After upgrading to OJS3 (now in 3.0.2) I started to have a lot emails from the users of our journals which were basically replies to notifications they had received. This is due to the fact that in many notifications the reply-to address is set automatically to site contactEmail. This is the case for example when a new issue is published.
Also, in many cases the sender of the notification is also set as the site admin account, instead of the journal user accounts.
Some operations are considered journal-specific – most of them, in fact – but some are considered site-level, like password resets. Notifications are a bit of an odd case, because the same mailing code is responsible for a large number of different types of notifications. Given that you’re running a fairly typical service provider type of service, I’d be interested in more information on what your expectations are. Is the problem that you (as the site contact) are receiving messages that would be better directed to the journal contact? Are the problem messages specifically notification messages?
Public Knowledge Project Team
I think that the most cases where journal users have contacted me while meaning to contact someone in the journal have been the cases where a new issue has been published.
This notification uses the default notification template ($this->getMailTemplate('NOTIFICATION')) but it is a message I think should come from a representative of the journal.
In this case both the sender and the reply-to address is site specific, not context specific, and at least in my opinion they should be context specific.
Another case (which I have the check tomorrow) is that many of notifications going to for example editors have a reply-to address of the site contact. Here as well I think that the reply-to address should be the context main contact.
I keep on getting a lot of emails that are supposed to go the journal editors.
The most recent one was a respond to a discussion between the editor and the author. The author responded to a notification about an update in a discussion. The message is signed by the editor, the sender address is the editor’s address, but still the respond comes to my email. This does not make much sense.
I actually started to think that besides password changes, what are the other notifications that need to be site-level? And why can’t even the password changes be signed by the journal’s technical contact?
Make certain kinds of emails use a no-reply address and contain a “do not respond” note. (Discussion notifications are good candidates for this, because we want to strongly discourage out-of-channel conversations, as they break the audit trail.)
Make certain kinds of emails come from the site contact when outside of a journal context, and from the journal contact when inside a journal context. Password reset and notification emails might be good candidates here.
Public Knowledge Project Team
Regarding 2, the sender is really not a problem here. The sender address is usually what you would expect (see OJS3: notifications, sender and reply-to address - #4 by ajnyga). The problem is the reply-to address that is in many cases the site main contact address. The problem with the reply-to address is that you only see it when you start to answer an email and usually you do not notice that it is actually different than the sending email.
Regarding 1, I do think it is a good idea to have the “do not respond” tag there. Should that be in all notifications (meaning that the change would apply to the NOTIFICATION template) or should it only be in the case of discussions? However, if you are suggesting using a no-reply address like “email@example.com” then I think that this is something that is usually discouraged?
Is this really not a problem with other multijournals installs using OJS3? Today alone I have received seven emails as responses to notification messages that I have to forward to the correct editor. One of the responders noticed that the reply-to address differed from the senders address (the editor) and got scared that this was some sort of security issue.
On balance I think your proposed solution (to use a context-level reply-to when available) makes the most sense and is easiest to explain. Would you mind opening a pull request from your branch above if it’s ready?
Public Knowledge Project Team
Hi everyone, I am trying to find the most recent thread that discusses the problem with email notifications automatically being sent out to all users for our journal whenever a manuscript is published despite us having unchecked every box in the Notifications Tab for each user. I’m wondering if this issue has been resolved in any way? I see that there has been considerable talk about it for at least the last year?
OJS 3.1.0 introduced a change that allows you to choosen whether you want to send the “new issue” notifications
OJS 220.127.116.11 (latest release) introduced a new user setting that allows a user to opt-out of new issue and/or new announcement notifications.
Thank you for the reply.
We have recently updated to 18.104.22.168 and unfortunately the problem persists. In the attached photo you’ll see what settings we have the notifications on for each user; however, we just uploaded another manuscript and every user received this email: **A new article, “Quinze nouvelles especes d’oiseaux observées en Guadeloupe (F.W.I.),” has been submitted.**
The problem that we are having is that every one of of our USERS is receiving notification emails even though we have un-checked all of the notification boxes in each of their user profile settings. And so we are thinking that there might be a more fundamental problem here that we are not understanding.
In the Settings > Workflow > Emails tab, I am noticing that we do not have any control over the Notification Email options. Please see the screen shot that I have attached below. You will see that the on/off check box is a light gray color that we cannot change.
Any additional help would be very much appreciated. Thank you!
So are you saying that for example users that only have for example the Author role are receiving notification emails when someone else sends a new manuscript? Can you double check that these users do not have any other roles in the system. Because that is not supposed to happen and have not seen anything similar before.
Great suggestions. I think we’ve found the problem. We were having our users change their email notification preferences, which we thought would be enough, but indeed we had to go into their user “roles” to uncheck the Journal Manager option, to which it seemed that many of them were defaulted to. Probably a remnant of the upgrade to our new OJS version. Easy fix now!!!