* More content cleanup
* Add missing TOC entries for CLI commands (I know there is planned work to change this outlined at https://github.com/pulumi/docs/issues/1491
* Apply get-started-note shortcode to other pages
Update intro to Getting Started index page
Signed-off-by: bermudezmt <bermudez.mt@gmail.com>
Add link to Pulumi platform GH repo
* Address feedback
Clean up duplicate TOCs in tutorials, check for consistency
Signed-off-by: bermudezmt <bermudez.mt@gmail.com>
* Move CI/CD docs under console/continuous-delivery
* Rename CI/CD docs
* Update references to new page locations
* Add CI/CD items to left-nav
* Migrate rest of Service docs
* Update a few more broken links
* Address PR feedback
* Normalize several cloud-vendor SVGs (these are now stored in Figma as well)
* Add an OpenStack SVG
* Use these logos on the docs home, get-started home, and supported-clouds pages
* Add a hover effect to the logos, to indicate interactivity
* Adjust the border and hover colors on the get-started page, to match those on the home page
* Remove unused PNGs
* Remove an unused partial
Signed-off-by: Christian Nunciato <c@nunciato.org>
* Add get started callout
First pass at informational note styling
Highlight Pulumi in headings and text for SEO purposes
Clean up home page
* Make a few little style tweaks
Lightens the background, adds a pixel border and uses a blue tint and underline for the link color as well.
Signed-off-by: Christian Nunciato <c@nunciato.org>
* Address feedback
Apply get-started-note shortcode to installation and tutorial index pages
Remove duplicate table of contents from Kubernetes tutorial
Signed-off-by: bermudezmt <bermudez.mt@gmail.com>
* Incorporate feedback and edits
Signed-off-by: bermudezmt <bermudez.mt@gmail.com>
* Clean up styles and add a warning note
Revert style change on single docs page layout
Address 1473
Signed-off-by: bermudezmt <bermudez.mt@gmail.com>
* Remove note around skipping version checks
* Clarify alternative installation options for all operating systems
Signed-off-by: bermudezmt <bermudez.mt@gmail.com>
This change prefers meta_desc over summary for controlling meta
description tags.
It also undoes the check for a missing meta description in the
prior commit. Since we unconditionally use Hugo's auto content
summarization via .Summary, it doesn't do anything meaningful.
We may want to revisit this as part of pulumi/docs#1475.
This change fixes up some titles and meta descriptions, in hopes
that it'll have a positive impact on our SEO (and SERP summary).
- Ensure all pages have titles and meaningful meta descriptions.
- Error out if pages lack a meta description. Also add the ability
to write `nosummary: true` if it's intentionally omitted.
- I'd like to error out if pages lack a title too. Unfortunately,
the Python API docs are missing titles (pulumi/docs#1127), so this
check is not currently enabled (easy to uncomment out later).
I also made the possibly-controversial change to stop emitting an
automatic `- Pulumi` at the end of page titles. This
- Leads to illy things -- like `About Pulumi - Pulumi`.
- Might make it more difficult for Google to extract a good title for
our SERP summary; is that why "Documentation" is currently
awkwardly labeled simply "Pulumi" on Google at the moment?
- Probably has a net negative effect on things like blog articles,
which we would like to index naturally based on the content.
I propose that anywhere we want Pulumi in the title, we manually add it.
- Indent H3 items
- Provide a way to opt-out of showing "On This Page" and use it on some pages
- Only show "On This Page" if there are items in the list
- Added left padding to the right nav (to match the right padding on the left nav)
- General style and code cleanup
- Delete `mini-toc.html` partial
Removed the "LEARN" button on the docs home page as there's no longer a need for it. Previously, it was a way to get to the Reference page to see the expanded Reference TOC, but now that we're expanding by default, there's little need for it. And it never felt quite right having it named "LEARN" but linking to "Reference". Now we have the single CTA we want to "GET STARTED".
The `{{ else if ne .Summary "" }}` check is always evaluating to true (probably because `Summary` is typed as `template.HTML` instead of a `string` in Hugo's source), so we go down this path, but when then using `.Summary`, the value is empty. This results in many pages having empty `<meta name="description">` instead of falling back to using the default description. The fix is to check `{{ else if .Summary }}` which will only evaluate to true if `.Summary` is set and is a non-zero value.
While making changes here, I cleaned up how we access other params throughout. We'll often use `if isset .Params "foo"` to check if the `foo` param is set, but this only checks if the param is set -- it doesn't check if the param actually has a value. The fix is to just use `if .Params.foo`, which will only be true if `foo` is set and is not: `false` (boolean), `0` (integer), or a zero-length array, slice, map, or string.
This change removes the `uppercase` text transformation currently applied to `h6` elements, and applies it instead to the instances that call for it.
Signed-off-by: Christian Nunciato <c@nunciato.org>
* Do not use "latest" GKE engine in examples.
Replicating changes from https://github.com/pulumi/examples/pull/343.
* Fix gcp.container.Cluster version for more occurrences.
We currently generate RSS feeds for all sections/taxonomies, which means we're generating and publishing a large number of `rss.xml` files. This change makes it so we only generate a single RSS feed for the blog at `/blog/rss.xml`.
I noticed our broken link checker was checking the home page and then would complete without checking the rest of the site. It turns out our recent addition of adding a `robots.txt` with `Disallow: *` for non-production builds was preventing the checker from checking links throughout the rest of the site. This change passes the `--follow` flag to the checker, to force it to follow robot exclusions.
It also includes some additional links to exclude, and fixes a couple blog posts to avoid broken images on summary pages.
