Reduce the number of new submission notification emails for editors, and simplify their management

[jmacgreg]

Right now it's possible for editors to receive three emails notifying them of a new submission in the system:

1. They can be BCC'd onto the submission_ack email to the author (and BCC’d editors, potentially) - see https://github.com/pkp/ojs/blob/ojs-stable-3_1_1/classes/submission/form/SubmissionSubmitStep4Form.inc.php#L85. They cannot control whether they receive this via notification settings.

2. They can be notified via a general new submission notification (NOTIFICATION_TYPE_SUBMISSION_SUBMITTED - see https://github.com/pkp/pkp-lib/blob/ojs-stable-3_1_1/classes/submission/form/PKPSubmissionSubmitStep4Form.inc.php#L125). They can control whether they receive this notification via notification settings, AFAIK.

3. a more specific notification, in event that no editor is auto-assigned (NOTIFICATION_TYPE_EDITOR_ASSIGNMENT_REQUIRED, see https://github.com/pkp/pkp-lib/blob/ojs-stable-3_1_1/classes/submission/form/PKPSubmissionSubmitStep4Form.inc.php#L133). They cannot control whether they receive this notification via notification settings.

1 and 3 are problematic, and the general existence of 2) and 3) which are so similar is also potentially confusing.

Suggestions:

• I'd suggest removing the BCC option from 1), and reserve it just for author informational/instructional purposes.

• I'd also suggest only ever sending out one of 2) and 3): 2) only if an editor has been assigned; and 3) only if no editor has been assigned.

• Finally: there should be a way to opt out of 3) via the user's notification settings. Either there should be two separate settings (one for each notification), or the one setting should be applied to both cases.

[asmecher]

@jmacgreg, the purpose of 1) is to notify section editors who are automatically assigned. It's pretty likely that none of the other cases will notify these editors; are you sure that should be removed? (I suspect this is less useful on smaller installs, but is probably pretty useful on larger ones.)

[asmecher]

(Also note that 1) and 3) are mutually exclusive.)

[asmecher]

@jmacgreg, your original post says 4. and 3) are problematic -- I think github's Markdown tried to renumber your list and made it less useful. Which 2 are you referring to?

[jmacgreg]

Yep, Markdown munged my comment - I was referring to 1 and 3. By "problematic" I'm referring to the fact that there's no option to opt out.

It's been a while since I've looked at this, but I think the larger issue (which is only hinted at here) is

• there is the appearance of being able to control some of the emails that are received from the system, eg. via the user's notification settings; but in the end, editors still receive emails which they can't disable
• some of the emails, in particular 2 and 3, may be sent at the same time, which can feel overwhelming.

[asmecher]

PRs:

• pkp/pkp-lib#3770 Reduce "new submission" notifications/emails #4269
• pkp/pkp-lib#3770 Reduce "new submission" notifications/emails ojs#2199

Here's what I've done...

• Left #1 in place (BCC on SUBMISSION_ACK), at least for now. This is the one place that new submission messages can be sent to non-users of OJS, like mailing lists. We can revisit this for 3.2.
• Made #2 (NOTIFICATION_TYPE_SUBMISSION_SUBMITTED) go only to users who have been automatically assigned
• Made #3 (NOTIFICATION_TYPE_EDITOR_ASSIGNMENT_REQUIRED) only go out when nobody was automatically assigned (i.e. mutually exclusive to #2), and added configuration options for it to the user's notification settings.

There look to be a lot of parts of the notification system that are dead code at the moment (e.g. the whole NOTIFICATION_TYPE_EDITOR_ASSIGNMENT_... set of constants, at a glance). I'll file that elsewhere for 3.2.

Meanwhile, @jmacgreg, could you test this for me (once the tests pass)?

[asmecher]

(Filed an issue for the investigation of the nonfunctional notification code: #4270)

[jmacgreg]

Absolutely!

[ajnyga]

Another perspective on #1. We have had journal editors asking why some of their editors are receiving "Thank you for your submission" messages from the system. The reason is of course that the editors are section editors in specific sections and receive a bcc of the author email.

I know that new email templates is something to avoid, but just sending a copy of email that is meant to other user / role is a bad practice in my opinion and creates confusion.

There should be a "New submission in your section" template that you could turn of if you wanted.