Alan Orth comments

Results 201 comments of


                                            Alan Orth

RegexPasswordValidatorIT class is apparently a unit test, not an integration test.

Your proposal sounds good, though it seems the opposite was done in https://github.com/DSpace/DSpace/pull/8924 a few years ago?

Map a Handle to the object path, not the old "handle" path, avoiding double redirection.

Thanks @mwoodiupui. I ported this patch to DSpace 7.6 and tested resolving a handle using the built-in HTTP resolver. [9864-dspace-7_x.patch.txt](https://github.com/user-attachments/files/17227594/9864-dspace-7_x.patch.txt) (only contains the functional changes in `dspace-api`, not the tests...

Map a Handle to the object path, not the old "handle" path, avoiding double redirection.

Another data point for this: we've been having problems with Altmetric not tracking our content since we upgraded to DSpace 7. After discussing with them we realized their bot does...

Map a Handle to the object path, not the old "handle" path, avoiding double redirection.

> A version for DSpace 7.6 is in production here. Running in production on DSpace 7.6 for a week now here as well.

Map a Handle to the object path, not the old "handle" path, avoiding double redirection.

Hey @misilot. Which patch did you use? I manually ported it in my branch of 7.6.3-SNAPSHOT as of a week or two ago and it's working: https://github.com/ilri/DSpace/commit/374069565a9a963c00d21e0de1437eb0dc91b8a4

Map a Handle to the object path, not the old "handle" path, avoiding double redirection.

The patch in this PR is against `main` so there has been some divergence from 7.6 I guess.

Map a Handle to the object path, not the old "handle" path, avoiding double redirection.

I've found a side effect of this patch: collection subscriptions now display the object paths instead of legacy handle paths. Email from before I applied the patch: ``` This email...

Map a Handle to the object path, not the old "handle" path, avoiding double redirection.

Another side effect of this patch is that it breaks the DSONameService, used to find the names of DSpace objects, for example here on the site statistics page: ![Screenshot 2025-02-12...

Map a Handle to the object path, not the old "handle" path, avoiding double redirection.

> @alanorth Regarding the subscription emails: is there a problem or only a surprise? What should we be showing there? Hey @mwoodiupui, historically we showed the Handles here and this...

Map a Handle to the object path, not the old "handle" path, avoiding double redirection.

Hi @mwoodiupui, according to my notes from #9994 the DSpace 6 behavior was to show the canonical hdl.handle.net form: ``` New items are available in the collections you have subscribed...