Web Client: Serials - Predict New Issues Ignores Entered Values

[John Amundson]

Tested on Evergreen 3.0.3 - Web Client.

When Predict New Issues is used and issues exist for the subscription, any values entered into the dialog box are ignored, and the new issues predict based on the latest created issue.

For example, if your last issue is v.9 no.2 (2018 Feb), and you select Predict New Issues and enter v.15 no.5 and 2018 May, and enter prediction count of 2, the two issues predicted will be v.9 no.3 (2018 Mar) and v.9 no.4 (2018 Apr).

Currently, the only workaround for this seems to be to edit the holdings code of the latest issue and then predict new issues off of that.

This can lead to some frustrating workflows, especially when a prediction pattern is changed. More information on this bug:https://bugs.launchpad.net/evergreen/+bug/1745427

Predict New Issues should not ignore entered values.

This behavior can also be seen when using "Add Following Issue"

[Beth Willis (willis-a)]

Still a problem on EG 3-2-4

[Lynn Kauffman (aploregon)]

Affects workflow.

[Erica Rohlfs (erohlfs)]

confirming in version 3.7

[Irene Patrick (iepatrick)]

I appear to be experiencing this problem on version 3.7.2. In my case, it involves a pattern change. The publication in question switched from bimonthly to quarterly. We deactivated the old pattern, created a new active pattern, and predicted the issues based on the new pattern. However, instead of using the initial values we entered into the Predict New Issues dialog, it ignored them and predicted the next issues based on the ones that had been already received. (It predicted v.45:no.7, etc., instead of predicting v.46:no.1, etc.)

We were able to use the "Edit issue holding codes" option to correct the enumeration and chronology to show what they should have been, and prediction of subsequent issues seems to be working. But it was unexpected to encounter this problem in the web interface, since we never saw this behavior in XUL.

I have added the description about the pattern change to this bug, since it seems to apply here more than it does to bug #1797452. That bug describes the Predict New Issues option failing with a new pattern, while we're instead seeing the same behavior as described in the initial report for this bug.

[Chrisy Schroth (cschroth)]

I just experienced the same as Irene (comment #4) on our 3.9.4 system. I was attempting to change the pattern for a publication that changed from bimonthly to quarterly using seasons (some quarterly titles use months for their enumerations). Even though the pattern it was predicting was created with, and displaying YEAR Season, it predicted YEAR Month issues instead.

I had created the pattern with a template, so I assumed I'd accidentally chosen the wrong template, using one that makes Quarterly with months instead of Quarterly with seasons, so I went in and edited the pattern to use seasons, saved it, and predicted the issues again. After 3 tries I deleted the whole pattern and started over, but got the same results.

My only option was to predict the first one, edit issue holding codes to display the proper YEAR Season, and then predict the future issues from there.

I can add a further complication with this, that Evergreen doesn't allow you to choose when you want to start your issues. The title in question was bimonthly, and had already had an Oct. 2023 issue several months ago. The new issue, first as a quarterly, is Fall 2023. I began by deleting the pending Dec. 2023 issue that will never be published and deactivating the bimonthly pattern. Created the new pattern and predicted the issues. When predicting, I changed the Predict New Issues: Initial Values to October 21, 2023, which caused it to display an 2023 Autumn enumeration to predict. Despite the fact that Evergreen only recognizes Winter as being in December, never January (some quarterly titles begin their year with Winter, not end it) Evergreen predicted a Jan. 2024 issue over and over. If Evergreen had been following the previous pattern (which is not active), it would have predicted a December or February issue.