Add a rough draft of ERD

[KevinKelly25]

Adds a ERD for the detail tables mentioned in issue #138

[smurthys]

Thanks @KevinKelly25 for getting this started.

A few things to consider:

• The title of the HTML document needs revision. It presently says "Copy of ClassDB_Detail_Log_20171230", which is obviously not relevant
• Entity name connectionActivity needs to begin with upper case
• Some of the points enumerated in an earlier comment still need consideration
• In general, I recommend looking through comments to Issue #138 to make sure all requirements are met (perhaps they are, but do check)

I wonder what people think about the naming convention for attribute names in tables: Thus far, ClassDB has made first-letter lower case, whereas Gradebook makes first-letter upper case. Which approach should we use?

I prefer first-letter upper case because that seems to be the industry norm, and it distinguishes attributes from variables in code.

[KevinKelly25]

Added corrections as discussed in today's meeting and issue #138 .

[afig]

Thanks for making the changes @KevinKelly25, they look great.

I have a few minor suggestions:

• The StatementSucceededAtUTC attribute in the DDLActivity entity should start with a lowercase letter to match the convention currently used elsewhere in the diagram and project. @smurthys made a comment earlier about starting a discussion related to this convention, but we have not yet committed to a decision.

• I believe that it would make more sense for the new file ERSchema.html to be renamed to ER-Schema.html. This follows the naming convention of other files in the docs folder, and matches the title located within the <title> tags in the HTML file.

• @smurthys proposed adding an attribution box similar to the one used in Gradebook's ER diagram. I agree that adding this information to the diagram would be beneficial.

---

In terms of naming conventions for attribute names in tables, I do not have a strong personal preference between the two major conventions. However, since first-letter-uppercase appears to be more common overall, then I agree that we should consider switching. If we agree to switch, then a separate issue should be opened.

[wildtayne]

Overall, the latest ER schema looks good.

One additional observation I have is that StatementSuceededAtUTC is not the best name for this attribute. Currently, we are using statement_timestamp to determine the timestamp for DDL activity. According to the Postgres docs, this returns the timestamp when the Postgres server received the current statement from the client. Perhaps StatementStartedAtUTC would be a better name.

I also agree that beginning attribute names with capital letters is a better convention, and that this should be addressed in a future issue.

[KevinKelly25]

Thank you @afig and @srrollo for the comments. I fixed the issues you brought up. I did leave the attributes in the current system we have, lower case first. However, I also agree that switching to what is the industry norm according to @smurthys is probably a good idea.

[smurthys]

Good changes @KevinKelly25

A few observations:

• Relationships need names
• The relationship cardinality on the side of ConnectionActivity should be 0..*: a user is not required to have a connection. The cardinality on the other side of this relationship should be 1..1.
• Partial ID attributes in weak entities should have a dashed underline, if that is possible in draw.io
• The HTML title should be ClassDB ER Schema, unless draw.io requires the title element's content to match the file name (I doubt that is the case)
• I recommend adding the text ClassDB ER Schema at the top of the diagram to make it appear like a "heading" (with sufficient space before and after the heading text)
  • The heading will be useful when the image version of the schema is embedded in the docs
• I recommend moving the attribution box to the bottom of the diagram with its left aligned with the left of entity User
• Names in the attribution box (and in general in any artifact) should reflect people who have made "meaningful contribution". IIRC, that should be Kelly, Rollo, and Murthy, but please feel free to correct me
  • Past contributors are familiar with attribution guidelines; new contributors are yet to be appraised

[smurthys]

I have added the Issue #140 about letter case. Please discuss that specific issue over there.

Because the ER Schema needs to be edited, it is OK if the issue is fixed in this schema. However, I am also OK if it is addressed in the next round (because we know this schema will soon be outdated). I'd prefer if this PR is approved and merged before this afternoon's call (but of course only without sacrificing quality).

[KevinKelly25]

Updated ERD with corrections for @smurthys comments and used new standard for attribute names discussed in issue #140

[KevinKelly25]

Updated the Attribute box to include authors with meaningful contributions as per the discussion in the meeting today.

[smurthys]

Good work @KevinKelly25.

After all reviewers have approved, please hit the "Merge pull request" button to merge your changes to master. I believe the changes should merge without any conflict.

Remember to delete your branch after the merge is complete and you have verified the file appears in the master branch.

[afig]

Also of note, when merging a pull request, there are three options, which can be seen by clicking the drop-down arrow next to the merge button. The three options are:

• Create a merge commit
• Squash and merge
• Rebase and merge

So far, we have been using the "Create a merge commit" option, and thus this Pull Request should also do the same. "Create a merge commit" should be the default options, but is worth checking. If you wish, you can read more about the differences between each method here.

[KevinKelly25]

Thank you, once approved by @baconbm I will do the create a merge commit