This blog site will be based on the same code base than https://github.com/adobe/business-website. We'll need to decide a time where we copy / paste the code and start diverging the 2 code bases.

This issue acts as a container to reference all business site issues that are "blocking" this copy / paste, i.e. some todos that need to be done before we start diverging the 2 code bases (to avoid duplication of the work).

Note that it is given that after the copy / paste, we'll necessary have to do fixes in the 2 code bases, the idea here is the minimise the effort.

• adobe/business-website#87
• adobe/business-website#80
• adobe/business-website#79
• adobe/business-website#72
• adobe/business-website#63
• adobe/business-website#30

"filtering" includes:

Products / industries filter UI on the topic pages.
Products and industries lists are coming from the taxonomy document.

Sample page: https://blog.adobe.com/en/topics/creativity.html

We probably simply need to add a "filters" option to the article feed that will show the UI and update feed query filters.

Current code is here: https://github.com/adobe/theblog/blob/master/scripts/v2/filters.js

This depends on #2

• add article feed

getSocial disappeared during the migration. We need to port it back.

See https://github.com/adobe/theblog/blob/master/scripts/v2/post.js#L603-L621

Check page: https://main--blog--adobe.hlx3.page/ko/publish/2021/10/21/digital-transformation-determining-the-future-of-healthcare

The author links points to https://main--blog--adobe.hlx3.page/ko/authors/------.

2 issues here:

1. During the content import, the "author" is converted into a meta property. But in the source document, the author can be a link: this is how we solved the problem when the author link cannot be deducted from author name (especially true in KO and JP or for authors with special characters in the name). The link gets lost during the import (I have added now the link on the example page to observe the next issue).
2. If we change the author property in the metadata table to be a link, the link is removing during rendering.

Several options:

1. Add support for a link in the author metadata property in Word - the link should then be outputted as another meta because the author meta tag does not support it. There is no standard or OG representing it. We should invent one (author:url ? like in og:image:url). The pipeline should then be extended.
2. Split the 2, have another row with author:link or authorLink (no necessarily user friendly).
3. Create a new block for it (really complicated).

In all cases, the importer must be adjusted.

@fkakatie @davidnuescheler @rofe WDYT ?

Check https://main--blog--adobe.hlx3.page/en/authors/paul-robson

We should show the twitter / linkedin ... icons and make the look better.

The date are not shown at all on the article cards.

If you compare date format between the EN site (https://blog.adobe.com/) and the JP site (https://blog.adobe.com/jp/), the date format is different.

Here is the missing function: https://github.com/adobe/theblog/blob/master/scripts/common.js#L236-L248

After import, the links in the Recommended Articles block are on the same line. Not really user friendly.

Please include URLs and screenshots (drag-and-drop on to this text box) where applicable in your description.

Just dev console on https://main--blog--adobe.hlx3.page/en/publish/2021/10/26/new-creative-cloud-releases-enable-creative-collaboration-drive-innovation-empower-creative-careers

The sitemap must be configured for all locales.

• en
• ko
• jp
• it
• fr
• es
• de
• br

@dominique-pfister Could you help here ?

Open https://main--blog--adobe.hlx.live/br/topics/news
No article is flagged with the "Novidades" topic. The page shows no result but also no message that says "No result".

Open https://main--blog--adobe.hlx3.page/en/publish/2021/04/22/photoshop-and-adobe-experience-manager-accelerating-creativity-through-apis

Check the Javascript error:

Auto Blocking failed TypeError: Cannot read properties of null (reading 'insertBefore')
    at scripts.js:537
    at Array.forEach (<anonymous>)
    at buildImageBlocks (scripts.js:532)
    at buildAutoBlocks (scripts.js:654)
    at decorateMain (scripts.js:975)
    at loadEager (scripts.js:1081)
    at decoratePage (scripts.js:1158)
    at scripts.js:1163

This is cause by the 2 side-by-side images (the phone screenshots) in the source document. They do not end-up in a block.

This was somehow supported on the hl2 version: https://blog.adobe.com/en/publish/2021/04/22/photoshop-and-adobe-experience-manager-accelerating-creativity-through-apis.html:

At least, there should not be an error.

Everyday, the content is auto published. This requires some configuration in the query-index AND the new coming "crontab" feature https://github.com/adobe/helix-slack-bot/issues/249.
Publication times must be aligned with the current ones.

• en
• ko
• jp
• it
• fr
• es
• de
• br

@dominique-pfister Can you help here too ?

We only get 99 to LH because of TBT: https://developers.google.com/speed/pagespeed/insights/?url=https%3A%2F%2Fblog.adobe.com%2F

Matt would like Title and Description to be "mandatory" metadata properties. In order to influence authors who will copy/paste content from older articles, we should add the Title and Description to the metadata in all the cases.

Open https://main--blog--adobe.hlx3.page/es/latam
No articles.
But https://main--blog--adobe.hlx3.page/es/publish/2021/02/18/5-ideas-creativas-con-photoshop is tagged with LATAM. It should be visible on the latam page.

https://main--blog--adobe.hlx3.page/en/publish/2021/10/13/unleashing-limitless-creativity-see-the-world-differently-with-immersive-tech

Add a custom plugin to the sidekick so authors can generate, upload, and publish a new atom feed

Currently used to debugged and track invalid content during content migration.

Different type of missing captions:

• simple image caption: https://blog.adobe.com/en/publish/2021/10/26/new-creative-cloud-releases-enable-creative-collaboration-drive-innovation-empower-creative-careers
• double image caption: https://blog.adobe.com/jp/publish/2021/10/15/cc-stock-collection-review02

The gnav and the region picker links currently point to blog.adobe.com which makes it impossible to stay on the same host and navigate the site.
The links must be adjusted when not on the production environment.

Seen on page:
https://main--blog--adobe.hlx3.page/en/publish/2021/09/15/adobe-for-all-do-one-thing-today

The current Analytics code in the repo is the one from the business website. We need to take theblog analytics code (which is currently evolving) and copy it in this repo.

@fkakatie follow up from #13.
Weird border (I assume it is the adobe logo in the background)

What needs to happen here is not fully clear to me - I assume Go to Fastly and edit "blog.adobe.com" service backend url: change from theblog--adobe.hlx.page to main--blog--adobe.hlx3.page.

@tripodsan @stefan-guggisberg @trieloff @rofe @davidnuescheler : is this really the only thing needed ?

This depends also from #49.

Follow up from: #10

At the moment, the category of an article is the first tag of the list.
Problem: an author can specify a tag defined in the taxonomy document as non visible. Consequence: the tag should never appear on the site. The category is usually part of the LCP block, thus we cannot check early enough thus no client side check is performed.

This happens today in the content quite a lot and this is handle client side on the current site. Several options:

1. post LCP and once the taxonomy is loaded, we replace the category by the correct one
2. we load the taxonomy before LCP thus we can check (might have an impact on the score)
3. we do nothing client side (we can still fix the imported content). The tagger is now always moving a visible category first but if an author insists a non-visible category may appear. The link is set to # if this is the case.

With the hlx3 migration, we are getting rid of the .html in the urls (see #7). Sitemap will be re-generated and will not contain the extensions (see #46).

Once the site will be migrated and the sitemaps will be regenerated, we will need to re-submit them to Google.
We need to find who has access to do so.

Pre-go-live

• Adapt monitoring automation to new site

Go-live

• Disable monitors prior to go-live switch
• Run new monitoring automation
• Verify and enable updated monitors

Post-go-live:

• Update static homepage response

Current gnav on https://blog.adobe.com:

Our gnav on https://main--blog--adobe.hlx3.page/

• box size is smaller
• overall spacing is missing
• promo background should be white
• promo link should be left align
• mobile version looks broken

@auniverseaway any chance you could take a look at this one "quickly" ? Go-Live date for the blog is Nov 9.

Implement the author pages and add author link on article page.

The taxonomy is handled differently with business-website. We should migrate the taxonomy from theblog site.

• Visible vs non visible tags
• Show parent tags
• Exclude logic
• ...

This depends on #1.

List of recommendations for SEO purpose. For now, this is just the raw list, we might need to define the ones we don't want to do.

• Homepage changes
  • Add H1 title - done with hlx3 migration
  • Add sub text below the H1 title - styling / CSS missing
  • Use image alt text for card images - done during hlx3 migration (image alt text indexed)
• Topic page changes
  • Add sub text (155 characters short description) - styling / CSS missing
• Author page changes
  • Use H1 instead of H2 - will done with #53

Open https://main--blog--adobe.hlx3.page/ and try to search for "photoshop".

Results show "UNDEFINED" as a category

See page: https://main--blog--adobe.hlx3.page/ko/publish/2021/03/31/how-to-font-pair-basics-what-to-do-and-not-to-do

Current link to articles in content, especially in promotions, featured / recommended articles contains .html. But requesting an article with a .html returns a 404.
While there is some thing that needs to happen globally, while we import the content, we can certainly remove the .html.
Also, we could add client side logic to remove it.

List of things to remember or todos for the content migration:

For each locale:

• promotion pages:
  • before import, generate a promotion list for conversion to banner
  • copy all promotions docx
  • lower case folder names and filenames
• author pages:
  • copy all
  • convert all to docx
• topic pages:
  • copy all
  • copy taxonomy.xlsx
  • transform - add "article feed" block tag config
• gnav
  • copy / paste existing HTML gnav into Word gnav
• footer
  • copy / paste existing HTML footer into Word footer
• articles
  • copy (overwrite .docx and remove .md) imported articles from temp sharepoint to theblog sharepoint
• manually sync index pages with latest content

Same than adobe/business-website#127
Must be integrated when the other one is fixed.

https://blog.adobe.com/en/uk.html
https://blog.adobe.com/en/apac.html

Homepage like.

Certainly some content adjustment required. Depends on #2.

The block seems to be used in several places (102 usages in total):

https://blog.adobe.com/ko/publish/2021/10/27/ocean-ar-coral-reefs-aero-smithsonian-thehydrous
https://blog.adobe.com/jp/publish/2021/10/28/cc-design-max-challenge-winners-2021

The current index config use the title metadata which leads to some unwanted behaviours (like displaying | Adobe in article cards - they use it in title metadata for social sharing). Indexing the H1 is more accurate.

cc @dominique-pfister @davidnuescheler

business was EN only. We should make sure the site works for other languages.
Pre-requisite: import some content.

In the current implementation, we load the full query-index.json. The EN site contains 13k articles, the file is huge (time to download and time to parse). We should load batch of 1000 articles.

@davidnuescheler said he will take care of the implementation.

• append the article-feed block
• convert h2 to h1 (SEO feedback)

Rough idea:

• modify the edge dict for adobe/blog to use adobe/theblog contentbus id - i.e. adobe/blog uses a temporary sharepoint but the final contentbus id
• preview / publish all content via adobe/blog - i.e. fill production S3

for the go-live:

• change the fstab of adobe/blog to point to the adobe/theblog sharepoint (theblog Site)
• switch the blog.adobe.com edge DNS to point to adobe/blog - production ready and live
• behind the scene, update sharepoint, i.e. replace old content by migrated content

cc @tripodsan

The gnav content has to be fully created for each locale: copy/paste from existing content.
Note: some locale like JP have a fully customised gnav.

Same applied for footer and placeholders.

• en
• ko
• jp
• it
• fr
• es
• de
• br

Theblog sidekick has plenty of features that need to be ported into the new code base. See https://github.com/adobe/theblog/blob/master/tools/sidekick/config.js

I assume this is the same than #8.
The problem occurs mainly on localhost but I assume it can happen on prod if everything goes super fast.

https://main--blog--adobe.hlx3.page/en/publish/2021/09/15/adobe-for-all-do-one-thing-today
http://localhost:3000/en/publish/2021/09/15/adobe-for-all-do-one-thing-today

On localhost, the Youtube video never loads.

Open https://main--blog--adobe.hlx3.page/en/publish/2021/04/01/36-days-of-type-2021#gs.fjtoin in Incognito window: gnav is missing.

This seems to be some css priority that is applied differently depending on the loading sequence.

https://main--blog--adobe.hlx3.page/en/publish/2021/10/26/new-creative-cloud-releases-enable-creative-collaboration-drive-innovation-empower-creative-careers

mp4 are missing.

https://main--blog--adobe.hlx3.page/en/publish/2021/10/26/new-creative-cloud-releases-enable-creative-collaboration-drive-innovation-empower-creative-careers

While image captions are displayed correctly, the others (video, animations, embed) are still italic text under the assets.