Decide approach to migrating "Author(s)" and "About the author" information

mlncn commented

2019-07-27 15:15:34 +00:00

(Migrated from gitlab.com)

Issue articles (story content type) have a multivalue Author(s) text field and a single value "About the author" rich textarea, and a separate "Author picture" field.

Instead of migrating these over to article, we could instead migrate them to a separate person/author content type, which would be referenced from articles.

The primary practical difference is that updating an author profile would update it globally, for all of their content.

The question is, would that be preferred?

Not to bias the question too much, but a non-hypothetical case is Ted Rau.

Example: http://www.geo.coop/story/if-were-all-equal-then-why-cant-we-agree

cc @gnuget @cedewey

Issue articles (`story` content type) have a multivalue Author(s) text field and a single value "About the author" rich textarea, and a separate "Author picture" field. Instead of migrating these over to article, we could instead migrate them to a separate person/author content type, which would be referenced from articles. The primary practical difference is that updating an author profile would update it globally, for all of their content. The question is, would that be preferred? Not to bias the question too much, but a non-hypothetical case is Ted Rau. Example: http://www.geo.coop/story/if-were-all-equal-then-why-cant-we-agree cc @gnuget @cedewey

mlncn commented

2019-07-27 15:38:41 +00:00

(Migrated from gitlab.com)

@gnuget we haven't migrated the regular contributor field, have we?

That could and should be replaced with the expanded author profile approach also.

EDIT: OK the field is actually just a reference to a do-not-migrate vocabulary; makes sense to leave it behind. See below.

@gnuget we haven't migrated the regular contributor field, have we? That could and should be replaced with the expanded author profile approach also. EDIT: OK the field is actually just a reference to a do-not-migrate vocabulary; makes sense to leave it behind. See below.

mlncn commented

2019-07-27 16:07:47 +00:00

(Migrated from gitlab.com)

Some context:

select COUNT(*) from field_data_field_author_picture;
+----------+
| COUNT(*) |
+----------+
|       15 |
+----------+

select COUNT(*) from field_revision_field_author_bio;
+----------+
| COUNT(*) |
+----------+
|      323 |
+----------+

Of the 323 field_revision_field_author_bio, 307 were unique, because eleven bio fields were reused two to four times each. Many of the remainder, of course, were almost identical (Yavor Tarinski "in Bulgaria" vs "from Bulgaria"), but that suggests that the about the authors should be historical and not from a potentially ever-updated biography, as does another one indicating GEO just learned of Graham's deportation.

Note: None of these happened to have any revisions, so that's an accurate count of content.

There are 685 pieces of content with authors, with a total of 770 authors listed (a piece of content can have multiple authors listed in separate text fields); out of those 770, there are 439 unique authors.

Queries used (won't show multiple authors but row counts are informative):

select COUNT(*) from field_data_field_author;
select COUNT(*) from field_data_field_author GROUP BY entity_id ORDER BY COUNT(*);
select COUNT(*), entity_id, revision_id, field_author_value from field_data_field_author GROUP BY field_author_value ORDER BY COUNT(*);

Make that 438 unique authors with trim() applied ;-)

Some context: ``` select COUNT(*) from field_data_field_author_picture; +----------+ | COUNT(*) | +----------+ | 15 | +----------+ select COUNT(*) from field_revision_field_author_bio; +----------+ | COUNT(*) | +----------+ | 323 | +----------+ ``` Of the 323 `field_revision_field_author_bio`, 307 were unique, because eleven bio fields were reused two to four times each. Many of the remainder, of course, were almost identical (Yavor Tarinski "in Bulgaria" vs "from Bulgaria"), but that suggests that the about the authors *should* be historical and not from a potentially ever-updated biography, as does another one indicating GEO just learned of Graham's deportation. Note: None of these happened to have any revisions, so that's an accurate count of content. There are 685 pieces of content with authors, with a total of 770 authors listed (a piece of content can have multiple authors listed in separate text fields); out of those 770, there are **439 unique authors**. Queries used (won't show multiple authors but row counts are informative): ``` select COUNT(*) from field_data_field_author; select COUNT(*) from field_data_field_author GROUP BY entity_id ORDER BY COUNT(*); select COUNT(*), entity_id, revision_id, field_author_value from field_data_field_author GROUP BY field_author_value ORDER BY COUNT(*); ``` Make that 438 unique authors with `trim()` applied ;-)

mlncn commented

2019-07-27 16:22:34 +00:00

(Migrated from gitlab.com)

There are 198 regular contributors cited across 195 pieces of content (select count(*) from field_data_taxonomy_vocabulary_14; select count(*) from field_data_taxonomy_vocabulary_14 GROUP BY entity_id;).

The regular contributors taxonomy was marked "Don't Migrate", but that's how attribution is shown on posts like http://geo.coop/node/2667 ... but that's a blog, OK, so can show attribution that way, and the other content seems to have redundant attribution of authorship. OK so that can stay dropped, but the motivation for referencing author profiles standard is stronger.

There are 198 regular contributors cited across 195 pieces of content (`select count(*) from field_data_taxonomy_vocabulary_14; select count(*) from field_data_taxonomy_vocabulary_14 GROUP BY entity_id;`). The regular contributors taxonomy [was marked](https://ethercalc.org/geo-migration-plan-13sdfsdfq34 ) "Don't Migrate", but that's how attribution is shown on posts like http://geo.coop/node/2667 ... but that's a blog, OK, so can show attribution that way, and the other content seems to have redundant attribution of authorship. OK so that can stay dropped, but the motivation for referencing author profiles standard is stronger.

mlncn commented

2019-07-27 18:57:37 +00:00

(Migrated from gitlab.com)

marked this issue as related to #5