob/master/bars.geojson, your embed code would be:
By default, the embedded map 420px x 620px, but you can customize the output by passing height and width variables as parameters at the end, such as
ref can be a branch or the hash to an individual commit (like
If your map contains a large number of markers (roughly over 750), GitHub will automatically cluster nearby markers at higher zoom levels. Simply click the cluster or zoom in to see individual markers.
The underlying map data (street names, roads, etc.) are driven by OpenStreetMap, a collaborative project to create a free editable map of the world. If you notice something's not quite right, since it's open source, simply sign up and submit a fix.
If you're having trouble rendering geoJSON files, ensure you have a valid geoJSON file by running it through a geoJSON linter. If your points aren't appearing where you'd expect (e.g., in the middle of the ocean), it's likely that the data is in a projection which is currently unsupported. Currently, GitHub AE only supports the
Additionally, if your
.geojson file is especially large (over 10 MB), it is not possible to render within the browser. If that's the case, you'll generally see a message that looks something like this:
It may still be possible to render the data by converting the
.geojson file to TopoJSON, a compression format that, in some cases, can reduce filesize by up to 80%. Of course, you can always break the file into smaller chunks (such as by state or by year), and store the data as multiple files within the repository.
When you add Jupyter Notebook or IPython Notebook files with a .ipynb extension on your enterprise, they will render as static HTML files in your repository.
To view a fully interactive version of your Jupyter Notebook, you can set up a notebook server locally. For more information, see Jupyter's official documentation.
If you're having trouble rendering Jupyter Notebook files in static HTML, you can convert the file locally on the command line by using the
$ jupyter nbconvert --to html NOTEBOOK-NAME.ipynb